Slashdot Mirror
Rolling Your Own Business Desktops?
mike asks: "I'm mulling the logic of my company building its own desktop computers. As the IT Manager (plus sysadmin, janitor...) of a struggling-yet-thankfully-still-alive dotcom, money is really tight. We have around sixty ~400MHz desktops which are increasingly showing their age. Acceptable P4 systems from the big guys run at least $1000. By recycling the OS (Win2k), case, cdrom, floppy, and K/V/M, I figure I can assemble a good AMD system for about $600. That's a 40% savings. Is it worth it? The cost difference could very well determine whether this project proceeds or gets put on the back-burner again."
"Some negatives about rolling my own:
- Management: I won't get the special business features offered by some manufacturers. Dell's OpenImage, for example, looks awfully nice. But how much does that really buy me in a company of 60 machines? I don't use such stuff now; am I missing out on nirvana?
- Time to build: Even though we'd leverage Ghost wherever possible, handmade systems nevertheless take time to build, load, & configure.
- Supporting different platforms: Because money is so tight, I can at best afford a capital replacement rate of 25%-33% (15-20 units) per year. That means I'm committing to the support of 3 or 4 different platforms. Having just one platform is great, but how many companies, even ones that actively strive for it, truly enjoy that luxury? I inherited two platforms (Micron & Gateway); support isn't that bad. With proper planning, I don't see why we can't support four.
- Hardware quality: How much can I trust a popular Athlon chipset in a business environment? I feel silly bringing this up because I have a few Athlon systems at home, each with a different chipset, and they've been nothing but rock solid. But I know the lack of a really good chipset has been a large contributor to why AMD's aren't more prevalent in the business world. (well, that and long term bullying by Intel).
- I don't get a proven, prepackaged system that works right out of the box.
- Cost savings. Plain & simple.
- Increased horsepower per dollar spent.
- By choosing my own equipment (mobo especially), I suffer fewer OEM shortcuts.
- I have to admit that I'd enjoy the pure geek satisfaction of rolling out 'my' creation to the company.
For those that are curious, Ask Slashdot did an article on the AMD issue, here.
I see it this way. You are the one that will be working on these machines. You must factor in the knowlege that you made them and know what is in them. Just make sure you get a warranty on all the parts since you will not have one on the entire machine
Maintaining all of them would give you plenty of job security.
Will Microsoft even allow you to recycle your Win2k license on a new computer?
...but Microsoft might be. You might want to take a look at the EULA from M$ and see if they allow the transfer of operating system. Not that I'm suggesting you follow that load of malarky, but it may be a consideration.
Personally, if they're just office type machines. Get Star Office and Linux and see what you can do. Experiment with a couple of your users to see how much trouble it might be.
I really don't see a downside to the project... if you had a few people you trusted to help upgrade the systems, you could assembly line the upgrade and get things up and running in a couple weekends. The only things that I would see as a concern would be the age of power supplies, hard drives, etc. But if you do regular backups, that risk is minimized.
I wish there was a choice that said "Factually Wrong -1" when I mod.
If, two months from now, one of them dies and dies hard, you're on your own to figure out what went wrong, find a replacement part, try to get warranty service from wherever you bought that component, etc.
Most of the majors offer very good service. Often it's just a cross-ship for the whole system, and you're in business the next day with no time invested by your IT department.
Those that actually need them?
I mean really suzy in the phone center has no need for over 400 mhz, I'm striving along just fine on my 667.
Don't forget having to run your own assembly and tech support shop as well. I can usually coerce somebody to come out from Dell and replace my broken (video card, motherboard, CD-ROM drive) with little effort here at work if the need arises and it's covered under warranty. At your shop, YOU are the warranty guy.
Also, factor in the labor costs (which will be substantial), count the amount of time it will take for you to assemble a machine, the cost of ESD straps and mats (you will be using ESD mats, right?), the time it will take to set up an assembly area, and the space that will take up, etc.
I used to build machines for other people (family members, etc.) Now I just tell them all to buy a Dell because the hassle on me to maintain them is WAAAAAY less. The only machine I build myself anymore is my personal box, because I spec out stuff that is too high-end for a manufacturer like Dell anyway.
What's your salary/the salary of the people that will have to build 60 boxes? How long will it take? Are you sure $600 + labor costs + no manufacturer support will be less than $1,000? If not, there's no business case to do it yourself.
-matt
depends on what your time costs the company. If you are willing to eat the extra hours to build the systems then yes you can save a lot of money. In this economy the extra hours put in to save your job might very well be worth it. YMMV but everytime I have done this the first couple of boxes have taken a long time then once I had working with the hardware being used down the time to build went down a lot. I would not worry about support you are most likely better than anything Dell could provide.
Cypherpunks: Civil Liberty Through Complex Mathematics. Those who live by the sword die by the arrow.
Don't buy more processor than you need; It's expensive. You can always upgrade the CPU later if you pick a good platform. You can do the whole thing for about $450-$500 for each box.
Incidentally, I picked the GF2MX because it has good drivers and VERY fast 2D. If you are doing cad or something, get something from matrox, they have a much better DAC. The 3D is just icing.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
One thing you've got to look at is warranty coverage and who will support these boxes once they're made. Most of these components will have warranties, but there's different coverage for each one you'd have to keep track of, you might get a bad batch if you order in bulk, which can cost time and money in the long run. For a business environment, even small business, I'd recommend looking to a solid company that has a good service record and see if you can get a bid war going between two companies who have small business plans to have your business. Bottom line, it's probably more hassle than you'd want.
Use RIS.
You won't be tied to hardware configs (unless you have funky hardware that doesn't have a Microsoft driver) and you can just plug your machines in with a floppy telling the machine to RIS itself (or certain NIC cards.. was it newer 3coms or Intels?).
There are some things that are not fun about doing this, like popping older apps in to MSI's (something I have had difficulty doing), but it pays off in the end.
Will the receptionist who plays Solitaire all day need a new machine.... consider that there are probably only a few folks that would need the upgrade.
My company tries to save money all the time by doing things ourselves rather than buy some solution off the shelf.
They end up paying as much or more and taking twice as long. Problems come up - all the while the employees doing the work are on the clock.
As cheap as some pre-built systems come, I'd be willing to bet that doing it yourself and saving much would be unlikely.
.
It's hard to believe that's how Micronians are made. Why don't we see it right now by having you both kiss one another?
how bout some dell p4 1.6's for $350 shipped?
http://www.gotapex.com/deals.php
Figure out how much you make an hour, then multiply that by the number of hours it'll take you to put together each machine. Then you'll see if building your own really is cheaper.
You should put Linux on those computers... Microsoft sucks.</mindless>
SlashBorg
Well I can talk from some kind of experience.. we had a bit of a botched attempt at trying this one year and decided never to bother again.
:
:-)
Someone had the great idea to buy a load of AMD K6's and some cheap generic 'all in one' motherboards.. our team of 8 or so techies all sat and built about 30 of this machines in an afternoon or two.. but the machines are pretty stubborn and are already very very out of date - we just used old cases complete with 2GB drives which were more than enough at the time. Now the CPU's are still quick enough for office tasks, but the drives are much to small.. and its too much hassle to go around adding new drives and re-imaging.
I think buying complete systems is the best way to go about this for a number of reasons
a) Standardised hardware (makes imaging a lot easier)
b) Probably more reliable (you know the hardware combination they give you IS going to work.. sometimes you can put together a troublesome combination of parts and never get the system working right)
c) Having someone else to blame if the system gives you hassle.. (just call their tech support and get them out to fix it!)
"Hey! Unless this is a nude love-in, get the hell off my property!!"
You may save your company a lot upfront, but be wary of those hidden costs; lots more manpower down the road, no vendor to call, nobody to blame. Plus, you've got to factor in the manpower of doing it.
i did the same thing you are thinking about doing now. I think the man-hours involved with building the machines ends up costing mre than buying complete machines. It depends on the amound of machines. We had 30 to build. All identical, so we just applied the same image to all of them. the long part was building them. In the end we would have saved money and time if we had simply ordered them.
Be mindful of the power supply that the new mother board uses. Many of the new P4/Athlon require a four pin power plug for the proccesor that wasn't on the older ATX's
It was me, I did it, I moved your cheese
Perhaps you should consider Walmart's Microtel PCs without Windows. Assuming you don't need software or monitors, you can get a 1GHz Celeron for $400. The trick is the legallity of transfering your Windows licenses (Which piece of the original computer does the license go with, the hard drive? Can you swap that piece into the new system). [Of course, if you could convert to Linux, that would be cool, but that's probably a separate battle.]
My experiences building AMD systems on my own have been great. They are easy to put together and maintain. The savings IMHO are absolutly worth it and you will most likely come out with better systems than if you went with the big guys, more affordable and better performance. This is how it has been for me and my friends, but we aren't every situation.
Well, are they failing? Or are they just not sexy and shiny enough for you? If they're not failing, why do you think new computers are going to increase productivity? If they are failing, what other bad decisions were made 2 years ago besides choosing substandard office machines?
400Mhz is usually enough for most business apps. Or just buy a new PIII and some more memory. Perhaps you should reconsider what OS you are running.
There are a few advantages of building the computers yourself, but it's not something I would do in your shoes.
Consider first the labor costs. Even assuming you can ghost your software and buy exact matching hardware, you're still looking at 2-3 hours per machine in the actual hardware construction/testing phase. Depending on what you could be making billing out to clients (again, depends on what kind of business your in, and your position in the company), you may loose your cost savings.
Second is system hardware management. You know for a fact that a solid system from Dell or another giant will most likely have every component working together and all the neccescary drivers functioning right out of the box. Most of the time off the shelf components play nice these days, but you never know.
And, of course, there is the licensing issues. If you plan on migrating your current software licenses to the new machines, make sure they all work ok.
Just a few things to think about.
The Internet is generally stupid
If you must, go out and get some low-end consumer PCs and buy a bunch of spares: it's less work than building your own and still very cheap.
Speaking as someone who has to support about five hundred of Some Other Guy's Product(tm), the main issue I'd have with us rolling out so many of our own custom built systems is just that. Systems from some other guy (say, Dell) come with pretty comprehensive service plans that lets me make Dell deal with dead monitor/mouse/HD/power supply problems in 24 hours instead of me having to track down the manufacturer and get him to ship me a replacement within a couple weeks time. :)
If you're already supporting the systems, though, as you make it seem... then this may not be an issue for you. Just find out about RMA policies of your vendor beforehand!
Maxim: People cannot follow directions.
Increases in truth directly with the length of time spent explaining them
I don't know where you are getting $600/system but I can get a 1ghz duron system complete with no scavenging for less than that. I would think w/o software that you could get about $400/system if you really skimmed.
----------
while (alive) { Work(); PayTaxes(); Eat(); Sleep(); }
Bool
Be sure not to get the board and the chip seperately. Buy from a vendor that will install and test the chip for free. You can scratch the die on Athlons, and ruin them or cause them not to clock as high as they're rated.
Also, some Athlons just don't go the speed they're supposed to.
This has happened to me and a couple of my friends, so save yourself the hassle of taking the parts back.
WWJD? JWRTFA!
We have around sixty ~400MHz desktops which are increasingly showing their age.What apps are you running that is causing them to show their age? With sufficient memory, P2-400's are still plenty fast for office apps and email.
Personally, I'd spend my money on a 256MB stick of SDRAM and some time to reinstall the OS.
While I can appreciate the geek factor here, I think you'd be nuts to roll your own systems here. It will eat up loads of your time, overall costing your company more than it would to just pay more for each system. And I'm not just talking build time. When (not if) one of the systems go kerput, you'll end up diagnosing it yourself, RMAing the defective component, replacing it yourself, testing, reloading OS (if needed), etc etc. Compare to getting a Dell or something, where you determine software or hardware. If hardware, it's under warranty, you don't have to so much as crack the case open. Saves a lot of time and therefore cash.
Even if they cost a little more, I think you'll find yourself grateful for a warranty to fall back on. Plus, when machines go boom, you aren't instantly blamed. If you roll your own, any system that crashes will be pinned on YOU, and you alone.
I know that's not a situation that I'd like to be in. Would you?
End of lesson. You may press the button.
It depends what you need them for. I'm not sure where you get the $600 per unit quote from, but assuming you recycle most of the things, I'd say $300 a unit is likely sufficient.
For simple office work, a $50 Duron and something like a $100 moderate quality motherboard should suffice, throw in a $100 hard drive to increase speed, maybe add 128 megs of ram for a little boost too, without topping $300
For 3-d or crazy amounts of compiling, you can probably upgrade to a 1.6ghz Athlon XP and a new hard drive as well as DDR RAM for under $500.
What I would do is build a couple dual Athlon linux servers and compile code on them while doing development and small compiles locally on the current 400mhz machines, but it depends on your application.
--------
It's OK to be social, just don't tell anyone about it.
I think that you might be well served looking at the various desktop refurbished deals on ubid.com as well. Seems pretty decent, especially in desktop (vs. laptop) arena.
Also I know the name may be tainted, but I cannot stress the quality of E-Machines. If your tech staff knows a bit about hardware, their horrible tech support is not an issue. We have about 50-60 E-Machines here, and only 2 or 3 have ever gave us a problem. These PC's are insanley priced and the components are name brand. You can a 1Ghz+ machine for under $800 with a monitor if you look around.
Remember these big guys buy in bulk that is why they offer good prices. Plus most of the time the PC is ready to go (as long as it comes with the OS you want which you can customize with Dell.)
ya know... if you order complete systems not only do you not have to service them but with those other 400mhz boxes you could build one mean cluster... mp3 streaming server anyone?
----------
while (alive) { Work(); PayTaxes(); Eat(); Sleep(); }
Bool
Why is 400mhz so bad for desktop systems ? What are your users' needs ? Must every system be upgraded to a 'blazing' Ghz+ processor ?
Cobbling together parts saves cash initially, but what about technical support and part replacement ? Do you call each vendor for each component when something fails ? How do you prove you bought the part and deserve support ?
Example: buy an OEM system - say, a Dell, and you call them when anything breaks that came in the box. Hard drive, keyboard, mouse, monitor, etc. Are you now going to keep track of Viewsonic, Maxtor, Microsoft (periphs), Xircom, Intel, 3com, Logitech, Samsung, Sony, etc etc etc! support contracts ??
So basically I'm curious as to two things -
Why the need for a processor upgrade across the board, which is what I'm understanding this to be ? You're keeping everything else from the original systems, right ?
Do you have a system to manage proving you deserve support to a dozen vendors ? Will you no longer have support from the original OEMs who built the systems you're canabalizing ?
__
When figuring the cost, you must take your own time spent on the project into account. Time spent building and configuring your systems is time not spent filling other roles (sysadmin, etc). The other side to that coin is that 60 systems will take a good amount of time to build - moreso if you're interrupted every so often with a question or a tech issue. If you get an assistant to run the business operations while you build computers, or vice versa, that's still a tangible cost which must be accounted for. The time saved by getting 60 systems ready to rock out of the box seems considerable - particularly if you're on a tight schedule for this roll-out. Randomly, Phased rollouts might not be a bad idea to start off with - maybe 50/50 or 60/40, as needed - let the hardware that's working well continue to do so for another budget cycle.
Never underestimate the potential of Human stupidity. -Heinlein
Even better than building new systems, just upgrade the old ones. You can easily upgrade the motherboard and CPU for $100 to $150. Save $800 per system
Most business desktops are built around Intel processors. AMD does not have much of a market outside of the enthusiast croud. Many businesses are hesitant to use AMD processors, and you may want to consider some of the reasons behind that decision. AMD processors generally run hotter and require more power than their Intel counterparts. This isn't a big deal for most home users, but in a business with a large number of computers in a relatively small amount of space, the additional heat output and power requirements could become a significant issue. Also, AMD CPUs are not considered to be nearly as reliable for mission critical applications. They may work out fine in the average desktop, but in a server environment where uptime is more crucial, AMD chips simply have not proven their reliability, whereas Intel chips have built their reputation upon years and years of usage in a real-world environment. There is also the issue of application compatibility. You never know what kind of weird bugs might crop up when you're using a CPU that is not 100% x86 compatible such as the Athlon. Intel DEVELOPED the x68 standard, and their processors are guaranteed to be fully standards-compliant. I'm not saying that AMD's are useless. They are certainly cheaper than Intel processors, but like the old adage goes, you get what you pay for!
-atrowe: Card-carrying Mensa member. I have no toleranse for stupidity.
...and I can't really recommend it.
I worked in a 50-user shop, and provided services and equipment to a 200-user shop under contract.
In our case, the only way to get decent specs and meet the client's budget was to roll our own. The other options were too few systems, or systems too cheesy to contemplate. Cheesy as in crap, not as in creamy goodness.
If you go down that path, my suggestion would be to make sure you have confidence in your component choices, and that all your component choices interoperate flawlessly. Any system you have to see again will blow the savings - your first callback or return could be fatal. Make sure you source quality components, and if you're trying to minimize the number of discrete configurations, buy all your components at once.
Spend money on decent cases with good power supplies. Don't yield to the urge to "cheap out" on components that "don't matter" - they all matter. Don't buy cadillac parts, but make sure everything you do buy is good quality, sound, and durable. Keep extra original parts on hand, especially a mobo or two.
Come up with a logo and have the stickers printed - it amazed us how many people would readily accept a brand they'd never heard of, but would never accept an unbranded system.
Your initial problem will be evaluating a number of different hardware options, then settling on those you want to standardize on. Once you get to that point, what do you do with the bastard love children of your prototype period? Don't deploy them to users, you'll water down any faith and confidence your production systems should inspire.
http://drteknikal.blogspot.com/
With all the issues with the VIA K7 chipset, it's natural you'd feel a little queasy about going the AMD route. Also there's the heat death issues to consider. I understand there are now safety measures in place to save an Athlon XP if the chip fan/heatsink fails, but that was not the case with earlier Athlons. But keep that fan on tight...it's important.
Knowledge is power. Knowledge shared is power multiplied.
You may find that your brand name PCs have some strange form factor (i.e., not true ATX). IF so swaping out mother boards may be unworkable. Also how much can you update and still pass the M$ EULA??
Acceptable P4 systems from the big guys run at least $1000
What's the definition of "Acceptable" for your application? You can get a pretty decent Small-Business Dell for under $800. Not top-of-the-line, but good. What are these machines going to be used for, what kind of power do you really need? You might be able to go cheaper than you think, and have support besides.
First I think you have to figure how much you make now and figure how much money it's going to cost you to build these things. If you're the catch all now are you really going to have enough time to be building machines, as well as maintain them. It sucks but people time costs money too, could the money spent on you be spent better having you do something else.
The warranty thing is an issue, make sure you are covered under warranty for every part - there is somethign to be said for having an entire machine under warranty and being able to call dell and say I have a bad drive, I need a new one by tommorow morning
I honestly gotta think you can get new machines for under a grand each. You don't need new monitors, keyboards or mice do you? All you are buying is CPUs. You probably need a P4 1.4ish, 256 meg of ram, 20gig hd and you're fine...that should be had by Dell (I am dell biased sorry) or anyone else for around 600-700 or less.
Micro$oft - I seriously doubt you can role your licenses to new computers. If you upgrade a few components you're probably fine but I am fairly sure it says they are non transferable. At what point does gutting a system to upgrade it constitute a new system - I don't know but it may be an important point if you can't transfer licenses.
(plink, plink).
You didn't mention exactly what these desktops are used for, but I imagine it's the typical office apps/web surfing/email duty that most business desktops are used for.
Given that, what's wrong with 400MHz? I don't see how a CPU can "show it's age" by doing the same tasks it was capable of a few years ago. My grandpa is a different matter altogether. He's a *lot* slower than he used to be three years ago.
I assume these PC's got the job done at the time they were purchased. Have your users' needs changed? Maybe that upgrade to Win2K was counter-productive. My boss keeps buying Dells with Win2K and XP even though he says every day how much he hates it. I have refused to give up my reliable 98SE desktop (333MHz Celeron) because it is everything I need to do my job.
I guess I just don't understand how the performance of a CPU could be perceived to degrade over time.
I have found there are just two ways to go.
It all comes down to livin' fast or dyin' slow. -REK, Jr.
How many boxes are you talking about? You could very well find it would be CHEAPER to get a configured solution and maintenance from a computer maker like Dell rather than hire half-a-dozen techs full-time to build, troubleshoot, and maintain your company's gear.
Dell has the benefits of economy of scale: ie they build and maintain computers at a very low cost because they do it on a big scale. You'd be hard pressed to match that economy of scale at anything beyond maintaining a dozen boxes. That $400-$500 a box difference gets eaten up pretty quickly with the amount of hired hands the company will need to put these together, test them, and then maintain them.
Maybe a better way to go is to learn how to use the OEM Preinstall Kit. Make all the software packages part of the image, then sned the image to the machine. Use DOS boot disks to connect to the network and start Ghost. Your setup from then is minimal (printers, network connecitons, etc). this way you can guarentee the same software versions are installed on every PC.
Also, Seriously look into Novell's ZenWorks. You can roll out software to the PCs remotely, and give the users icons that can re-install the software without a desk visit from you. I believe it's available on Linux (as the server), as well as NT,Netware,Tru-64.
I'm a small company and even though we only manage about 12 desktops we went with Dell. Over the past two years we've had one motherboard burn out and another had a bad video chip. Both were covered unde r warranty and it was easy to get it sent out and repaired. Also, it's easy for us to keep our insurance info together when all the hardware is purchased from one vendor. These are perks for pre-built systems and I think it's worth the extra money for the peace of mind.
Honestly, with the possibility of needing more storage (I'm guessing that 400mhz systems probably had what? 10 gigs or so?) I don't see what problem you have. These are office machines, not servers whose load increases each week. Even then, these not only meet the minimum requirements for win2k (Which, btw, I think is a bad choice), but should be spiffy enough that no one dies of terminal annoyance using them.
Am I missing something? Are these software development systems (where compile times have alot to do with productivity) or maybe web design/graphic arts systems (where someone is bitching for the latest Macromedia tool) ? You've given absolutely zero compelling reasons for such a massive upgrade, with you willingly admit that you are on a shoestring budget. It's a wonder that your dotcom isn't dead like the rest.
another suggestion would be to simply upgrade the ram on all the machines. a 400 with 512MB of ram is more than sufficient for any office task outside of video editing. instead of $600, you would only spend around $100 per machine.
As someone who's been in your shoes, I'd recommend it IF time I'd take to build them is less than the savings.
The beauty of just ghosting another clone and handing it to a user, and troubleshooting the bad PC back in the rear is worth A LOT of hassle. You just need to get 5% more PC than what's on the floor (about 2.5% for replacement and 2.5% for spare parts/bad parts found during the initial build), and be prepared to spend the time to build them out.
If something in an OEM machine breaks, you can be assured that they can get you something to replace it. If you made it yourself and it's been a while (>6 months) there's a good chance the entire component you need is no longer available.
:)
So you put something else in it. Next week something else breaks. A couple weeks later another one. Now you already have four different setups, and the ones with replaced parts will give you trouble if you put a GHOST image back on them. Not to mention the hassle when you have to install new applications or drivers.
I personally prefer the OEM workstations with lots of stuff integrated: video, sound, controllers, NIC with lots of features. And you can be assured these machines will be tested when some ISV who's software you use (Microsoft?) brings out a patch or update. If you have self-made Athlon boxes sitting on all your desktops, what are you going to do when some crucial piece of software doesn't work? Blame the guy who sold you the 60 Athlons?
sure, it maybe cheaper per machine, but at the cost of your time. economically, how much will you spend building these machines, and what do you get paid, if the money you get paid is even close to what you'd gain, then factor in warranties (or in this case lack of.) with all this in mind now, pitch it to your manager, see what they say, so either way, you can't be held responsible, as it wasn't your decesion
--fetch daddy's blue fright wig, i must be handsome when i release my rage
Bang for the buck is always a great exercise to play, but how about maximum buck?
Why $600 per machine? Why not.. $400?
Worst case you've got a power supply, motherboard, CPU, and ram. Everything else (peripheral cards, video cards, networking cards, sound cards, monitor) stay the same.
Best case, you can reuse the power supply.
Go for 800-900 MHz, rather than 1.4GHz.
Go for 266DDR, rather than 500+
So you spend about $60 on a CPU, you spend about $110 on the motherboard, you spend about $180 on 512mb RAM... that's $350...
How much performance do you need, how much performance can you afford, and how much performance can you settle for?
GPL Deconstructed
It depends on what kind of apps you're running but have you considered dumping client/server entirely and opting for thin client (or server-based computing, whatever they call it these days) instead? TCO is much less and so long as the apps aren't highly graphical you can cheerfully run all you need using the machines you've already got. Citrix MetaFrame would be a good starting point (disclaimer: I'm not an employee, I've just done a TCO white paper and thin client came out heaps cheaper than client/server based architecture).
I am a leaf on the wind
My experience has been that when you're too busy to handle your own hardware/software support, you should find a competent local firm who can build machines to your specifications, support them, and provide warranties.
I have found that name-brand systems (i.e. Dell, Gateway, Compaq, etc.) are overpriced, underfeatured, and have a very limited hardware upgrade path.
When you find a local computer reseller who will provide you with the support you need you can get the AMD systems you want with the componentry you want, without the hassle of taking the time to order, build, and load them.
This arrangment is especially valuable if any of your hardware is DOA. The vendor will take care of any returns. You only get working hardware.
Finding a competent local vendor is tough. Everyone thinks they know their hardware and their hardware is the best. It pays to go with someone who has been in business at least a couple years. Talk to their customers and get feedback. Check out ResellerRatings.com for comments on some of the larger resellers.
Good luck.
well, take this for what its worth, but at my last company, I was the company co-op for two years, and hence, low bitch on the totem pole... all new computers from oem's went to full-time employees (never mind the fact that i had worked there practically full-time for longer than most of those bozos). the first four months I was there, i worked on a pentium II laptop ~350mhz (i was doing 3d graphics--that's a woefully underpowered platform). When i finally got my own desktop, they brought me a big cardboard box full of parts--unassembled. so i put it together, and used it the rest of the time i was there. Never crashed on me once.
;), six or seven machines died in the same week and people were fighting over my box :)
all the OEM desktops? well, anytime one of them crashed, they'd snag my desktop for a replacement while i was out of the office and I had to go track it down when i came back. this happened at least once a month, sometimes more frequently. there was this one time, (at band camp
moral of the story? don't be afraid of a generic desktop, just don't buy shit parts. as long as you buy stuff that's fairly easy to get a driver for--nics and so forth--you'll have no problems.
Lease...
:-)
AMD is NOT the way to go for a business (i know from experience). Intel's chipset's are more robust and are proven for business'.
Call up dell, get a sales rep and start a dialogue. price out some optiplex's (or a dimention if you wanna go cheapie). your monthly payment on ~25 dell boxen @ ~$800/ea might be ~$300/mo.
The only way to know is to price 'em out and run the numbers. Is using a whitebox gonna save you money? maybe...BUT! what is the 'TCO/TCA' gonna cost you for those whitebox's as compared to a dell? What about warranty? parts replacement? Drivers when you format? Licensing?
I have to go over the same scenerio with my clients day in and day out; for them a Dell is cheaper.
I am not part of dell; I just preach them, my statements are an opinion, YMMV...of course
Appended to the end of comments you post. 120 chars.
I see you already have a Pro/Con list. Here's a few more thinmgs to consider:
;)
DIY Rollout:
Pros:
You know exactly what parts are, or are not, in your systems, and can (usually) get spares easily.
OS installation/options/configuration is(are) also a known quantity.
Can be extremely cost effective to roll out.
Cons:
Warrenty is provided by whoever you bought your parts from (new), or long-past dead on recycled parts (in most cases).
Tech support? Look in the mirror!
Large Scale network support? See above.
There are a few good reasons for a DIY rollout, but the long term support may be the price you pay later. If you have confidence in your skills, and have a friend or two that can help you out when "it" hits the fan on the next "I Love You" type virus hits, I'd say, Save The Cash, and Go For It!
If your Boss (the guy signing the cheques) want "Guarantees", you just might have to talk to a Big Name company.
Here's a thought: Try selling off your older componants. The extra revenue, however small, might be enought to help get things rolling.
Good Luck eigther way!
You could pull off $600 in savings if you did it right -- I just replaced the CPU & motherboard in a 266 mhz box, kept all the peripherals (although I did buy newer, faster RAM), total cost was $325.
But anyway, you think you'll get $400 in savings per machine. OK, how much do you make an hour (on average, if you're salaried)? Let's say you make $40/hour, roughly. OK, so if it takes you 10 extra hours to custom-build the box, then you break even. Because you'll have to do without a support contract -- which I find is rarely used, anyway -- you may want to factor in cost for that, too. OK, so let's say you'll spend 3 hours, on average, servicing each machine yourself. So if you can put together the box in less than 7 hours, it's a savings. But it's really a good savings only if you can custom assemble those boxes in something like 2 or 3 hours. Then the numbers start to show promise. If you save $100/machine, that's $2,000 a year on 20 machines. So-so.
I guess for me, if I could replace the machines for $400 in parts, that's a $600 savings. If I then could assemble the thing in just 2 hours, that's roughly $100 of "savings" that I lose. That's 20 machines/year X $500 = $10,000. Yeah, that starts to sound worth it. If I was your manager and you came to me suggesting this big plan which would save the company $2,000 a year but suck up a lot of your time, I'd say no, let's have you spend your time doing other things that might have more bang for the buck. But if you come to me with a plan to save $10,000, and you are demonstrably capable of pulling it off, it starts to sound like it might be time well-spent.
My Greasemonkey scripts for Digg &
I would guess a change of CPU is "not an upgrade" but a completely new machine in the mind of M$. Perhaps that's why Intel put that damn serial number on their CPU's...so M$ could track how you moved their software around and get you for more "pay per use".
If you buy a machinefrom a ``real company'', you get support. If a hard drive breaks, Dell will forward-ship you one overnight. If your AMD system breaks...um...you'll have some guy breathing down your neck while you hope CDW has some spares in stock.
In a corporate setting, there's simply no reason to roll your own systems.
- A.P.
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
(Note: I don't work for Dell, but after buying this latest round of systems, I wholeheartedly recommend them.)
;)
I got two Celeron 1.1GHz systems and a Pentium IV 1.6GHz for $588 each (shipped!) Here is a Slashdot post that details my experiences with them.
There was absolutely no way I could undercut Dell on price by building my own -- especially not when you include the cost of Windows XP (preinstalled), one-year on-site warranty, and the awesome cases that open with the press of a button.
It really doesn't make sense to build PC's yourself anymore when manufacturers are offering PC's like this for bargain-bin prices. Plus, you can always recycle monitors as well -- that's what I did with this set.
Building your own will certainly give you job security (as someone else mentioned), but it will also give you no end of headaches. Why doesn't video card A work with motherboard B? And installing Windows 60 times is enough to make even the bravest person run away in fear. Even with a copy of Ghost in hand, you still have the daunting task of putting everything together (and charging the company for your effort). In the end, it's really not worth it to either you or the company. Besides, do you really want to spend the next two weeks testing out RAM and hard drives by hand? Bleh.
Simpli - Your source for San Jose dedicated servers and colocation!
If you really want to save bucks and cut down on licensing headaches, use open source software. If all you need is replacements for Office and Windows, standardizing on Linux and one of the handful of free / cheap office suites will save you a pile of money and a few headaches (no more macro and e-mail viruses, for instance).
AMD is much more cost-effective than Intel. If you do go with Linux desktops, NForce-based boards should be sweet. There are full Linux drivers for all built-in components (sound, IDE, network, video, etc). Otherwise, check out reviews for ECS K7S5A mobos. They're cheap, fast, and reliable. No built-in video, however.
Finally, look around for recent reviews of customer service from Dell and company. With the downturn in PC sales, most PC manufacturers have slashed and burned their support staff to meet earnings expectations. If you know the hardware in all your PCs, it's much easier to do your own hardware support.
I've considered doing the same thing in the past. I've recently built a nice AMD system for myself at home, and I would love to save the money for the systems we buy at work.
But I'd be hard pressed to find the time and tolerance for 60 machines that I end up building and supporting for the life of the computer.
We buy Dell Dimension systems complete with monitor, Win2K and usually Office XP SBE for less than $900 shipped. Depending on your needs, you could probably come closer to $800 each. I know that extra $200 adds up.. but how much is your time worth? More importantly.. how much is your sanity worth?
Rather than buying from Dell/Compaq/HP...
You might want to consider what a local computer assembler would charge you for a generic PC with equivalent specs. Around here, at least (SF Bay Area) there are a number of mom-and-pop shops that consistently beat the large manufacturers on price. It's helpful to have someone local to call for repairs, too.
Also, a lot of these places will do the upgrading labor for you (and test/warranty the machines, as well).
-Mark
im going to guess your doing 3d design? since thats the only reasonable conclusion as to why you would need to upgrade the system (short of upgrading hd/s and memory)
:
the intel 400 with 512mg of memory and a decent (30-60gig) hd is more than adequate to be running any type of business applications (ie: ms word, powerpoint) and even apps like photoshop and illustrator dont need more (unless your busting out some crazy sheeeiiit)
now if your doing 3d design and using tools like lightwave, softimage, 3dmax, etc etc then i would suggest the intel p4s for 1grand - reason being alot of the 3d apps will run better on the intel chip (not that the intel chip is better than the amd) just that the software is "optimized" for intel processors.
also if you build your own systems that 40% you saved will cost you 60% when the machine breaks and there paying you over time to come in on your days off and fix it since you have all kinds of shit you have to take care of on your normal hours. not to mention running from computer store to computer store finding the lowest prices/compatable parts, etc.
in short
1) dont upgrade just because you want "faster computers" unless you really need them.
2) dont build your own period. its only a good "idea" much like communism is good "ideally"
Ave Molech Setting
We have about 50-60 E-Machines here, and only 2 or 3 have ever gave us a problem. These PC's are insanley[sic] priced and the components are name brand.
Here are potential points of failure on an eMachines:
Strictly, the last is not a point of failure, but more an annoyance that is easily remedied.
My suggestion: if you go the eMachines route, replace the boot hard drive right away with a boxed Maxtor and use the Samsung as a slave data drive. Also get a spare Sparkle SFX-L form-factor power supply for each machine...the power supply WILL DIE. I guarantee it. Maybe not this week, maybe not this month, maybe even not this year, but IT WILL HAPPEN.
Also I strongly suggest using the expansion slots to replace the video with something that doesn't suck memory and processor cycles. You can still find decent PCI video cards.
Do this and you will avoid most of the eMachines' endemic problems. It's better to build from scratch, but if you must buy a box with a name, you can do worse (cough*HP Pavilion*cough) than eMachines.
Knowledge is power. Knowledge shared is power multiplied.
You need to budget an extra $100 for the fans.
I know its not communistically correct to suggest this, but both Gateway and Dell both offer 1.2 Ghz Celeron PC's for under $600.
m en _1_dimen_150d.htm
.. Yes. Walmart.
Here is the URL if you dont believe me:
http://www.dell.com/us/en/dhs/products/model_di
It comes with Windows XP, which is a lot more stable that what you have there. you can buy OS-free PC's at walmart for $400
If not you local walmart, check online.
I think u should continue to look at Dell, and if not Dell any of the smaller manufacturers.
Otherwise everyone WILL blame you for issues, even if it is not your fault. It is not worth the risk. Trust me when things go wrong its easy to point fingers.
So the main reason for getting outside built stuff is so you dont get blamed if and when computers crash or users screw up the systems.
Nobody ever got fired for choosing Dell and/or microsoft.
-Johan
I am glad I do not work there anymore and would never do the same again.
Having once worked in a place where all the desktop PCs were of 3 models (all Compaq), hardware maintainence was a breeze. Personally, I would try to sell the boss on picking up last year's models from E-Bay. Compaq has a policy of not accepting returns from distibutors, so distributors tend to have stacks of last year's models. Get a great deal on something never used, and still quite suitable for smaller businesses.
I do not work for Compaq, and I think their home systems are no where as reliable as their office products.
If you know where to get stuff, it can be cheap. Definately check Pricewatch for your stuff. You'll even get discounts on shipping if you buy in bulk from most places!
I suggest going with the ECS line (60 bucks for their AMD/SIS board) and an older duron (older ones can get had for 30/each) and figure you'll spend about 60 on ram (for like 256 megs ddr, if you really need that much).
/. and you've heard this a ton before, but switching to Linux in this senario would save you A HUGE AMOUNT OF MONEY. That might not be extremely true if you've already bought enough licenses of windows, but I highly suggest moving to Linux, for the price at least :P
At any rate, if you can manage to keep things homogeneous you can use a *nix box to copy the intall on one drive to another. Doing a dd of just the partition will get you a system that's not bootable, so you'll either need to boot every machine with a dos disk, or if you switch to Linux just run lilo after using a rescue/boot floppy (or cd).
I know this is
PS: I don't reply to ACs.
I've learned a few important lessons from having set up networks of white boxes. The little problems you have setting up your home gaming box just aren't acceptable when multiplied by 60...
- Buy good RAM. I've never had a problem with the Crucial stuff before, and have had problems with just about everything else. Bad RAM can cause intermittent failures, disk corruption, and a heck of a lot of wasted time.
- Buy a mainboard from a reputable manufacturer with a solid chipset. Don't buy anything cutting edge, get something that is stable and proven to work. Normally I go with one of the more mainstream Asus boards.
- Buy retail boxed CPUs. In my experience, the brand-x bundled coolers WILL FAIL within a year or two. Even the supposed high-end ones. The boxed CPUs don't cost any more when you factor in the fan cost. The retail boxed CPUs come with a 3 year warranty from AMD.
I used to work for a University where I was a sysadmin. We had 70+ machines, and not a single one of them was the same configuration. (No exaggeration, here) We were under the same constraints budgetwise if not more so, and had some real lusers on the systems. It was totally not a problem, though. I'll grant you that we had our fair share of maintenance work to do, but when I left we were rolling our own desktop machines, and people were loving them.
We even found that the newer machines needed less upkeep than the older ones. Start upgrading and by the time you do a full cycle, I think you'll see a reduced workload.
IANAL, but most windows EULAs are tied to the machine. The question is how much of the machine do you replace before it isn't the same machine. If push comes to shove, by the componenet for the new machines and stuff them into the same case. Or if you really feel daring, buy entire new systems and just use the PCI blanks from the old boxes. Whose to say you didn't replace the whole machine on piece at a time...
Play Well
You know, he's going to be at work whether he does this or not. So it'll eat into his /.-trolling time. Seriously, though, you can't simply take the number of hours he spends at this and multiply it by his hourly rate to get a "cost." Oh, you can do it as an exercise but the dollar figure you get isn't as valid as the cost of, say, the mobos.
Personally, I don't see where he really needs to do this in the first place, unless his SETI@Home stats need a little help...
Chances are, the cases for the 400mhz machines don't come with the 300 watt power supplies that are necessary to run an Athlon.
test
the first big issue is the microsoft issue. if the win2k license you have is an oem license, then you can't install it on a new machine (you might have 60 oem licenses and 60 new machiens, but once you get rid of the old machines you lose your rights to the license). if you're using the old machines and "upgrading" them, you might be able to get away with it. i don't know if putting all new hardware in the old case is upgrading or buying a new machine, that's questionable. i think it's upgrading personally, but ianal or overly familiar with what microsoft might say about that.
another issue is the cost of labor. you save $400 per machine, but do you spend the $400 saved by paying people to build these machines? also, how well can you trust the people that will build them? they'd all have to go through huge amounts of testing because unless you build them all yourself with no help, they might be all slightly different. and then there's supporting them yourself and the company has to pay for any support or repairs. dell is great for getting hardware. if something breaks, you tell them, and they send you a new one and you send them the broken old one. also with a company, you might be able to get some sort of a deal on buying 60 machines from them. once again, i don't know the answer to that one, but it's a thought.
how much power do these machines really need anyways? if it's not intense work they would be doing, you can get computers for pretty cheap from dell, and if you already have monitors (and you didn't sound like you cared to replace the old monitors) you can just not get a new monitor.
what you need to do is really look into the microsoft issue first, because if you will need to buy new licenses, then it probably isn't worth building them on your own. and even if you'd just normally transfer the operating system to a new computer at home, the bsa (business software alliance) sounds like they're going to start cracking down on piracy. i've been hearing lots of radio ads from them. so you'll want to make sure that what you're planning on doing is totally legit.
please me, have no regrets.
Seriously, if you can't shake loose more than 10 grand a year for pure hardware upgrades, your older boxes will be falling apart long before you get to them. (And your users will probably rebel and have you drawn and quartered too)
About two jobs back, I worked in an engineering firm with lots of bright tech talent. Management decided to purchase ten new PCs for the programmers. They built one from parts, but determined that the time it took to get it fully assembled and running was great enough that buying the rest pre-assembled made more sense. (I recognize that your requirement to stay on Win2K will have an impact on what is available for purchase now - sometimes life drops a litter of puppies in your lap)
In your 600 per box budget, does that include the costs of the time it will take to assemble the boxes? If not, you had better reconsider, and recognize that when you bring up the new system with the old hard drive, it will most likely complain quite a bit over the changed hardware configuration - so allow an extra two hours per box to be on the safe side in tracking down all the nasty driver changes that no one ever thinks about. (I've done this sort of thing over a dozen times - sometimes you wind up reformatting and installing from scratch)
Why are you recycling the cases?
1> The old 400mhz boxes may well be AT and not ATX form factors- if so, you will need new cases
2> Cooling. If you go with AMD, will the existing cases be able to provide enough airflow?
3>Power supply. you should at least replace the power supplies - they are still the single greatest point of hardware failure in a PC and the ones you have are all probably at or near the end of their service life anyways. The newer systems will have greater power requirements, especially if you go with AMD. Nothing like an increased load to bring a "mature" power supply to its demise.
You either believe in rational thought or you don't
Buy real business PCs with support contracts. Work with one vendor, so that all your PCs are theirs. When there's a hardware problem, it's not your problem. They'll be forced to provide some sort of continuty when they can't get replacement parts for an older but still-covered PC, which is not at all the case when you decide to standardize on a particular motherboard/NIC/video combination, and then it gets discontinued 3 months later when you've deployed 50% as many machines as you'd planned.
:-P) than enough to retire on. Remember, your one criminal act had better be unexpected, and lucrative enough to let you retire in a foreign country with no extradition agreements with the US.
Even if you're planning to be dishonest and pocket the $400 apiece for 60 PCs, it's closer to a quarter's salary (half-year post tax
As for OS licenses, bite the bullet -- it's sad but true, but Microsoft owns the business market, and it's not going to get any easier for business users to do things like recycle licenses. What are you going to do when USB2 is the standard, but W2K doesn't have drivers? Or when your new Hammer server has to run XP Advanced Server with CIFS2, but there's no CIFS2 client for W2K? They're masters of making it not worth your while to try to make the old versions work. You may want to make a statement against Microsoft, but making it at the expense of your company's productivity is no way to do it!
Hate stupid software on freshmeat? Laugh at
How about:
. jsp?motherboardId=21
0 chassis.jsp
...
http://www.viavpsd.com/product/epia_mini_itx_spec
and
http://www.viavpsd.com/products/epia_compatible%2
I'm considering using these myself. at £100/$100 price margin they really rock
I'm about to evaluate these beauties, they seem to really be quite neat
This was the big deciding factor for me. The initial capital expense is lower for rolling your own, but there's a lot to be said for picking up the phone and having a replacement mobo shipped to you overnight at no additional cost.
Besides, with the way vendors change their product lines these days, you'll be lucky to get the same box next year that you standardized on this year.
As for AMD, I've been using Athlon-based machines from Micron PC and haven't had any trouble related to the platform. They're about $300 cheaper for the same relative power and I haven't noticed any problems that the Intel-based boxes woudn't have had as well.
In the end, it's up to you. You might also consider that you might not always be the computer guy at this business, and whoever follows you will have a harder time learning about the quirks of your home-rolled systems than they would learning who to call to get the pre-made ones fixed.
DON'T DO IT!
Always buy a machine that will meet your specs for at least 3 years out and two or three knotches below the "Most Awesome" (can you say price point?"). If it doesn't meet the job either give it to someone where it will or get rid of it.
What a waste you say!? Well three years ago hard drives were 33 now they are 133, ram was 66 now it's 333, video cards were 2 megs, now they are 64, CD roms were 4x now they're 48x+! The only semi-stable part of an old computer is the power supply and even that is not stable.
I suggest you ask your accountant about the tax implications of rebuilding (keeping some real old and slow hardware) v.s. the tax writeoff from giving to charity.
Build it yourself or buy from Dell, doesn't matter just forget about the retro!
Yes, if you can keep sufficient staff to handle the workload, and remember you'll need some space to do the work.
But I've always found it better to use clones, and the more control we had over the design, the better. The real savings come from being able to set up all your systems consistently so that you can manage them more effectively. You might have to deal with a variety of hardware (video cards and NICs are hard to stay consistent with, they change every six months or so), but you can account for that.
Using brand name systems, even if you stick with one brand, involves extra work, because they change so often now. You always end up managing a heterogeneous environment that was designed by a marketing department. And desktop PC's are so disposable that any extra warranty you'd get on a brand name is going to cost you more than it's worth. Much easier to just replace a hard disk than to ship a PC out for service.
Your biggest cost is staff, so if you use the opportunity effectviely, you'll make everybody's job easier, and the hardware savings won't even matter. It will also shorten your lead times if you set it up right.
Also, be careful that you don't over-use spare parts, thats' one problem with having a lot of new hardware hanging around. My motto is, "If we have too many spares, we'll use them." Sometimes you swap out a hard disk, throw the old one in the "we'll test it someday" pile. But if it turns out that wasn't the porblem, then you've pretty much throw away a hard disk without realizing it.
- You can get a reasonable deal from a Dell or IBM for 50 PCs. This includes putting your own image on the drive, support, a decent salesdroid who will likely help with any issues down the road, a sturdy warrenty to back your purchase up, lotsa help in the drivers & spares market, etc.
- You can get 50 PCs assembled at ye local screwdriver shoppe for about what it would cost you to build-your-own but insofar as support & such you are own your own (unless it is some gross defect they can return to the manufacturer.)
- Or you can do it all in house and assume you've got the time to do it all, keep up with everything, and of course document it all in case of a proverbial bus hitting you.
My own argument would be if the business is a real business it should invest in its tools that are a critical part of it's operation. If this eats into the other budgets tough - employees need a roof, lighting, and decent computers. Trying to nickel & dime on hardware is foolish because you invariably end up with a herd of increasingly quirky systems slowly becoming Frankensteined. Unless the tech support (you) is free they're going to end up spending any savings in your time as well as the downtime of the aging & rebuilt systems plus the increasingly irate rest of the staff.Put this all on paper, generate some good estimates of costs & time allowences, failure rates & resolution times then present it to the CFO. Even for a company in a cash crunch these are generally compelling arguments that are well understood by the numbers folks.
They they don't bite then ask yourself if you want to hang around babysitting these monstrosities as the rest of the world moves on?
I don't read ACs: If a post isn't worth so much as a nom de plume to its author then I wont bother either.
Comment removed based on user account deletion
-atrowe: Card-carrying Mensa member. I have no toleranse for stupidity.
Congratulations. You just made my day.
One ring to rule them all. The (_O_) in Goatse.cx
I would buy from a trusted local store. You might end up spending a bit more, but at least you can go exchange a component without having to worry about S&H and delays.
The benefit of paying less for your components means that you can keep a spare set around to fix any failures that may occur. This will give you a much shorter time to repair compared to invoking a warranty and waiting for a replacement machine.
You mention ghost. There's also Partition Magic's drive image program. There's also using any of the free cd-rom imaging programs to make an image of the final install. Slap that image on a harddrive and since you'll have the case open anyway, just put that hd in as a slave and copy the image over. There's a few things you can do since you'll have the case apart.
One thing you are factoring in is your time. You will be using it to do these upgrades. You know what you are paid and you know how long it will take you to do the upgrade. You do the math. 2 hours a machine can easily shave a hundred bucks off the price difference.
Yes, your time is a sunk cost for the company, but your time is valuable and could be spent on other projects which must now go without you.
Plus you've got the added cost of ordering multiple parts from different vendors, tracking these parts as they come in, etc.
I'm not saying you won't be able to save money, but be aware that there are these hidden costs as well.
I've got a 700mhz athlon at home, and a 1.4ghz athlon at work. They've both got 256mb ram, except the 1.4's got DDR ram. I've got 2 30gb 7200rpm drives at home and a big 60gb 7200rpm drive at work. Everything else is pretty similar, except the odd part that happens to be slightly faster because technology apparently moves fast, and the home machine is 2 years old now.
When I go home, the only problem I ever have with the machine is that it runs WinME (sound card is not compatible with WinNT/2k/XP or Linux), but I dual boot. I run Linux at work.
For 99% of the stuff I do on it though, including running a local PHP/MySQL/Apache server, leaving SETI@home on in the background, testing pages in IE, Mozilla, NS4 & Opera, plus having Photoshop open half the time, listening to Winamp, sometimes downloading some stuff off Gnucleus in the background as well, oh, and hitting record whenever I get a new idea for a tune and want to get it down in Saw32 and Fruityloops, I swear the machine acts like a newborn. Just blazing. I never notice the age! Of course, I've got good specs all around, which helps a lot, but assuming you're only 300mhz away from me, and I can't tell the difference between my two machines with 700mhz difference between them, yours shouldn't be aging all that noticeably.
Perhaps you should up the ram or try things like that before junking them all.
putfwd.com - 1GB Free file storage with a twist
Your company name, address, phone number, and you and your manager's contact info. We will be glad to assist you in any licensing issues you may have, especially in the area of OEM licenses. Glad to be of assistance.
Sincerely,
The Business Software Alliance
I tried every decent and legal way I could think of to resolve the issue w/the business before I rented the chicken suit
It's like this:
If you look you can probably find systems (not THE latest systems but good high powered systems) for around $600.00. Forget Dell, Gateway, and the like. You are paying for the name more than anything else. Look for those hard to find small companies looking to make some money because they are getting squeeze by the larger companies. Their products are just as good and you can take out everything you do not need before purchasing the base system.
Example: When 850mhz systems were costing $1,000.00 to $1,300.00 I bought two stripped down systems for $450.00 each. I used all of my old equipment but bought new CD-RWs. Now we have some very nice systems which work quite well.
The problem you will have now is that the big name players have taken over in many respects those sites which used to find you great deals on various items. PriceGrabber is one such site. It used to feature many small companies selling computers but now all you get is IBM, Apple, Dell, Gateway, and the like. So now you have to hunt. Try Google and use phrases like "+500.00 +pc +desktop" to help eliminate clutter. (The plus means to use AND instead of OR when querying Google.)
If anyone else knows of a good site to look for cheaply priced computers - they should list them here.
Later!
Though you'll save $400 per box, the noise that all those AMD-powered boxes will generate may not be worth the savings.
I'm not sure how they do it but Dell boxes are extremely quiet.
I own and operate a small dotcom like business and we always bought the bits for our machines, and built them ourselves.
:-) Get someone else to burn it in - it's a waste of your time.
We bought the best components, big cases and were able to ensure everything worked as desired. But reliability is a BIG issue.
We recently stopped this practice and decided to buy from a small but reliable company (armari.co.uk). I bought a test machine (dual amd 1800+, 1GBram, etc.) and the build quality is amazing...we are now purchasing these machines (plus dual monitor) for all the team.
It's a big relief knowing that I can just call someone and have it fixed asap. Armari even provided named Win2k login, partioned the way I like, and system rescue CD's that in 10 minutes put the os, drivers and configs all back to factory ship.
No looking back to the dim and dark days of spending hours trying to get a SCSI card to boot a CD
I am currently typing on a brand spanking new Dell PowerEdge 500SC with a 1.1G Celeron that I bought for $550 - a $250 Mail in Rebate. Total cost was $300. Sure, it only had a 40GB HD and no sound card, but let the users provide/screw that up. This stock box beats out my HP 733 with 640MB RAM.
B
Flamebait
Serious inquiries only.
to give me all the ~400MHz mobos left. They make great firewalls. ;-)
As I am sure many people will mention, a simple way to look at it is as a question of how much your time is worth. If it takes you 6 hours more to setup each custom built machine than a prepackaged machine, then is your time worth $67/hour to the company? You also have to compare this down the line. When a machine's hard disk, etc. dies, is a quick call to the box maker easier than spending time dealing with the drive manufacturer?
Of course it isn't quite that simple. This last point used to be a big reason for me not to build my own machines. Flaky CD, noisy hard disk, sticking keyboard? Call Dell and a new one will show up tomorrow morning. However, in the last year or so Dell's service has gone to crap, so now I spend an hour on the phone with them to get a dead CD drive replaced. When the cheapest CD drive at newegg is $29 shipped, it is more cost effective for me to buy a new drive than deal with Dell.
There is also the issue of finding a company that will build the machine you want. I want to spend $1300 on a dual athlon. I can't get that from any of the big box companies, and the smaller companies often have markups too large for me to swallow, or don't quite offer what I want. I am sure I could work with many of them to get just what I want, but by the time I have done that I could have chosen the pieces I want from newegg, mwave, gogocity, etc. 1-2 hours of my time to put the pieces together is worth $50-100, it isn't worth $500.
I find the balance tends to go in cycles. For a time I can build boxes better and cheaper than I can buy them ready made, and then for a while I can get what I want ready made for only a trivial markup. The combination of being in the "I can't get what I want" and the decline in customer service I have experienced recently unfortunately puts me in the building boxes phase.
Of course this only applies for boxes I care about and will have to support. Any box that even if I helped buy, I don't have to support, I just order ready made. That way if something goes wrong I just shrug and tell the user to call Dell, or whoever.
You also might look at the no OS boxes from Wal-mart. there is an article at NewsForge about setting Linux up on one. Spending $450 (including upgrading the memory) to get a 1Ghz Duron all put together isn't bad. Just wipe the commercial OS from the drive of the old machine, and install it on the new machine.
1) You have Micron and Gateway. If these are standard ATX cases, swap out HD, MB, CPU, RAM. This should keep you legal with the software license. Since you have the original case, these parts are upgrades to the system, not a 'new' system. This is very important if your company is using the OEM license and not a select agreement or retail license.
2) Did you know you can't ghost an OEM license across different PC's? I've run into this at work, even though we do have a select agreement, we cannot take the image that comes on a Dell PC, add our software to it and ghost it back down to multiple PC's with the same hardware. Violates the terms of the EULA. We actually have to blow away the Dell pre-installed image, load our select software CD's and then use the Dell website and driver CD's to get it back to where it was before.
3) Have you considered just upgrading the processor and RAM? on a PII 400, you should be able to upgrade that to a PII 1GZ (possibly higher). Not quite state of the art, but for most office tasks, I can't tell much differnce between my old 400 and my new 933.
I'd say that the best and least expensive option is doing nothing.
Seriously, unless your line of business requires you to run heavy duty applications, there's no good reason that your 400 MHz machines won't do just fine.
Are you sure that "showing their age" is not merely a symptom that could easily be cured by either upgrading your network or adding some memory to those old systems?
On a personal basis, I've been seduced by the low costs and good performance of the recent hardware, but like many others, I found out that implementing a solution like that chews up far too much of my time. [I can't tell you how I hate to have learned that reseating memory modules sometimes makes a difference!] In the future, I'd rather go with an almost turnkey system made by a manufacturer with a reputation for reliability. My time is too expensive to do anything else.
Now you may have enough systems that you can recover your time investment: eg, you find out that the power supplies on those cheap cases go out and give weird symptoms that you learn to recognize. Time to fix problem first time: 2 days. Time to fix problem second time: 2 hours. Time to fix problem third and nth times: 20 minutes.
Shoot, if the systems you have are old and common, you might be able to pickup some backup reserves for parts or hot replacement systems for next to nothing.
If it ain't broke, don't fix it.
"Provided by the management for your protection."
Remember they're running Win2k, not Linux =)
I say go for it and build your own but have you checked out all your options, like
Dell Outlet
Dell Optiplex refurbished systems
When the system I built started overheating I had to come in on my weekend off to install additional cooling.
When the Dells we have fail (very rarely, mind you), I get a call at home and I tell the office manager to call Dell.
Guess which option I prefer.
One thing to consider is whether your existing power supplies are up to providing nice, stable voltages for your AMD chips, which can be a little fickle.
Factor in time (yours) which can be worth a lot, versus just buying new boxes which will be "good enough".
What do you run besides Office and IE? If you're not planning upgrading your software, why do you need faster computers? Have you considered performing a clean install on the boxes?
I used to work for a shop of about 50 computers. We considered all options when we needed upgrades and new computers. In the end, we determined that a $600 Celeron Toshiba desktop machine was better than trying to roll our own and have compatibility and quality problems. However, we determined that rolling our own servers was VASTLY cheaper than trying to buy them. We went with straight Intel, right down to the motherboard, which was the most stable at the time. Built 3 new servers in just a few days. Most of the components had already been through a burn-in test, but I put them through my own anyway. Last I heard the fileserver had been up for > 365 days, and it was Win2k machine. If you do decide to roll your own, make sure you've got some extra equipment to replace dead stuff. Memory, motherboards, processors, etc. Then you can stand to wait that extra 2 weeks to get the replacement parts, because you are just waiting for replacements for you spares. In our 50 person shop, we always had a machine or two lying aside that could be swapped in for a problem child at a moment's notice. That allowed the users to maximze productivity while I diagnosed the problem on my own time and/or waited for a replacement. Roaming profiles and a good backend setup allowed a virtually seamless swap of machines.
Asset management is critical.
If you do decide to build them yourself, do a huge favor for your future self - keep a hardware inventory of what components are in what machine. If you do the planned 15-20 machines a year, the hardware in your shop is going to span 4 years, which is an immense amount of time. Considering that by the end of the fourth year, you will have done maintenance and replacements on the 1st year machines, you are going to have a huge mix of parts.
Grab up something like IRM so that you can keep track of what is where and which machines are having which problems. It will take a little more time to setup initially, but you will thank yourself profusely for it in about two years (or less).
I wouldn't be so sure that it is legal to recycle the Win2K license. It's most likely a license "for distribution with a new computer only".
Depending upon your organization, and what stuff is done on these machines, I would consider lowering TCO (total cost of ownership) by phasing out desktop units with servers and dumb terminals.
The benefits of a back office full of servers, and a series of dumb terminals are:
1. Terminals are cheap and easy to replace
Keep a couple of spares for when one goes
down.
2. Terminals are usually all identical, with
no issues of drivers, etc.
3. Central management. When time comes to
upgrade the software or to do a virus scan
it only needs to be done once. Not to
mention security.
4. Central backup....
5. Scalability.... Need more people, or more
processing power, upgrade a server, not
20 machines. You can do things like fiber
channel, clustering, and gigabit ethernet
much easier.
6. Geek factor (never forget the geek factor)
Servers are way more fun to fool with than
some old windoze machines.
IMHO, most organizations would realize a lower TCO with a properly organized server farm and a set of dumb terminals. The only problem is software licensing. Under Linux/Unix terminal servers are unlicensed, and you can have as many clients as your hardware can handle. Under Windows/Citrix you have to think about licenses, yuck.
my $0.02 worth.
Time to build: Even though we'd leverage Ghost wherever possible, handmade systems nevertheless take time to build, load, & configure.
Yes. But make damned sure that you're building them as an assembly line. The principle is that building a second one will only take 50% more effort than building the first; the third will take only 33% more effort than the other two, etc. Whatever old Henry Ford's theorem was. It works.
Set aside a room where no one else will bother you. *GOOD STATIC CONTROL* is mandatory. Do all stages of assembly at once, that way you're not wasting time fumbling back and forth for screwdrivers. Get going at a good clip with quality cases, and you should be able to assemble 100 systems/day - but that assumes you have *everything* where you need it when you need it, there's good padded shelving, and you've got a grunt taking care of taking cases out of boxes for you. It also excludes software load.
Just make sure you get a warranty on all the parts since you will not have one on the entire machineAbsolutely. But, assuming a competent builder (ie. not blowing processors with bad jumper settings or blowing boards by not having them seated right), the parts themselves should be pretty reliable. If you're buying good stuff, the biggest source of problems will probably be static handling.
Keep in mind that a modern memory or processor chip has literally millions of CMOS transistors. CMOS transistors have an incredibly thin layer of glass between the gate input and the source-drain circuit. A voltage applied to the gate influences the flow of current through the source-drain circuit. Trick is, the layer of glass involved is so thin that you can punch a hole in it with 30V. Next trick is that static electricity generates kilovolts (thousands of volts) with sufficient current to blow holes in the gate layer, but be imperceptible to you.
All it takes is one transistor out of the millions inside a modern chip to be defective and the computer will crash apparently at random... you know, when Windows VMM writes a 0 to a memory address and gets it back as a 1 later on... BSoD. Kernel Panic. Choose your flavor.
Wrist straps, static baggies, conductive floors, grounded workstations are *crucial*. Dell, Compaq, Asus and Abit spend millions of $$ on these things, and for similar reliability, you should demand the same standards every step of the way for your home-rolled machines. Make sure your computer store hasn't "helped" you by opening the static baggies. Write that one into the contract with the computer store. And make sure that the hard disk drives are still in their packing "egg-crate" things. You really don't want a box with a stack of hard disk drives. (Western Digital had a great video on hard drive handling floating around the 'Net, you should view it if you're building en masse.)
Fire and Meat. Yummy.
The major problem is post-install maintenance. If you know your way around troubleshooting a PC then it isn't a big issue, though. The only catch is keeping track of warranties.
To do that yuo need a big database project and the ability to track eahc individual item (since you'll likely buy a few dozen parts at a time, instead of one huge buy).
To take care of that problem, here's what I do: I ignore the warranty past the first 90 days. If it fails in that time then I'll bother to get a replacement. If it fails a year later, then it isn't worth my time to pursue the issue - my time alone is worth more to the company for two hours of work than a single CPU, motherboard, HD, etc.
So you need to decide up front what you are going to do in the post install world before you install.
As a side note, for my corporate workers, I build celeron 1GHz machines - $75 each for the mobo and chip, a GOOD case and a few fans for $50, 128MB for $30 or so, and a $65 HD (10G - everything usefull is on network). They don't need high performance (well, most of them), and network speed actually affects more of what they do that just about anything else I can do with their system.
-Adam
I think while it might look good on paper, it will be a major hassle in the long run. I know I sound like a broken recond, but Dell has some incrediable deals and service is the best of all the major players. Here is a 499 1.7gig Intel that should comparable to what you could do if you tried to do it yourself. http://www.dell.com/us/en/bsd/offers/specials_3x_d imen_outrageous.htm
...a stick full of 256MB PC100 SDRAM.
I've had a PII 400 BX system with 196MB ram, and the biggest performance boost by far came with swapping out the stock 5400 HD for a Barracuda.
It ran Win2k very well, (fast boots) and now runs XP for my parental units several states away with zero problems.
You could do both for ~$200 USD.
About in your shoes. I was a one man shop, and supported 100 machines, plus servers. I replaced most of the desktops with machines leased from Dell. Great support, new machines every two years, and you don't have to buy licenses through Dell, I got all mine through CDW because they offered a better deal. Building your own machines is nuts, trust me. I came from a company that hammered out whatever a customer wanted, which didn't always work well. Dell has already figured out which works best and sells them as such. I have a video card go bad in a Dell I call Dell. I have a video card go bad in a machine I built, I have to call the video card manufacturer, and then they'll blaim it on the motherboard, etc, etc...
My opinion, lease machines through Dell, sure it 'costs' more in the long run, but it spreads the money out so you're not dipping too heavily into reserves. Servers are up to you. I custom build each free OS box, as in my experience when a free OS box goes, generally it IS hardware, so no guessing, whereas with windows it could be either software or hardware. For the windows servers I used Compaq.
The only tip of advice I can give you about leasing machines is to make sure the machines are spec'd out for the next two years. As our engineering department grew, they required more memory and more CPU, requiring us to lease more new machines and cycle the older engineering machines down to accounting.
In the end, the machines that were destined for the scrap bin, were auctioned off with a silent bid to the employees.
i'd rather roll a nice phatty
-- ganja man
What you are talking about can be done quite easily, the important issue to becareful of is to get some high quality components. Remember high quality != expensive, just older.
... (you get the idea), then the store threw assembling the systems in for free. The really nice thing about it was the machines we got from this had all standard parts, and we knew exactly what was in them. It made finding drivers, and installing different OS's much easier.
At a company I worked for our team couldn't weren't allowed to purchase computer systems, but needed new machines. The manager went to a small local computer store and bought 30 cases, 30 motherboards, 30 videocards
The biggest complaint I have with Dell, et al. is the use of "custom" solutions. Old Compaq DeskPro's were the worst things I've dealt with, I absolutely hated having to find the proper EisaConfig disk for the things everytime I added RAM. I _really_ hated the time I nuked the invisible harddrive partition Compaq had put EisaConfig on.
Don't get systems with special cute management "features" they all to often break with a minor OS update, and the vendor rarely if ever keeps the systems upto date.
I think I'm missing some critical component of your question but what exactly makes a 400MHz P2 system slow with Windows 2k and why exactly do you need new systems? Upgrading is great and all but do you REALLY think you need to at this point? I've got Win XP running on a K6-2 with gobs of RAM and it runs great it has a slow hard drive which makes app loading a tad on the slow side in some cases but most of the time it doesn't appear much slower than my Athlon XP 1700 (in anything not dreadfully processor dependant). For most things I wouldn't consider these systems slow unless you're doing heavy graphics work or doing a lot of compiling or plain number crunching.
If you insist on upgrades a good strategy is to stagger the upgrades spreading the process out over a longer period of time. Grab a couple new systems and get them integrated and slowly but surely phase out the older P2 systems. If possible recycle them to take the place of more specialized systems like file servers, firewalls, domain controllers and the like. If you stagger the upgrades you eventually get more for your money because of continued development and you have time to build or merely integrate all of your systems. Getting five workstations meshing with your network is much simpler than getting 50 meshing all at the same time. If you're building them you'll definitely appriciate not trying to do it all at once.
Instead of building them grab refurb units from the major vendors, considering the number of high power systems purchased in the past couple years you can end up with a speedy system for a pretty good price. It makes ghosting a single drive image a bit more difficult than having a homogenous system but it saves you a bit of time. Besides how expensive is it to buy a pack of CD-Rs to burn images for particular families of systems? Some organizational elbow grease and this process is not too difficult. I like color coding systems to make it easy for tech monkies to find the right CD to image a system with. It's lowtech but it works. One color for the manufacturer and a second color for a particular model or family. Just organize your CD images by color and it works fine usually.
I'm a loner Dottie, a Rebel.
I manage desktop support for an office not much smaller than yours. For budgetary reasons, I started putting systems together myself instead of buying from Dell/Compaq. It's been a tremendous success for me. Here's why:
1) I like picking the components myself. The components I choose are almost always better (and cheaper) than what Dell or Compaq would choose for me. We got some name-brand systems not long ago that had no-name network cards in them that the installed OS had trouble supporting. That doesn't happen when I buy my components.
2) I can pick components based on the needs of the user, not the need of the ISV to make money off of me -- so the receptionist doesn't get an Audigy and a GeForce 4. I usually get the cheapest video card I can (which is still more than enough for Word/Excel/E-mail) and sink the money into RAM. I'm spending less than I've ever spent per system and my system standard has 512MB of RAM.
3) I'm always one step ahead when troubleshooting hardware problems.
4) I'm able to purchase OEM licenses of the OS and Office suites from the various discount vendors (newegg, etc) because I've bought hardware. That saves real money.
5) I _hate_ Level One tech support at every computer company. Spending an hour on hold so a 17-year-old can treat me like an idiot and not solve my problem is just a waste of time. Onsite support has been similarly disappointing. I don't miss it -- I get the support I need on the web and from Google Groups.
6) I've been doing support for a part of my job for a long long time. Putting stuff together is much less boring than buying a pre-built system. If I see one more "getting started" setup poster, I'm gonna cry.
7) Having homogenous systems isn't what it used to be. Time was, you needed to make sure all your systems were from the same vendor to make sure your apps would all run, your network cards would all behave, etc. That just ain't so nowadays. Frankly, Windows doesn't really care what it's running on (once it's running).
8) Don't even get me started on how much better Linux runs on systems I put together!
I put several Athlon systems together using pretty basic components -- recycled cd-rom, floppy, network card, etc. -- and it's been terrific. The systems are louder and uglier than they'd be if I bought them from Dell/Compaq, but then, I paid half price for them.
this question is pointless. Never underestimate your PHB's ability to ignore everything you say, spend three times as much money, and end up with something that doesn't work.
We (I) just replaced the 15 workstations in my small office here with these. Gateways. P4 1.2's. 20 GB HD's. 128 MB of mem. $600. Even comes with a monitor ($70 less without). XP Pro installed, so no windows licensing issues. And your supporting all those artists, dude!
Th
I'd check the form factor real close to make sure the new main board will match the power supply and case size. That is of course if you you need a new main board, chances are you will, if you are moving up to a better cpu. Even so, a new case, cpu, main board, DDR ram and a much bigger hard drive ran me about $600 retail, and that was about 9 months ago. The system has been real stable. Good luck!!!
For example, $360 for a 1.6 GHz P4 w/ 3 year on-site service.
They're putting dimes in the hole in my head to see the change in me.
Setting aside all the considerations of what is easier and what is the least expensive ... etc.
The thing you really need to take a serious look at is do they REALLY need to upgrade?
My personal opinion is unless it's so old that it can't run the core applications that are required it isn't worth upgrading.
That's not likely to be the case. So I wouldn't do it. The money is better spent on something that will have a real return to the company.
Now if we are talking about machines where you just don't have enough memory, disk, processor to do the bare functions....such as running Visual Studio, a CAD program, or massive graphical work of some sort.....then yeah. Get an upgrade and do it right. Buy that new system and pass the old system on to a better use or sell it off to recoupe some of the cost.
------------------ Still Surfing
My 1.4 Tbird at home runs a "cool" 60-64 C with the case off, a fan sucking air off the board, a 1.5Ghz+ rated CPU fan and Arctic Silver 3 compound spread like butter between the chip and fan.
No way I'd use AMD on a at work machine stuck inside some box that'll be beneath someone's desk being used as a footrest.
http://slickdeals.net ---- has a lot of good deals on PCs a lot .. P4 1.6ghz with all the stuff cept a monitor for like .. $400
not bad stuff i guess
Answer these questions, and you will have your answer.
Do you pay for a maint contract? or do you maintain your own systems? maint contracts will cost more on home grown systems.
do you have folks with extra time on their hands? If not, and you will need to hire to get the job done, you have to take that money into account.
Do you have a corp license from micro$loth? Read the w2k EULA. changing motherboards probably will cost you another license fee.
You can actually build them cheaper than $600, especially if you are reusing a lot of stuff. I buy ABIT's for under $250, and those are the best mb's going, bar none!
Just three more hours seapeople and you can finally take me away from this crappy God Damned planet full of hippies
If you sell the old computers on e-bay you may be able to make up the $300 difference in buying the computers new. Also why do you have to upgrade all of them. The old ones are probably fine for many of your users. Why do you need a 1-2ghz p4 to do word processing? You can get a complete 1ghz celron from e-machines with win xp for $400 and still sell all the old ones (or give them to your employees to take home). For $700 you can almost buy two for each person and when one quits working just throw it away if you can't think of anything better to do with it.
I work for a school district that had really old machines setup, such p90's etc. I came in and spent about $40k on 200 new machines and built them MYSELF. It has been pretty good. I would suggest getting some people together to build your machines. Have six friends come by, have them each build 10 machines and use one to play 2v2 warcraft 2 when u guys are done :0). This is an economical way of doing it, but its lots of work on your end, software and integration from the employee's old machines. If you have any questions, drop me an email at whoinow@hotmail.com
hang out at deal sites and get something from dell. here's a deal that was posted a few days ago at bensbargains.net
t wallet.com
OptiPlex GX240 1.6GHz $358 shipped from Dell Business
Pentium-4 1.6GHz, 128MB SDRAM, 20GB Drive, 16MB Rage 128 Video, XP Home
Integrated 10/100 3Com NIC, 48x CD-ROM, Integrated Audio, 3-yr Warranty
This was done using a 20% coupon and a $100 rebate. Dell does these things kind of regularly and if you're diligent you can jump on one of these. The Optiplex is Dell's Business line of machines and the 3 year warranty for their business machines is pretty good. even without the $100 rebate those machines will only run you about $450. Sure it's not the best marriage of all components (sdram instead of ddr, lower end video card, xp home) but for that price you get some pretty good warranty and they're decent work machines.
If you wanna go for even less money, try compgeeks.com's specials, Get a grade A celeron 700mhz with 20gb hd, 64 or 128 mb ram, cdrw, keyboard, mouse, nic, modem, and it'll probably run you $300, just recycle your old monitors.
hang out at deals sites like:
http://www.bensbargains.net
http://www.fa
http://www.anandtech.com
It's funny to me that someone can take the time to write out a long, descriptive article about their problem, and then half of the responses question whether or not this is REALLY their problem. Without knowing the business environment, rules, politics, and the like, we can't begin to make assumptions about the nature of this gentleman's business. So don't automatically assume he's an idiot, and tell him he doesn't need to upgrade... Just answer the damn question. ;)
"Goodness, how did you people live long enough to invent tools?" -Hobbes (the tiger, not the philosopher)
First buy some parts and build one. Figure out exactly which parts will work with each other. Format the drive and do a fresh reload of all your software with all the service packs and Windows updates on the master machine. Remove any of the crap you don't want the end user to mess with. Setup everything on this first box to a T. I mean get their templates, addressbooks, favorites etc all lined out. if you're using roaming profiles on a PDC then it's not as important to get it right. Samba works great!
Then buy 20 more sets of the final parts and clone the drives. You don't build the systems one at a time you mass produce your 20 systems per year in one setting. Just lie Eli; you install the ram and processors, you bolt in all the motherboards, you install the cards, all the while keeping your cloning operation going. You don't need Ghost, an xcopy works just fine if you have a boot floppy with fdisk to set the partiion active. Or alternately use the find command on a linux box to build a tarbar or do a cp. Have you ever bought a new Dell or whatever? My god they're so full of crap! You've now built 21 machines... Put two of these machines on your desk one as a spare. Keep the spare updated and when a user complains about this or that, swap their PC out with the spare.
If you have roaming profiles and the users trained to save files on the server drive it's even easier to swap out workstations. You have to keep that spare computer updated!
Temporarily loose a user's files when you swap out their PC with the spare to teach them the importance of using the server's hard drives to store their files.
Buy a slighly different but high quality case next year so the users will know what model year workstation they have. This makes support easier.
400's don't seem too bad for a business environment to me though.
Let's say I have a business with several dozen computers. I have "illegal"/reused copies of Windows on all of them. Who cares?
What is Microsoft going to do to me? How would they even find out? Why are people scared of this?
In my opinion, do as much as you like with Windows installations, up until your conscious tells you maybe it's time to give Microsoft some $ credit.
I wouldn't touch this one with a ten foot pole.
Something is going to go wrong during the installation (bad parts, weird software issues, etc.)
If managment is not demanding it I would steer very clear of this bottom end stuff. You will be married to it for the rest of your tenure with the company (and possibly afterwords).
I once did a DSL install for a retailer on the cheap. Using a low end computer that he had around I concted a video kiosk of sorts. Of course, the thing never worked like it should, and it reflected badly on me.
Isn't your job hard enough, that you should leave the OEM business to the OEMs?
"Roll your own ."
Honestly, this phrase needs to die a quick, quiet death. Please help me by refusing to use this phrase. That is all.
When building your own systems you have a number of advantages. It took my boss a few months to learn them after buying some shit hardware from a "professional PC builder"
Advantages:
1) Good hardware in each system, you don't have to deal with low-end entry-level shit motherboards that are unstable and will be the leading cause of crashes.
2) Easier to upgrade. Sure, you can buy an OEM system from HP, Dell, IBM, or wherever... but in two years, when you want to upgrade.... you have to buy an entirely new box instead of just upgrading the motherboard, cpu, and ram.
3) Easier to fix. Sure, OEM's offer supposedly "good" support and will come onsite to fix it. But usually you are stuck mailing off your system to a factor somewhere for hardware replacement. This is just not an option when the hard drive contains sensitive business information.
Since we moved to building our own systems here at my office, we've had increased reliability and performance... we've also seen a drop in the time spent dealing with bad hardware.
If you have qualified people that know PC hardware or you yourself are qualified... then go for it, there is no reason not to.
x
Ever need an online dictionary?
This isn't the situation where you'd see the advantage to custom configuration. As I see it, the two main selling points of the big builders like Dell are 1) a good support base + warranties, and 2) great prices on the "mainstream", lower-end parts (i.e. the ones you seem to be looking at, which they get fantastic volume discounts on). So in your case, they've got the market nailed down.
Even if you feel these advantages aren't enough, then I think dealing with this many systems will just get out of hand too quickly to make it worthwhile.
If I owned the company and was paying you salary, sure, I'd go for it. Free Labor for all. Need more incentive? I've got a roll of stock to hand out.
Seriously, in a shop your size, it's a good short term solution, but hard to pull off long term. Dell, IBM, Compaq and the ilk have an advantage that they spec out thier Mobo's exactly. They can have a single driver install that works for 100 models of computer.
Back in the days I did break fix I came across too many systems that people have put together that were hard to support. You'd get odd errors, but nothing would ever check out as bad no matter how you tested it.
The worst problem was no system was like another. You'd go to a desk. Windows is iritiating itself to support. Let alone trying to figure out what's inside with out cracking the case. With a big name you look at the front of box.
You might want to see if you can go somewhere down the middle. First, call around to several business electronic resellers. Tell them how many manchines you're looking for, and what ballpark range price you want. Also let them know you're calling around to several vendors. Sometimes you can find bargins. In Particular these days in the Cel 1Ghz area. Which is more than 2x than your currect crop of computer. Companies like insight and CDW. Might not hurt to check out the likes of TigerDirect.
If that fails, local IT shops might be able to puts something together for you. The fact that you're looking at 15+ machines means you can most likely deal. Your support costs may not be all that lower, but there is a chance you'll be able to get a replacement part quicker. At the very least you'll save time over having to do it yourself.
Last resort is doing it all yourself. Sure at home, you don't want no stink'n pre-built. Specially a consumer one from a retail store. But at work you have...well...work, and it's doubtful you have the work bench to setup 15 computers at one like an asembly line.
You should have AC'd your question, mike, because now the BSA is going to be all over your a$$ about paying the 'protection' fee for your little construction project.
. . .
Which piece of the original computer does the license go with, the hard drive?
I can answer that first one straight up : MS licenses software according to a complete configuration, usually specified according to model number.
Moreover, as I understand it, if Dell or whoever change *any* component specification, they have to seek a *new* license _every_ time this results in a materially different *system*. I understand that system is defined as mobo + processor, disks and ram et.c. don't have any effect. The system system (are you with me? :) is not mutually exclusive with the model number system of licensing - both seem to have simultaneous effect.
How do I know this?
Well a year or so back, I ordered up a bunch of IBM "M Pro" dual PIII/i840 machines for my company. Firstly, IBM were sharp enough to take our cash (yup that's cash by direct transfer to their account) stating they had shippable product. Rubbish. Weeks later we were still being fobbed off. So at that point I called the legal department at their regional HQ and pointed that they had a material breach of contract and had better sharpen up. We got our boxes pretty darn quick. But with NT4 loaded instead of Win2k. (we'd ordered W2k)
In trying to fix our fulfillment problems I had a direct line to their assembly/engineering management, so this info is near as dang it from the horses mouth. IBM couldn't just switch us a new license for Win2k. Moreover, once an OEM license is accepted by the end user (like when you power up and configure :) , you're bound by the same OEM terms. You are *supposed* to keep the base system.
Yup that sucks. FYI IBM set us up with a bunch of nice SCSI 18Gb 10k drives by way of apology, and the machines are rock solid, service since then good et.c. It was an interesting education.
As far as the real world goes - not that I advocate this - how exactly is MS going to be able to tell you replaced the whole underlying System?
If that made any sense to you, I guess it's a result! I'm too tired to unravel the rest of the gobbledygook that was pumped into my mind when I got irate and pressed for answers why I couldn't just get IBM to hand us the licenses we originally ordered.
Good luck to ya, hope the BSA doesn't catch you at anything you shouldn't be doing:-0
I was just researching building a new AMD system... I was using augustustech.com as the part reseller. I came to about $420 for an XP 1700+, 256 MB ram, onboard sound (Shuttle MOBO), Geforce 2 MX-200 (32 MB), 20 GB HD (60 GB, and still under $500), CD-ROM (CD-RW for about $50 more), and a new case and floppy...
The number one problem with "locally built" low cost PCs: They're cheap. There's no way around it. The advantage OEMs have is that they can put quality tested components together at prices impossible for normal folks to match. PCs are an economy of scale, just like any other mass produced product. There's no way you can build up a QC infrastructure to match Dell's. Whenever a friend asks me to build them a PC, I ask for his credit card, order him a Dell and then setup all the software on it. I only build PCs for myself or close loved ones, because a temperamental homemade PC can really erode confidence, believe me. I shutter to imagine what would happen if I had dozens of them out there.
There's no place like 127.0.0.1
You might try Connect-Computers.com 1.4Mhz with ample memory and disk go for less than your $600.00 price of building your own. For $280+ you can simply get the upgrade kit with processor.
I am in the inverse position. You forgot one HUGE factor. dollars X hours you waste building/repairing/studying them etc...
There are about 80 systems where I work, and it is a royal pain. Rolling your own systems looks good at first but actual cost labor cost goes through the roof.
But that's the only way some of us can get the ladies to go anywhere near our "sensitive equipment."
Does MS alow that in their EULA...... I would read it very carefully...
You could ofcourse use something free. won't save you much right now (since you already have an OS, or assuming it is allowed) but it will save you a lot of maintenance and thus time that you could use to build more than just 25% to 33%
Jeroen
Secure messaging: http://quickmsg.vreeken.net/
Be sure and also factor in the cost of your time and another resources required to build the systems. The time you spend on this will prevent you from doing other work. Be sure that the number really make sense beyond the basic cost of the hardware. You may find that it is more expensive doing it yourself but cheaper if you hire a tech-temp for 30 days or whatever to build them for you.
Have fun
People who bite the hand that feeds them usually lick the boot that kicks them
One thing to watch out for: 400MHz may be plenty of speed, but not if every person in order processing or customer service has AOL IM installed on their machine. Before investing in new systems, I would look at what apps every person has installed on their machine and kill the system load bearing ones. Remind them that these machines are for them to get work done, not to fill with games and crappy applications. Keep a machine clean and it will run much faster.
Money is tight and you are upgrading 400MHz machines? Yow! My company is only now decommisioning some 286 machines, a 400MHz P2 would be heaven.
Anyhow, IANAEE (I am not an Enron Employee) but I suggest fresh Windows installs and some spiffy new case labels on the existing machines and pocketing the cash. No one will suspect a thing!
This is the way I'm headed (that is, if we ever find ourselves in a position to buy new equipment). Then again, I just went through an awfull experience with Micron (fully 15% of my new workstations have had their video card replaced in the last 18 mos, and 2% have needed new mobos.) so I'm prejudiced.
Display some adaptability.
Quite honestly if money is as tight as you say it is, I wouldn't even set out to replace all of them. I would only replace 1) computers that are broken 2)computers where there is a significant business advantage to having a faster machine (compiling movies... get a faster machine, compiling word documents, who cares).
In my opinion, if you have anything faster than a 200 Mhz for most uses, you're probably spending your business money on tech toys instead of your business.
Of course there are exceptions to the rule, but I vote for buying a good terminal server and downgrading everyone to a dumb terminal.
...that can and will go horribly wrong.
After doing my own independent upgrading with friends' and friends' friends' computers, I can tell you one thing: if you make it or change it in any way, you will be RESPONSIBLE for it. The owner of the computer/s will expect service from you as soon as you can find you.
Every little change that happens to their computer? Remember, that's your fault. Why their floppy disk drive scans itself before shutdown (even if it did before you upgraded the computer), you may not know, but you better do something about it before the computer blows up! Believe me, there are times where they will hunt you down because a year after you installed that motherboard in their system a year ago, their computer now is "doing something that it didn't before"...you know, like when they open up a "cool attachment" from their relatives in the email program that they're just beginning to learn and the computer crashes because the person who wrote the program couldn't tell the difference between a divide by zero error and his own ass. Anyway, anything that happens to the computer is because you changed it in some way, hence, you're responsible for it.
It may seem like an extremely fun project to reach for, but realize that the managers are GOING TO HOLD YOU PERSONALLY ACCOUNTABLE for everything that goes wrong and for nothing that goes right. It's a road not even worth a million bucks to take.
It's the old computer with new motherboard, CPU, and memory... or as my dad used to say "This axe has been in the family for over 100 years... of course, it's had 3 new axe heads and 5 new handles in that time!"
We thought building our own would be cheaper, but the maintenance turned out to be a nightmare and cost us a lot more than the machines themselves. In once case, we had a machine that had a bad motherboard, then a bad replacement. Took almost 3 weeks to get that one machine up and running.
Dell is great. They'll come out and fix your machines for you. After the build our own fiasco, we went with Dell. The only problem we had of all the Dell machines was a bad IDE cable in one machine. Otherwise, things were great.
I bet you'll average more than $400 in labor time, for each machine, in the long run. Also, I think your math is bad. I bet you can get decent Dell P4s for $600 or so. A Dell, 128MB P4@1.7GhZ(without monitor), $500 after rebate.
I've got 350 machines and done it both ways. Been through this. It sounds good, but in my opinion, it just ain't worth it. Here's why:
Add up not just what you make an hour, but what your company pays to keep you around, including overhead. Most places you can double your wage. That's what you 'really' cost. Less? OK. Point still holds. We're buying perfectly servicable machines with giga wazoo drives and gigahz processors sans monitors (which have not broken) for between $700-$800 with a three year warranty. It breaks, they ship us a new one or repair it on site. Period. Your parts are costing $600. You're going to have to make money on a hundred dollar margin. That spread is too thin. You've got to manage the parts, store them, and get everything working quickly. That's a lot of prep time. Even if you managed to break even on paper, couldn't you be doing something more useful for the company? I know it's fun (I used to put together computers with nothing but a swiss army knife at trade shows), but you're supposed to be out there making a million dollars. Put your energy where it can be leveraged. --Just my opinion.
How about a moderation of -1 pedantic.
We roll our own because our OS licensing costs are negligible, and because we have had severe problems getting Dell to provide the level of service we need. We have enough expertise in-house that we can fix the problem in less time than we would spend on the phone trying to convince Dell to fix it... Besides, once you change a Dell machine away from their factory install configuration you'll have one hell of a time trying to get good help out of the monkeys they hire to man the phones.
Also, I'd be surprised if you really need to upgrade all your PC's. 400 mhz is still a pretty good box by most standards. And you needn't upgrade to the highest-end Athlon stuff except for maybe a few of the most power-hungry users (CAD, image rendering, etc). For most typical office applications, 400 mhz PII with enough ram (maybe all you need is more ram?) is quite useable and a 1000 mhz Celeron or Duron is plenty. I'm using a 750 mhz PIII right now and it wouldn't occur to me to spend money upgrading it.
They can just tell Microsoft to find a nice corner and autofellate if there is any complaint about "naked PCs".
What we call folk wisdom is often no more than a kind of expedient stupidity.-Edward Abbey
Last summer, we spent a few days upgrading 40 PCs running windows at my university. We changed motherboard, processor, graphics card and added more memory. (same case, harddrive and CDROM)
Start by doing one machine perfectly and then the other's won't be any problem.
Installing was easy as we were already using Ghost.
Your success will probably depend on the ease of installation / maintenance of the new boxes. It's worth spending a lot of time getting it right from the start.
As you have so rightly said, we have little information to go on. To answer the damn question, more information would be needed, such as who runs what applications.
If money is really tight, targeted upgrades rather than a complete set of system upgrades will generate more performance. An assessment of what causes bottlenecks to which users would save money, and allow specific systems to be upgraded with say
* more RAM
* better graphics systems for machines used for design
* faster networking for server apps.
Very few business workstations or applications require more than 400MHz. My advice would be find out how to spend less money better.
Anyway, as others have said, if you do persue an upgrade path, keep track of your Microsoft (and other) program licenses.
Obligatory ending:
Of course you could also save money by running a cheaper Operating system, such as Linux, FreeBSD....
Donte Alistair Anderson Roberts - hi son!
Karma: Chameleon
from the is-DIY-worth-it-in-this-case dept.
I dunno - is the ventilation good in there? Enough power? Just to be on the safe side, I'd get a new case, I like the Antec ones.
;)
This is a mine field, but if you must -- buy your machines as a bare bones kit that has the CPU mounted, heat sink installed, RAM installed, and a POST test before they ship it to you. Nothing cuts into margins like crushing CPUs. Trust me... nothing speeds things up like only screwing in the HDD, FDD, DVD/CDRW, and video - knowing you don't have to hork with mainboard settings or crushing CPUs. Did I mention how easy it is to crush a CPU? Buy good equipment too, cause you get to support it...
+++ UGUCAUCGUAUUUCU
I read somewhere that Dell's overall profit margin is only 5%. That means the components that went into their computer cost 95% of the cost you pay for the computer. I believe 5% is reasonable for someone to put together the computer, test it, and keep driver updates posted on their web site.
I will never build my own system again. I have had too many problems with drivers, things working together properly, and cooling the CPU.
My company used to pay a shop to build computers for us, we buy about 350 a year, and the problem we had was it was hard to keep a consistent set of hardware. If two years go by and you have to replace a video card, can you find the exact one you put in the originals so you can use the same ghost image on all your machines?
Personally, I would buy 60 machines from Dell at the best price they will give me. You should be able to negotiate a 5% discount at least. Since their profit margin is probably only 10% to start with, a 5% discount will be good.
If you get Dell certified, they will even pay you $65 an incident or something like that to fix your companies PC's when they break.
You're either crazy or just a masochist. Yes, it sounds cool and very geeky, but it's also rather impractical. The money your company would save by doing this in house would be lost on having to maintain them and pay you to set them all up (thus taking you away from your other duties.)
This would also take an insane amount of time. Sixty machines is a lot of boxen; optimistically it would take you an hour per machine to swap all the hardware around and reformat the drives and install Windows. In other words, you'd be down at least a week.
Order sixty new machines from an OEM and you're down 2 days tops. Plus you get the guarantee that the machines work (out of 60 boxes, you're bound to get some bad hardware) and you get a warranty from a reputable company, not to mention saving yourself a MAJOR headache.
If something goes wrong with one of the machines, you just call the vendor and straighten it out. If you roll you own, you have to spend time doing diagnostics, then tracking down the receipts, RMA from parts warehouses, limbo time for replacement parts.. All this time your company is paying you to not do the job they hired you for.
Sure, they can hire on another guy to help you, but then there goes all that money you saved having to pay his salary. So in the end, your company didn't really save any money, they just have 60 new machines with no comprehensive warranty, poor tech support, and probably a very frazzled and stressed admin. The geek factor sounds fun, but in reality, it would be more practical to order from a vendor.
Simply because of support issues. If you build your own machine and something breaks, who do you call to replace the part? Most likely you'll be spending time and money troubleshooting broken hardware and buying replacement parts.
When you buy a pre-built system, yeah it's more expensive up front (and not as fun) but in the long run you'll save time and money, and a big headache on your part because of the time and money you'll save with proper warrantees and support. Think about how you will look if these machines start failing and you have to start scrambling...
After all, this is a business not a home project.
It sounds like you are already a one-man band (IT manager, sys admin, etc.).
So how will you find the time to build and provide hardware support for up to 60 computers?
My team has built every machine in our company including the Desktops and Servers. About 8 years ago, we had to do a massive upgrade on many machines. I had just recently started there as the IT manager, and I only had one part time assistant. Money was tight, and since I used to work at a computer stores building and supporting the machines, it was a no brainer to do everything in house. We now have a main office with just over 50 employees, and 4 brach offices in the US and Canada. Except for a few Dell machines I purchased for the remote US locations, everything built at the home office :)
It's now myself, with one full time assistant, and the management of the "White Box" units takes up hardly any time. With our cost savings, in most cases we have entire spare machines ready to give to the user to replace a failed unit. Then, in time we fix the problems with the broken one, and it's ready for the next person.
One thing I would recommend though, is if your going to do this, DO IT ALL THE WAY ! Buy your equipment from a wholesaler. Get an account set up, and be prepared to pay COD to start. You may have to be a bit "Creative" on the application forms, and you might have to go through getting the proper licences for your company (Sales Tax Numbers, etc.) since your technically going to "Sell" to yourself.
Once your good at it, you can start selling machines to the companies employees at a slight markup and MAKE some money to pay for the upkeep !
Then spread out the upgrade - a few units a month - every month, and when you get to the last one, start again from the oldest one on the floor. In an office your/my size it's easy to keep up, no machine gets more then 3 years old, and there is never a massive capital outlay again.
Time travel is possible. We are quickly heading for 1984.
In my experience, when you factor in time cost, no support, no warranty, reliability, and TCO (do not forget TCO, a Favorite for the Bigwigs), the scale always tips to buy brand in a commercial environment.
But there is another route:
Optimize:
Make as many ponds of uniformity as you can (to leverage ghost), making sure that all (peripherals, chipsets, Video, all) is equal. Do a clean install of the OS, PatchIt with all your might. If you can assign machines on the base of needs alone, probably the secretary of the Ceo does not need a Dual Mobile P8 2Ghz Laptop with 1TB of video RAM.
Install Your applications, only a basic set, with a good custom install to get only the meat. If possible for you, do not install office if all someone needs is Works (or do not install StarOffice, if all you need is Koffice). This is specially valid for the afforementioned CEOs Secretary. Besides this will save in Licenses cost (even StarOffice 6 has a license now).
Invest a couple of $20s in some magazines with optimization tricks for the _OS *AND* Applications_ to disable fancy crap, remove fonts and junk, and generaly squeze your machine, then SET THE POLICIES SO THAT USERS CAN NOT CHANGE THIS (please, bear in mind there may be exceptions, otherwise you will alienate your literate users).
Once the machine is in TipTop form, use it as a ghost image for all those in your uniform pond, and move on to the next pond.
If you have a File server insist that user save ALL data files in the server, and applications are on the local HardDisk, that way you can erase the hardisk guilt free.
Then every six months update one machine from the ghost image, patchit... refresh the installation and once is rock solid use its ghost image to update its pond.....
You see it from here.....
*** Suerte a todos y Feliz dia!
OK, How many of you have ACTUALLY done this? I've been doing this for years (since the ATX standard came out) at my office for over 75 users. The key is to not buy crap. It only takes 30 - 45 minutes to build a PC, add another 15 min. to Ghost and your done. What is that $20 of labor per PC? And you have to Ghost most Compaq/Dell/etc to setup a standard system without all the extra crap anyway. We have had zero problems. Only failures have been a couple of power supplies and hard drives. All PC parts come with good warranties, usually - 1yr on MB, 3yr on HD & CPU, 5yr on video, lifetime on good RAM. RMAing parts is NO different than dealing with a Dell/Compaq return.
Recommendations:
Buy good cases.
Research your motherboard choices.
Don't buy the cheapest parts!
Buy boxed parts (CPU, memory)
If you are the person who specs the equipment for your office you will be blamed for failures regardless of you choice of "Roll your own" vs Compaq/Dell.
As many other people have said, you probably don't want to do this in a business environment when you can get reliable machines for pretty cheap from Dell etc. But if you do decide to, here are some general brand recommendations based on my system assembly business experience (I've built and sold ~40 machines in the last 2 years):
Motherboard:
Asus is reliable and fast but also expensive. Look at Gigabyte, MSI, or FIC. (I especially like the Gigabyte GA-7ZMMH as an Athlon platform.)
CPU:
While I have nothing against Intel, AMD consistently kills them in the price/performance ratio. In the past 2 years I have yet to build an Intel based machine for a customer. I just lay out the options and they invariably pick AMD, but that's just my experience. The Duron is a good chip but I don't sell many of those either, people are usually willing to pay a little more for an Athlon of equivalent speed.
Memory:
Buy name brand memory, not generic stuff. I like Micron/Crucial personally, never had a stick go bad. Corsair and Mushkin make good stuff as well, although I had to RMA the only Corsair stick I've ever bought after about 6 months. Corsair customer service was excellent about it though.
Video Card:
Despite their bad reputation some places, I've had good luck with ATI cards when I need a video card. The Radeon VE provides some nice features at a good price point. But in a business environment, get a motherboard with integrated video if you don't need much. If you buy an NVidia card, do some research on the card's manufacturer and look for a reputation of stability.
Hard Drives:
In machines I build, I use almost exclusively Seagate. Good performance, good price, and I've had zero quality problems with them. I feel about the same about Maxtor. I used a couple Quantum drives in my early systems with no difficulties. I don't have any recent experience with Western Digital so I can't really comment on them. IBM drives are too expensive unless you really need the slight performance edge some models offer.
Optical Drives:
I've had stupendous success with Afreey CD and DVD drives. Only had one fail and that was after some kicked the extended disc tray (we were able to open it up and fix it actually). Afreey drives are also very inexpensive.
CD burner wise I stick with Sony. We tried several different brands (Plextor, Acer, etc..) and found that Sony offered excellent quality for a very reasonable price. Plextor is good quality but you pay extra for the Plextor name.
Enclosures/Power Supply:
Antec cases are top quality and have excellent Antec power supplies. They are lots of less expensive cases you can get (and lots of more expensive ones), but Antec had never let me down. If you do get a cheaper case, get it without a power supply and buy a good Sparkle/Antec power supply for it.
(I don't work for any of the abovementioned companies and I don't profess to be an expert on this topic. Just sharing the experience I've had as a system builder who has dealt with a little more volume and long term support than your average hobbyist. We are by no means a high volume shop so I'm sure there are lots of people out there with more experience than me, hopefully they will respond if they feel I'm wrong about any of the parts I've recommended.)
-Sokie
------
Where are the slash-groupies? I distinctly remember being promised slash-groupies!
We have around sixty ~400MHz desktops which are increasingly showing their age. How about taking the viewpoint that the operating system is organic in complexity, which over time would result in less efficient performance. The human analogy would be, the physical capability of youth compared to the slow down that occurs to most with old age, that don't maintain themselves. One possible solution would then be, to just reinstall everything. I'd recommend doing this for the worst performing machine as a test. Also, just because a user has a piece of software installed right now, it doesn't mean they still use it / need it. This would be a useful task, even if you go with new machines.
Users and bosses expect the apps run "well enough". The hardware should pretty much work "all the time".
These are pretty much the minimum requirements. What ends up mattering most is cost. Which is about how you show things in your calculations... especially what it comes to YOUR work. Savings of a few hundred or thousand dollars in hardware end up being a small thing in comparison with the cost of a high-paid guy building and maintaining them. Make sure your TOTAL SOLUTION(tm) is the best one TCOwise.
Depending on how much you need Windows Apps, you may want to consider LTSP (www.ltsp.org) for Linux thin-clients.
The 400MHz boxes are plenty fast for video and graphics input/output via X, while the processing is done on fast servers. You can invest in one fast workstation/server (even a high-end desktop) per 5-20 clients. Budget about 50 MB RAM on the server for each concurrent client and study the ltsp.org site. Management cost is dramatically reduced when all clients remote-boot off a central image or images. If things get slow, upgrade three servers rather than 60 clients.
Many Windoze apps (including Office) are accessable via WTS/Citrix or even via Win4Lin or VmWare. Lindows is coming too.
just replaceing as many of possible is a plus, i am feeling the lag on my 866mhz, 400mhz if unimaginably slow
The Truth: There is no string:)
One thing to add is that summer is almost upon us, and with that season comes many students looking for summer jobs. A couple fliers taken around to the local high school should get your plenty of high schoolers who'd be perfectly competent swapping motherboards/ram/hard drives. Ten bucks an hour to a high school student is better than flipping burgers, and far less than a salried employees time.
- Pentium 166/128Meg RAM running OpenBSD as NAT/Firewall...it is never over 0.1 CPU usage.
- Pentium 120/32Meg RAM running Linux with WindowMaker...Rocks as a surfing station
- AMD K6-II/333, 384Meg RAM ruuning W2k...My sisters main machine (which I sponsored, she is stil at school). Never had a single complaint about performace
- And finally my sweetheart: Pentium Pro 200/256Meg RAM running W2K...Granted with Voodoo2 Card...Runs Halflife, Quake and Unreal like a charm. Return to castle wolfenstein is horrible, but I knew that when buying the game.
I really got sick seeing all these people dissing 400MHz machines. I don't know what they are running, but a 400MHz machines should at least last 3 years in a corporate environment. A well configured Windows (that is not the default unfortunatley) can be stable and last a long time. I myself am a BSD guyOh and I love your nick..however I'm more of an Audi guy (Drive a TT myself)
Ahhh...the great dumpster continuum. Many a free computer will be found there. -- sowth (748135)
AMD cpu and mobos are fine. I recommend the combination of the low-end Duron processors and the any of Via or SIS chipset motherboard.
:)
However, I am wondering about the reason why your colleagues find the limitation of 400MHz speed. Often users mistakenly think their CPU is slow when their disk drive is swapping for more memories. One easy workaround is simply addming more memories. I suspect 400Mhz machines used to be shipped with 128Mb of memories. Upgrading those memories to 256Mb or 386Mb will be one of the most effective performance-per-unit-cost upgrade solution.
Also, think about reinstalling Win2k. Win2k is much more stable and faster than Win ME or 98, but formatting and reinstalling the entire partition often cleans up the trashes built up in the system registry, leading to overall performance enhancements. A user might have a lot of autostarting programs installed hidden.
Investing on the displays and the human interface devices (keyboards and mice) can be another solution. If there are a lot of users complaining about their monitor performance, then try to *sacrifice* performance upgrade to the monitor upgrade. A decent 15" TFT flatpanel monotor can cost $400 and people often get less fatigue when they work with TFT flatpanel than CRT monitors. For me, I'd rather have a flat panel monitor than to have processor & memory upgrade. A combination of memory upgrade and the flat panel monitor purchase can be very appealing under certain circumstatnces if the workers have to look at the monitors for a long time every day. You don't need to squeeze your budget with a lot of possible component combinations in this case.
Good luck,
We have around sixty ~400MHz desktops which are increasingly showing their age.
You have got to be kidding here. 400Mhz? The fastest machine I have at home is a 350Mhz machine. And I am a serious nerd (with many machines at home). Is it really true that the amount of work these people do increases slowly? Is it? Are your secretaries typing 300% faster than they were when you got the 400Mhz machines? Are they? Really?
Or is it just the slow Windows bloat? We used to have it on NT, we have it on 2k. Perhaps you just need a reformat. Or to put a real operating system on them [;)].
Or are you trying to keep some job security by taking on this new job?
Either way, the more pointless upgrades, the more waste that has to be disposed of one way or another. (We used to have huge lorry containers which we filled with equipment and would go to get crushed. Kinda pointless.)
Malike Bamiyi wanted my assistance.
My 2Ghz P4 Dells are:
- 256MB DDR Ram
- 80GB 7200RPM Maxtor HD
- Builtin Sound, Ethernet
- CD-RW Drive
- Full Tower case (solid)
I've been building clone boxes my whole life, but I couldn't pass this Dell up. This is pretty typical on Dell's site.I usually check gotapex.com for deals.
Today they have a business class Dell P4 1.6GHz GX240 for $357.64 shipped. You can't build a loaded clone machine from scratch for that much, let alone one covered by a 3 year warranty.
If you're not using Windows, a 400 Mhz computer works great. My home computer is a pentium II 400, and I really have no need for upgrading it. Granted, I use Linux (though I still have windows 98 which runs. Sometimes.), but if cost is a big issue, then obviously the best plan would be the free one right? Not only would you not have to buy new equipment, but you get a better, more stable network environment. Obviously deploying Linux or some other free OS on a previously Windows network is not as easy as it sounds (depending on what software you need to run, how easily the users can adjust to not having their "friendly" os, and other factors), but it is something to think about if times are tough.
On a side note, I'd be glad to take those 60 "outdated" systems off your hands if you really are just aching to upgrade...
My suggestion is to go with a local IT shop that specializes in servicing businesses. Spec out your machine along with them. Ie. use parts they can get within a day from thier suppliers, preferrably dropped shipped to you. Also they should have some ie what is reliable and what isn't. A really good shop will even take care of warrantees and such. Don't expect to save money over Dell though. But you will build machines with better parts and an easy upgrade path. My experience with MSI and Gigabyte boards in a business environment has been good. Gigabyte dual-bios is a great redundant feature. Remember your users first though. Building small silent systems with LCD's might be the best for users in tight cubicles.
As long as you know what works try to leverage the same setup as long as possible. Building your own systems not only can save you money but can provide much better bang-for-the-buck. I've been using cheap ECS K7S5A motherboards and duron cpus. The systems seem very stable when used with quality ram and a good power supply. When the duron can't cut it I will upgrade the systems to an athlon XP.
IMO, Dell or Compaq support isn't worth a dime. Their systems fail as often as the built it yourself type, and I'd prefer not to wait 45 minutes on the phone to return a dead cd drive
Is spending the money to maintain homegrown systems outweigh the $400 you would have spent on a less resource (i.e. IT staff) intensive system?
You either invest in hardware or people. Cheap hardware means more time maintaining it, usually.
I look at it this way. How long do you plan on keeping the machines around. If you are planning on keeping them for 2 years. Divide 400 by 24 and that's what you save per month. Figure in your time to purchase, build and maintain the system as well as increased downtime because more stuff goes wrong with systems you build yourself.
I don't care what machines you use, but Dell workstations are tested to run as a unit. PC's you build are not tested as such and will have more problems and more downtime.
6 of one and half dozen of the other.
1) By N more systems than you need. That way when one dies you can just swap out the old hard drives (assuming it's ok) and put them in the spare system. Then you can fix the broken system when its convenient, and not in a rush with the end user hovering by.
2) In your design, consider including TWO of those slide-in hardrive bays in each machine. One for the hard drive in your original machine so *maybe* you can legally transfer your Windows License (does it work that way?). The second slide-in bay is for a new larger disk drive. That way, if someone's machines dies, you need only slide the drives out of one machine and into an waiting spare. Saves time and helps lower anxiety.
3) Buy 5 or 10 complete systems as a trial. I've found in my limited experience (the systems I've built number only in the teens) you're never sure what you're going to get until they arrive. And by ordering several systems you see if you the vendors send identical parts or mixes them up.
Do you really *need* to upgrade those 400Mhz machines? Look at your user's needs; could you be better served by looking at a thin client solution, with either a Linux/X11 or Win2K/Terminal Server config? Consider: you get better managability, since all the apps are on the server. The people who have extra power now don't have it going unused, it's reallocated to those who need it more. Furthermore, (especially with an all-*NIX solution) the liscencing costs are far less.
If your shop doesn't use Exchange, then Evolution is a more than adequate replacement for Outlook. If you do use exchange, look into the CodeWeavers Exchange plugin solution.
As for Word, Excel & Powerpoint
I know it would be 'nice' to get a little snappier systems for your company but these do the job, keep em. At my previous company 90% of the employees had DELL PPro 200's running NT4. it did what you needed for office type work.
save the money. You may need it if the time comes for determining severence packages ;-).
As for building the PCs yourself. Having worked in this capacity before. Building from scratch I could average between 9-14 PCs in a 9 hour day with interruption, thats just assembly with a quick P.O test NOT including installing the OS on the boxes. It's nothing you would get done all at once during a consecutive period of time. From your perspective the rollout could take forever (like a month) and then you have to deal with all the quirks that come up. Your time is more important than that.
Peter
www.alphalinux.org
When our company was only a tiny start-up, the IT Manager had the same idea and handbuilt our entired force of 30 pcs. They were great for the time and for cutting expenses, but it served more of a short term goal. Now that we are 300+ employees strong and the previous head of IT is gone, they cause problems. None of the assembly was documented, nor were any of the parts; we have found out now that there are a variety of different mobos, video card, NIC cards & processor speeds floating around in them, and no one knows what is where.
Kinda makes it a bitch for our current IT staff (me) to troubleshoot effiently and find current drivers, etc.
A plan that that can definately work, but be sure that there are long term standards in case business picks back up; like consistency among machines and documentation up the wazoo.
IMHO, we've never been able to take a box from Dell, stick it on a desktop and ask "Little Bo Beep" to sign in. There's usually a re-install of the OS required.
:-)
In terms of after sales support BIG NAME seems better at first glance...BUT something to keep in mind is...At a saving of $400 per box, if you buy 10 boxes you've saved $4000, that's more than enough to buy 2/3 extra boxes to keep as backup. Which can be used as a swopout when something goes wrong. Once the failed box is repaired that can be swopped.
Lastly, I'd suggest buying a full box with case and PSU. That way you can have your supplier do the assembly for you at a minimal fee. This saves you time that could be spent in the pub
What is your friggin idea of acceptable?? " issue.
You can get the entry-level dell or compaq or gateway for the same as that Athlon, and despite the fact that it isn't a 2.2 Ghz machine, the dell at 1.7 Ghz still blows the doors off of what you have now, not to mention a warranty, not to mention the whole pesky "what-do-you-need-that-much-horsepower-for-really
I think we are all aware that $100 in ram counts for more than a $300 faster processor these days especially since, the processor is rarely the bottleneck anymore. It is the bus or the amount of ram (dictating the need for virtual mem) or the speed of the HD. Jeez. How long have you been making the decisions?
Some questions to think about:
What are you useing these computers for?
Who's useing these computers?
Have you upgraded software that runs on them recently, so that it needs more prosessing power?
Is the new software nessisary, or just because?
Have you got enough RAM in each machine?
Are any parts failing?
Bu, if you are going to upgrade anyway, do something that makes sense. Auction off the MB and chip at the very least. get more money into the fund to upgrade more...
First off.. What the hell are you doing that requires more than a P-III 400?
W2K and office 2K run perfectly on it. unless you are running games or software development there is no sane reason to upgrade. (web development OUTSIDE graphics design does not, contrary to what the web-designers think, need any more power than that either)
Is your desire to upgrade purely for functionality? what profit gains or gains in productivity will you realize from the upgrades? if you can give solid financial numbers showing that profits will increase then go for upgrading.. If you are just stroking an itch or ego.... dont waste the money. W2K+O2K works fine on P-II 366's. and that is where I am keeping the sales department until machines break, or corperate loses their mind and tries to shove XP down our throats.. (we JUST migrated to W2K.. I dont see XP within 3 years) other than that it is just wasting money.
Do not look at laser with remaining good eye.
make linux boxes out of them.,. use them as test boxes.. "proof of concept" testbeds. I wish my old job had stuff like that laying around =/
..and here's what I've learned.
Go with proven hardware.
I went with ASUS A7V's (first run of machines) and A7V266's on the second batch. Basic CD-ROM, floppy, 20/30GB drives (Maxtor/WD *only*), Brandless Vanta cards (tho now I'd go with GF2MX400 minimum), 3C905's, Creative PCI16's.
I shouldn't have skimped on the video cards, I went with the cheapest no name TNT/Vanta cards the distro had, and have replaced 4 of them so far (image blurs & streaks off to one side). It's not that big of a deal except I've had to replace the cards (buy at least 1 spare vid/net/snd card, for mobo's they can handle a machine being down for a week, but not for components).
Try and get a board that's got a few BIOS revisions under it's belt, when I got the A7V's they were fresh and had a few issues (that I was able to work around, but still), the A7V266s have been solid since.
To my understanding the KT333 is a rock solid chipset out of the box (I've had a KT3Ultra running at the house for a few weeks now, no issues). Might consider those.
Buy your RAM from somewhere with lifetime warranty. I get all mine from crucial.com, the headaches saved are with their small price margin over the el-cheapo's on pricewatch.
Don't be afraid to use the cheap places for items with Manufacturer's warranty. Previously we alway purchased the nice fat 3yr warranty's from Dell, Gateway, IBM, etc....the only times we've had to use em they weren't worth the shit I'd like to have done in the reps mouth. YOU are the support, and the manufacturer's handle the warranty.
I've yet to use any of the junk software they like to bundle with the systems. If you have less than 100 machines it's just easier to track hardware with a simple spreadsheet.
Final reason I'd suggest not buying prebuilds is the hassle it is to get a clean machine...we've already got a site license for WindersExpy, why the hell woud I want some crap home edition from Dell/HP/whomever...?
...at least, it's a 450MHz Celeron, running Mandrake Linux 8.2 which came with OpenOffice.org 641d. It does all I ask of it, even OCR and stuff, without complaining. Not sure why you would want a P4 sace heater when a fast P3 gets 90% of the work done for 50% of the power. The reduction in the dotcom's power bill would probably pay for a second processor over a couple of years.
Got time? Spend some of it coding or testing
If this is really the highest valued use of your time, and if you are competent with hardware, you should go for it. You can build machines that work well enough to make support irrelevant, though they won't be cheap.
By the way, what in the world are you doing which can't be done on 400MHz PII's? If you have enough RAM, they should be able to handle W2K. If you are running compute-intensive stuff, maybe you should be looking at a compute-server for those few bottleneck apps, and keep the old hardware around for general use (for which I would consider it mild overkill).
My first post on /. ever!
If you use the nforce chipset from Nvidia you can install ONE driver from Nvidia for all parts (Sound, Video, IDE, 10/100 NIC, USB, PCI, etc.). They make drivers for Win95/98/Me, Win2000/XP, and Linux. As everything is intergrated, managment and assembly should be EZ.
I would go Micro-ATX for cost. Poking aroung newegg gives you a CDr&cable, 1.3ghz Duron&HSF, 512mb of ram, a case, 30gb 7200rpm HD, and a floppy for only $530. Note: The nforce needs 2 DIMMs for optimal performance.
Is life so dear, or peace so sweet, as to be purchased at the price of chains and slavery?
I know this is a little late, but here's a solution we've come up with that's in the cooker for some of our clients in a similar situation.
:-)
Basically you should invest $4,000 in a single server, RAID SCSI drives, dual athlon, 2 gigs of RAM. You've already got a 10/100 Mb backbone for your network, so you can slip this in just about anywhere.
Now here's where it gets fun. Load your favorite distro of linux, visit the Linux Terminal Server Project, and make a terminal server out of it. Then, check out MOSIX, or Sun's grid-computing (the later sports better redundancy, a feature I adore when working with end-users). Grab nics and boot-roms for each PC, install 'em, and boom, you've got a complete functioning cluster of what, 40 PII's? You have any idea the power those can muster?
Not only will you see a huge boost in computing power, but you also save money. Need to use quick books? What's a single liscence for Citrix cost? You can publish the app natively on your terminal server. Open Office works great for converting all those old MS documents.
Honestly, KDE 3.0 just came out. Use it.
Slackware forever. Honestly, what else would you trust when it absolutely positively has to be stable, secure, and easy
People say theres no such thing as a dumb question.. Well.. some of the more recent Ask Slashdot's have been a prime example of such.
However, when I first saw the "Rolling your own business desktops?" heading, my eyes rolled and I thought "oh no, not another one"
Thats till I read the persons question. He did his research and presented his facts, with links in a clear manner. I had to read the comments.
Some of the comments were REALLY good. I learned alot from them.
My point - Ask a good question, get good answers.
As the network manager of a Research company, we currently have about 450 PCs about 200 of those are less than 300MHz. We switched to this concept this year and have a saved an assload of money. Of course, our CFO/CIO who is an idiot, cut our upgrades in half even tho they are costing us much less. With our tax exempt status, we can get a full system (Athlon XP 1700+, board, TNT2 Video (32MB), 256MB RAM, DVD, 17" monitor, $89 Win XP Home OEM license (thank you enterprise select) all for less than $600. We saved enough to hire a person to build them. I say go for it.
No, he's building new boxes. Even if he wasn't, no MS OS licence has allowed `recycling' after Windows 95, so changing the CPU would axe the licence.
Mandrake Linux's licence allows you to recycle an installation. Come to think of it, you're also allowed to copy an existing installation, install as amny times as you like from the Download CD set, benchmark it against other things, use an unlimited number of seats, and comes with OpenOffice.org 641D. Oh, and even if you spot Mandrake $50 a machine, that works out at around 1/10 of the cost of MS-Windows+MS-Office, and no free viruses.
Got time? Spend some of it coding or testing
I've worked for a white-box builder/consulting company and now i'm a sysadmin. I've been on both sides of the build or buy question and i've finally decided to buy pre-built boxes...here's why:
Replacement parts availability.
That nifty Athlon board from MSI, ABIT, Epox, and the like won't be around 1 year from now. If you image machines, you will most likely have to create new images when you service/replace hardware. (Win2k doesn't like having its boot controllers moved around very much...you'll get the "inaccesable boot device error".)
The upfront cost savings may be attractive, but there isn't a free lunch...you'll have to spend more time maintaining different platforms.
-ted
I expect you can come in considerably less than your $600 estimate, particularly if you accept a CPU 33 mhz or 66 mhz slower than the very top end at the time you make your move. I see no downside and considerable upside. Since you indicate that the old systems are from major vendors, however, I would think twice about reusing cases, many such cases will not support standard boards, and very few if any will have enough power for a modern CPU. Still, you can buy cases with good, high wattage supplies in bulk at good prices, so this should not be a major obstical.
...which may or may not apply to your situation, but is something to think about if you are not a) totally married to Windows, b)it's not a huge network, and c) the network is otherwise well secured:
You can use old x486s as semi-dumb/thin clients to an app server using xdm -- the app server's X server just redirects the output to the users monitor.
I've set this up before and the performance is really very good. The x486s can have the OS installed locally so that they don't bog down the server booting from the network. Administration is centralized. Also, you can probably just keep using the workstation you have, otherwise, old boxes are cheap to find.
The downside of course is that it should only be done on a trusted network, etc. Again, if you need Windows, this is probably a useless idea.
All one has to do is hit the right keys at the right time and the instrument plays itself. - Johann Sebastian Bach
You may get a tax break if you donate the used computers to local schools etc.
If you could donate or sell each 400 mHz box for a $200 tax credit, upgrading would not be worthwhile!
Get a beefy server, run LTSP from LTSP.org along with Rdesktop which allows all windows apps. All machines will run diskless from the server. One point of Admin. All of the benifits of Linux with Win apps available.
If ya need help, there is plenty at ltsp.org
Mike
MS-Word 97 sometimes won't read docs from (for example) MS-Word 2000 or MS-Word XP. In those circumstances I've used StarOffice 5.2 to reprocess the document and make it work. OpenOffice.org would probably do an even better job of it.
Come to think of it, why bother keeping MS-Office? It's only another invitation for the BSA (BSAA in Oz) to bitchslap you for not having a matching holo-sticker.
Also, if office work is all that the machine is doing, upgrade it to Linux. Mandrake 8.2 even comes with OO. Let's see the BSA fine you for that. Buy a decent scanner for each office cluster out of the savings, and SANE it so all can play.
Got time? Spend some of it coding or testing
You are not going to get a better bang for the buck by building your systems yourself. You can negotiate a large lot of systems down significantly from one of the big makers (Compaq, HP, etc..)
You are also spending assembly time. That is significantly more expensive than shipping costs (which are usually free and part of the deal, actually)
You are missing all the management features which are indeed nice. Remember, the big guys have built the systems based on what enterprise customers have been asking for years for. Case-opening detection, remote bios flash, easy access to components, etc.
Also, you are missing tech support. Worth it or not, if a computer is fried, it's not your fault. If you put the computers together, you are responsible.
And the big one -- replacement parts. You may be able to make your cool PC and all, but are you going to guarantee that all the parts you put in there will be available a year from now? The large manufacturers will take care of you in that respect. Mom & pop assemblers will always try to sell you the latest and greatest. This is not what you want; you want stable and consistent for an enterprise platform.
I strongly recommend to do this professionally. Buy from the big guys (anything but Dell, since I hate the dude on the commercials)
If you don't already know the answer, you probably aren't the one who should be building the new systems. Most companies I know only replace a portion of their hardware at a time.
clancey
I'm getting 1.2 Gig celeron GX50 dells for 500 bucks
20 gig HD 128Meg RAM AGP 16meg video ect. Plus service and support.
I think you should shop some more...
Heh, here in .ro, at my previous SoftDevCo I've built all our workstations from components (12PCs + 1 dual power source fileserver, 1 dual CPU test server) in 3 working days from components. I think one week (aka 7 days) is enough for 60 PCs with another one to help disasembly. .ro is that the Company that offers best prices for components sell full systems only with M$ Win installed (another damn useless +100$ software).
Now about price, I know in Eastern Europe PC compnents are pretty expensive rather then US, but here with $520 (taxes included) you can put together: a Athlon XP 1800+/Cooler, 512MB DDR, MB/KT266A, DiamondMax D740X, 40GB HDD.
I think that's enough for a usual workstation and the difference from what you have is quite big.
Anyway I'll always choose to build myself all workstations even if my real job is Java serverside programming. And another reason here in
Regards,
c0d
Get a 1 GHz CPU upgrade for $130, add 258 Mb of RAM and your done.
http://www.powerleap.com/Products/iP3T.htm
No issues about OS transfer as it is the same box and no I don't work for them I just have used the product. They make socket 370 upgrades too.
The chances are that only a few of your workers really need higher powered rigs. Why not just assemble a few rigs yourself and put the rest on hold? Hardware is changing so fast right now that everything you get is obsolete in a heart beat. Put power where power is needed and give consideration to the power consumed by each unit. Electric bills might soon be a bigger issue than a faster Pc.
Don't forget the option of buying a batch of machines from some other failed
Have you factored in server upgrades? Or the fact that if the server dies everything stops? Please remeber that white papers are typically subjective, or argumentative.
I wouldn't recommend that they move completly to thin client, but rather have a percentage of computers that are quite capable of working on their own in case something goes wrong.
how many of those systems will die 2 years earlier down the line because of latent failures created when you worked on them. The FET transistors and high-density components in PC's are sensitive to static electricity. Just because it doesn't fail as soon as you plug it in doesn't mean you haven't damaged it. From the huge amount of faulty parts (RAM etc) I had when I was an engineer I'd guess lots of people in the industry don't bother with anti-static.
Don't do it; you give the reason yourself! If the company can't commit to $36k for doing all the upgrades at once then it simply isn't viable to look at it as a wholesale upgrade (licensing issues aside).
A case could be made for maintenance upgrades... RAM here, HD there, maybe even a processor or two, to the users who need it the most.
But, be careful proposing a capital upgrade (long-term project):
Will the existing equipment be viable for three-four more years?
Will the upgraded equipment be viable to anyone in the company in another two years?
Will the OS be maintainable in two or three years?
At what point will the company need to buy new equipment to support obsolescence?
Does the company plan on expanding or contracting in the short-term?
Is this proposal really to provide additional job security? Seriously... it can backfire if the company's needs aren't consistent.
I think you will find that it isn't in anybody's interests to make a long-term plan on upgrading the existing equipment, but... a short-term plan (with known comprimises) is great for deferring costs for a little while.
(If your proposal had offered a 60-70% savings, it might be another animal, but 40% doesn't cover the risks too well IMHO.)
A few comments...
1. Why buy P4's when you can do just as well with P3's? Why pay a premium for this year's models? If you were short on cash and in the market for a new car, would you buy the 2002 model or get a 2001 instead?
2. If you must make a choice between getting a faster processor or more RAM, you should almost always go with more RAM. If your computers are short on RAM, they will seem much slower than they really are. Get plenty of RAM and be your computers will last much longer.
3. I notice that you are using Ghost. Ghost is not the miracle tool that many think it should be. It sounds like you'd get more milage out of a desktop management solution, like Novell's ZENworks product. A well-done ZENworks managed network will cut your workload in half, which means you can let go of some people.. though, that may not be what you want.
To give you an example, with ZENworks and some talented people, you could easily manage 5,000 Win2K workstations with a staff of ten or less people. Right now, my group manages around 1,500 workstations with six full-time staff and we are not taking full advantage of ZENworks.
The Win2K distribution itself and our 120 or so applications only use about six gigabytes worth of disk space; much better than a lot of Ghost images. We can handle a broad range of harware and software quite easily.
4. This may not work for your business, but this is how hardware is handled at mine. The hardware shop sells computers to the various departments, essentially at cost (perhaps a little above to cover fluctuating prices). Any computer purchased has virtually unlimited maintenance, though there are some exceptions. The shop will replace any part that goes bad, usually the same or next day.
At the end of the year, the total cost of all maintenance is divided by the number of computers on maintenance. At that time, each department is charged a flat fee for each computer. The repair costs are averaged across the entire organization, which helps distribute the load.
Service is fast since parts are already on site and it is possible to replace entire machines if necessary. You might opt to have computers on standby that you could swap out when a computer has a tricky problem that cannot be solved quickly. It sure beats calling Dell or another vendor and beating your head against the wall for hours or days while the user is waiting.
One thing though, our hardware shop is not responsible for software; except motherboard and video card and network card firmware updates and they are only required to make those updates available to our software people. In other words, if Windows won't boot, it is NOT the hardware shop's responsibility to recover data. That is another reason that the hardware shop can get their job done very quickly.
Doing it this way, we manage a total of some 6,000 workstations throughout the enterprise.
Ouch! The truth hurts!
The company will want to insure the new hardware. At an estimated value of $36,000, they will be concerned about replacements and disaster recovery.
From past experience, banks and insurance companies are very nervous when non-brand name PCs are used.
Make sure that MS has no problems with you recycling that OS
Most large OEM use the cheapest crap they can get away with. Ever wonder why the rarely tell the consumer what brands the components are? On top of that, I just read a Dell ad that states an NVidia Geforce2 MX is a "high performance" card. Some people believe anything. PS: Anyone want to buy a lovely piece of land?
As the old adage goes "no good deed goes unpunished". If you buy Dell and there is a problem you can blame Dell and argue in your defense that you made a conservative responsible choice. If you try to be a maverick, you will be blamed for everything that goes wrong, the cost savings will most likely be forgotten, and you will spend a lot of time being questioned over and over again justifying your decision.
I have to agree with the previous couple of posters. Unless you're working in freakish conditions, or insist on scuffing your feet on the carpet to transport your items across the room and then insist on touching all the metal parts on it...you should be ok. Just don't touch any metal, take the time to secure everything appropriately...and you'll be just fine.
All of my machines are still running well, as I did my research and bought components that played well with each other.
If all you have is a hammer, everything looks like a nail.
Rolling your own box is great for a small network or a special system. All my home machines are self-rolled, it's worth the savings. However in a large, corporate environment, it ceases to be practical. Presumably, if resources are strained, you probably have better things to do then sitting down and killing tons of time diagnosing occasional hardware issues. It may even make you look less competent (what, he is still fixing that box? Dell would have fixed it by now). An OEM always has spare systems to swap out on a moments notice, and can do the swap and worry about more thorough diagnostics later, while you can not.
XML is like violence. If it doesn't solve the problem, use more.
Thought you could get that by us did you? We all know 400MHz is plenty fast for a regular office environment.
The real meaning of that statement:
"We have around sixty ~400MHz desktops which are increasingly showing their age during our office-wide quake tournaments".
or
"People are complaining that The Sims is running to choppy".
Don't worry, I won't tell your boss.
I have noticed that some people have suggested thin clients. But there is another 'half way' solution- blades. Basically, this puts the guts of the machine in a rack in a network closet and then runs a cable to each desk. When a machine breaks, you just pull the board and pop in a new one- no messing around under desks, no going out onto the floor. Since each rack has one redundant power supply, no case, and so on you can theoretically get a lot more bang for your buck and higher reliability.
When spec'ing a system, consider the following two things:
1. Overpowering a system is not bad if it doubles the lifetime of said system. No, web designers don't need a 1.4ghz Athlon today - but if your company can put off upgrades for another couple years, you've saved them a good amount of money more than if you had put an 800mhz Celeron in. At the same time, though, most people don't need a SB Live!, and you can skimp a bit on that.
2. When using integrated components on a motherboard, _always_ leave yourself an out if something on the motherboard fails. If the onboard video dies (I've seen it happen!), make sure you've got a PCI slot to slap a G200 into. This doesn't mean you need to keep the system like that, but a quick half-hour fix until you get time to do a mobo replacement can be a life-saver.
3. RAM, RAM, and more RAM. This cannot be impressed enough.
4. Remember that not all upgrades need to be "performance oriented". Your graphic designers will love you if you give them a nice 19 inch monitor to work with more than a GF4.
5. Reliability. You want things that are stable. This is harder than you think. Buy a test system and test it thoroughly before you commit to a thousand of the things. Stability is more than just your motherboard chipset.
-Erwos
Plausible conjecture should not be misrepresented as proof positive.
It is worth it if you can spare the time to build the systems. Try buying from a vendor like MWave who will put the MB, CPU, and Memory together for you and test it, it will save you lots of time. Also buy GOOD Power Suplies/Cases. I build systems all the time, and the only power supplies i will use are Antec, same goes with fans. I have found them to be the most reliable on the market. I have systems that run in EXTREMELY dust environments, and don't get cleaned like they should. I can say that i have never had an antec power supply die, and i have had dozens of other manufacturers die. Other notes. Buy a good motherboard, and stick a decent video card on it. Don't use built in Video Cards, they are flaky at best. Good Luck with the project.
60 boxes saving $600 each is $36,000. I doubt the roll out would take more than a month or so, but then again I don't know how long it takes to set up win2k uglies. At the $5.00/hr a MSCE is worth, the company should be a winner unless it takes more than 45 of them to do it.
DMCA, Hollings, Palladium. What might have sounded like paranoia is now common sense.
I used to work for a MAJOR web hosting provider who built their own boxes. They employed a full time hardare guy to put them together. The owner of the company was a real cheapskate so he cut corners whenever he could. To save money they used FreeBSD and Apache, and they built their servers from generic desktop towers, just like the one your mom uses. They made a ton of money though. With a full time hardware guy and enough UNIX gurus, they had no problem keeping systems up and running.
depends on how skilled you are I've been building systems since 1970 You can get screwed by: flaky motherboard mfgr bad return policy, etc prototype a test system and fire Linux up on it I'd have no problem with this but I don't build for others PS I build NEW AMD systems for $600 full tilt
But seriously I have worked on many many systems over the years and I have never had a problem which could be credibly linked to hardware failure brought on by ESD. I think that lots of less-experienced techs and help desk people blame any problem that they can't figure out on ESD. Got a BSOD that you can't explain? Here's a convenient line that no one can really disprove and makes you look smart. "It must have been mishandled by some other ignorant tech years ago and is just now showing symptoms." Right. How could even tell the difference between an ESD problem and a problem caused by irregular AC line voltages or manufacturing defects?
What the hell is a "latent failure"? As was stated above, ESD is measured in kV while CMOS tolerances are more like 30V. Either a transistor is blown or it isn't. I agree that ESD can damage transistors, and I also know that a computer may very well power up after suffering damage from this. However I think the notion that a system would power up and work normally for two years before going south is ludicrous. You seem to think that the static can somehow "weaken" the hardware without fully blowing it out.
My personal solution to the ESD problem is a compromise between the incredibly annoying wrist strap and "going commando" and risking relatively expensive hardware. I leave the power supply plugged into a grounded outlet while working on the machine. I know somebody's gonna flame me for this, but think about it. When the machine is plugged in the entire chassis is a path to ground which can bleed off excess voltage in the case of a static discharge. If you simply touch the chassis before you start working you will discharge any static electricity which is being carried by your body and you're good to go. Unless you are working on your computer while standing on a shag carpet in your socks while rubbing a balloon on your head then this is probably all you need to be safe. You could then unplug the AC line if you wanted, although I don't see any harm in leaving it plugged in during your entire operation. Outside of the power supply the voltages can be no more than 12V and low current so electric shock is really not an issue.
On a side note I think a much more common issue is the failure of the power supply itself, rather than motherboards and chips. In most machines I build the PS will burn out after a couple of years unless I spend a few extra dollars on a step-up model case like an Enlight.
Alright, flame away.
alex
--- Wherever you go, everyone is always connected...
While building a PC for home use might make good economic sense and be supportable (if you do your own tech support), if you build the machines you can be held responsible by your leadership for their failure. How dependant on PC's is your company? And unless your company is made up of computer geeks, tech support on home grown PC's could be overwhelming to a small support section. Also remember that when you build a PC you buy parts made by a number of companies. When a part breaks, you often are forced to go to that particular vendor and make them honor their warranty. On the other hand, if you buy a new PC the only vendor you'll need to deal with in the event of a problem is the company that sold you the PC. Time is money. You also need to get your PC replacement schedule on a cycle that supports the level of machine you need to do business. Example, if most of your PC's are used for word processing then think about setting up a budget that allows you to replace 25% of them each year over a four year cycle (or even 20% yearly for a five year cycle). If you need something more robust and your company is often upgrading to new software to perform critical functions (DTP, CAD, etc.)you may need to go on a three year replacement cycle. Just my $.02 worth....
I'm not sure what the chances are you'll even see this. 450 posts and the ones modded up are pretty negative. Oh well. . .
I'm currently rolling custom built machines for our 200 systems network. Oh, and I'm the only tech here. I do the servers, network, help desk planning, everything.
My place is a non-profit where a very small, chaotic budget. I'm never sure exactly when I'll have money to spend or how much. For strange reasons, when we go to spend money we have to go through a maze to buy complete equipment, but components are no problem. We couldn't buy a new company car, but we could buy all the parts to put a car together ourselves. Same goes for computers.
The savings we've seen building ourselves are huge. Adding the costs of the pieces and my time spent planning, building, and supporting these systems it is still cheaper than OEM systems and a support contract. A+ certified techs are a dime a dozen, so support of these wintel systems isn't really a factor if I were to leave. (They'd have to get a half dozen to do what I'm doing by myself in 40 hours a week, plus an MCSE and a CCNA, so I'm not worried about job security).
Here are some tips:
1) Plan out your configuration and use it for the next year. The most important component is the motherboard. It should be able to accept more RAM and a faster CPU than you are going to use initially. Spend lots of time developing a stable, user-friendly software config (OS and apps). As you need to replace systems throughout the year, use this config. After six months update the config with a faster CPU, more RAM, and maybe a larger HD. Update your software config with patches, fixes, stuff like that now also. At the year mark you can plan your new config.
2) Integrated components are your friend. I like the nVidia nForce boards because they have the (good) sound, video, and network integrated. Also, if one manufacturer stops making your board, you should be able to switch to another manufacturer but still use the same drivers. Very important for ghosting!
3) You really don't need the management software for 60 computers. That stuff is usually designed (and priced) for enterprises with several hundred if not thousands of systems. You should be able to keep most of that stuff in your head and in a small text-file database. Learn a little Python/tk and you can even build your own front end to the text-file. Cool!
4) Develop a relationship with a couple local component vendors, and a couple Internet vendors and have them bid for any purchase more than a couple grand. You'll definetely save money this way, especially if they know they are bidding and not just giving a price quote. I've saved thousands of dollars on a single purchase this way. Also, after a while the local guys will probably be able to send a couple guys your way to help out every once-in-a-while when you get swamped or stuck as a thank you for your business. Very Cool!
Following these tips, you only have four platforms to work with, you've saved money, you know exactly what you are working with, and you get a sense of pride from creating something from your own two hands.
I really can't recommend this approach highly enough.
obviously no deficiencies vs. no obvious deficiencies
Drinking is almost mandatory before a lot of computer work! How else would I summon the courage and forget the past pain of maintaining Windows?
Just install new celerons running at 1100mhz , they ar fast enough and compatible with most boards. And are a bargin, even though rich corporates wont update peoples PCs.
I'm a geek too, so I think that rolling my own desktop would be a great thrill, especially if I could make them linux boxen, but I also know that rolling your own is a huge investment in time that you might not be able to afford ... unless you get overtime pay. Good Luck. Hope that you build them.
which ran just fine on my 400. I really don't see this need to have faster and faster machines just because you are getting impatient with how long sometime takes to load. Remember, suzy in the phone center doesn't need anything over 800x600x256 and that will cut down the system drain by a lot. It isn't like she will be using anything more than database querrys and maybe word, which both work just fine in this environment.
At the last two locations I did a build out at we only format the drives for 3 gig because thats what the image size is...kinda silly really
[RIAA] says its concern is artists. That's true, in just the sense that a cattle rancher is concerned about its cattle.
If you are only building one or a few computers this might be an acceptable idea. If you are planning on several dozen or more forget it. If you are buying a single MB you probably won't run the risk of getting a DOA unit. But if you are buying a quanity you will run into bad units, and then will have to run the gantlet of returning the bad units and getting replacements. Ditto for memory, video cards, etc. The Tawain row outlets don't have the quality control that big guys like Dell and Gateway have. And you don't have the clout of ordering parts in the 1000's to get quick turn around on DOA parts. Unless you want the headaches of becomming a OEM buy your computers ready to go from trusted channels.
Another route, is to go used/reconditioned/surplus on last years models. Try
http://www.compgeeks.com I've bought systems from them and they are great. Don't know if they can supply in the quantity you need, but give them a try!
What's the point of spending $600 + labor, and getting crappy machines with no warranty? Wouldn't you rather have DELL come fix the problems? Or are you looking to secure your job for the next 3 years?
Besides, you ought to be able to sell the old machines for $100 each...
I've just left an aerospace start-up where I built, deployed and managed over 100 desktop systems myself; they were cheaper and faster than anything the "big guys" could have provided, and I ran into just about zero difficulties over my nearly three years with them.
My ending CAD workstations (running CATIA V5 and Unigraphics 17) were dual Athlons with 1gb+ DDR and a Wildcat graphics board; they were half the price and faster than anything IBM and Compaq demonstrated.
+++++++
"Look, dear, it's a crazy hairy scary man!"
What the hell is a "latent failure"?
Latent: Not visible or apparent; hidden.
Its not just the transistors. What I was told, which makes sense, is that the conducting tracks in the chips are tiny and even small static shocks can damage these tracks. They aren't damaged enough to stop current flowing but the damage gets slowly worse, eventually breaking. This is what I understand to be latent failure. I'll admit this could well be the invention of the static-gear manufacturors but it makes sense scientifically and I'm not going to risk it with my servers, although that doesnt mean buying anti-static equipment...
I know somebody's gonna flame me for this,
Well they can flame me too then because you just described my anti-static precautions. I leave the PC plugged in, if you turn the wall plug off the earth is still connected so you are earthed by touching the chassis.
http://www.thinknic.com
And take all those 400mhz beauties and linux cluster them.
You should also keep in mind that the longer you wait to upgrade, the less value you will bring in from selling your old systems. What you'll have left over could draw $2,000 in the used market - enough to buy 3 of the new systems that you are proposing.
Keeping wait and that figure will go down.
on one hand you have the knowledge of having built the machine, but on the other you don't get those nice warrenties.... that's a tough one
I've worked with a school board that decided to roll their own for every single school in the county. Each school theoretically had the same setup, but what really happened was that for each phase of the project the machines changed, both in terms of their hardware and their software. This became a huge headache for replacement of the machine parts.
For example, the vendor ran out of a certain video card and so provided us with the model one-up as an RMA when one of the video cards died. This means the user gets an improved card, however what it really means is a driver headache (the ghost image for all those PCs has to be upgraded to support the new vid card as well as the old one) and a possibly non-working configuration since the new card hasn't been tested.
Then, you run into the asset management of keeping track of the different phases of the project. This was a large project, so there were lots of different configurations to keep track of. If you are planning on keeping the number of distinct phases small, you may be able to make it work. However, keep in mind that when a 2 year old computer fails, warranty or no warranty you may have trouble replacing a component with exactly the same one. Even a hard drive can cause a headache, if the replacement is larger than some BIOS or OS limitation and your ghost image fails on it.
The keys to the success of this project are as follows:
1. Always have spare parts on hand, for each phase. I'd say you'd ideally have at least one spare box, maybe two depending on the size of the phase, and maybe a few hard drives and cdroms. As you find certain parts more problematic than others, get more spares in future phases.
2. Asset management (hardware, software, firmware, drivers, etc) is crucial. You must know what's in each box; sometimes even a BIOS "upgrade" will screw you up.
3. Using something like Ghost to transfer operating systems is great until users start storing data on local disks. You will need to make sure that they at least use a separate partition, or even better a network storage.
If you keep the number of phases small, it should ideally be no problem. In theory there are only a few different PCs to fix, so once you can fix one, you can fix them all. But once you let non-standard software and hardware creep into the different boxes the theory breaks down and support becomes a nightmare.
Mark
Well whatever you decide, do us a favor and sell the old parts cheap to us.
Do like the big corporations do and liquidate those dinosaurs... At auction (like dovebid?), you could probably unload all those machines at $100 apiece provided that they are in decent condition and performance. Then buy up obtain cheap new ones at "Walmart" or "Sam's Club" for $450. Seems that you are really only saving $100 in "parts" anyway via the upgrade path.
It depends on what your users are doing...but lets do the math. If you are looking at 60 computers at roughly $600/upgrade. Why not take that $36,000 and roll it into a Win2k Terminal Server. You can build several nice multi-processor servers and let everyone connect via the client...make the server do the work. The only licensing concers would come in if you had any win9x boxes on the network. Win2k Pro shouldn't eat up a CAL. BTW...you can really lock the users down so they dont mess anything up or load those goofy screen savers and other crap.
Once in a while you can get shown the light in the strangest of places if you look at it right. -Hunter/Garcia
I do the same, but just be careful. If you are working on an ATX system, and there is no "hard kill" switch on the back of it (a rocker switch usually), then the motherboard is still getting the 3.3 volt standby voltage. It's probably a bad idea to be plugging things in with this voltage on. That said, I've done it before, without apparent ill effect, but stil, probably not a good idea.
:)
Another option if the PS doesn't have a hard kill switch is to plug it into a power strip that can be turned off, and just turn off the power strip, ground is NEVER broken in a properly designed electrical system, even when things are turned off. Of course if you broke off your ground priong so you could plug your computer in to your 60 year old house wiring, all bets are off.
In the case of the broken ground, watch out, switching power supplies without ground float the chassis around 60 volts at low current, enough to wake you up if you are a better ground than whatever the case is sitting on, but probably not too dangerous, UNLESS there is a malfunction in the power supply, in which case you could become a crispy critter. In any case, good ground is a good idea, lots of things are affected badly by floating grounds.
This same thing applies when you use a UPS and just pull the plug out to test it, without that ground reference, the ground will float, so watch out.
As far as the "to unplug or not to unplug" debate, there is probably a credible argument that even though you don't have a good ground when the case if off and unplugged, things like static electricity will dissipate whenever you touch a large metal object like the chassis, due to leakage effects.
I've had enough abrasive sigs. Kittens are cute and fuzzy.
I leave the PC plugged in, if you turn the wall plug off the earth is still connected so you are earthed by touching the chassis.
It's better than nothing, but if you did that while working on the Space Shuttle's guidance computers, you'd probably be bludgeoned. And rightly so.
Fire and Meat. Yummy.
How about laptops motherboards and cases. Have seen something just recentlyy: the ones I saw were ecs computers, sis chipset for p4(desktop cpu!).
We used to have lovely Sun 20" monitors, and as good as they were, they do start to fade after 6 years.
Plus I do like having brand-name hardware on lease, all of it is under 3-year warranty so if it ever dies, it just gets sent back. There's a lot of attraction to PHBs in having that kind of predictable hardware cost rather than episodic bouts of upgrades, "If we spend Y dollars per person per year, everyone will always have a current model PC that is under warranty on their desk".
Xix.
"Everything is adjustable, provided you have the right tools"
If you do get around to reading this, you may may to investigate some of the computer companies in your area (CompUSA or whatever passes for the american equivilant of CompuSmart). I know example that the CompuSmart store here in Saskatoon Canada that I manage offers a $129.99 build fee (CDN funds) that includes labour and 2 year coverage (warranty/parts/labour/onsite service/24-7tollfreetech-support) on all parts bought from us and put in. Ie; if your buying a case, mobo, proc/fan, ram, etc and bringing your own OS, sound, nic, video, floppy, cdrom, etc then the coverage will be on all the stuff you bought from us. Also covers the labour.
And that way you have no worries from the doomsayers regarding static, having assembly lines, or anything like that. You just buy what you need (at usually not bad prices, or heck, buy them elsewhere and bring in the parts) and then you pick them up when they're done for a measly charge of $130 a machine.... which beats spending a week building them yourself.
Let me give you a little piece of advice. Slap linux on the client machines. Spend the money and buy sizable server like a 2 or 4 way with as much memory as you can fit in it. On that server just light off gdm. Now on each client modify the inittab to query the server for a login. You will now get at least another 4 years out of those machines. You can keep all of the money you would have lost in hardware, support and licensing for day to day operations. If you value your job take my advice.
Got Code?
But there's no reason to get new monitors if the old ones work. Buy systems with no monitor. Replace the keyboards and mice, though; they wear out, and they're cheap.
Okay, let's look...
You probably make an hourly wage which makes spending time quibbling over a few hundred dollars a losing game.
If your supervisors (yeah, you probably have more than one boss, right?) are pushing you to consider this, then it's plainly time for you to bail out and find a Real Job.
Screw that ltsp shit that is a pain in the ass to work with. It is just easier to light up gdm on the server and load linux on the clients (yes on the drives). Not just modify the last line in the inittab to query the server for a login.
I can have a full up setup running in 15 minutes.
Got Code?
...the sound from a custom built system. Too loud. When was the last time you heard anything coming from a Dell or a Compaq? The corporate systems are built to be quiet from the ground up. Some tin box full of components, with 7200rpm disks is going to be much, much louder. I go through pains to keep my desktops quiet, and this would certainly be a deciding factor in someone's otherwise quiet office...like mine...our very small company has a policy about buying only corporate built PC's, and this is one of the main reasons.
Of course we've got spare machines. But plopping another workstation in its place usually isn't a very elegant solution given that it would still require reinstalling the application suites and reconfiguring the machine to get it back up and connected to our Domain and NetWare tree and such. Very often we DO give out a spare machine to a user who's in dire need, but generally it's simpler to have theirs repaired.
And of course mice/keyboards are consumable. We've got what seems to be a bottomless box ofspares we give out when they bust, but it doesn't hurt to be able to get free replacements so that the box of spares stays bottomless.
Maxim: People cannot follow directions.
Increases in truth directly with the length of time spent explaining them
... Afraid of getting my hands dirty? I'm hurt. Mayhaps you would like to see the scars on my hands from prying components out of archaic old AT cases to try and recover data from a dead system? Or maybe you just need to see my 'hand rolled' dual XP 1800+ system?
:)
Parts may be cheap, but they're not as cheap as free. The support contract is included in the cost of the PC and is alot more cost efficient than paying me to do those sort of things.
Besides, it all comes down to time management. I work in a pretty understaffed IT department. Is my time better served by installing hardware we already paid someone else to install, or by working on problems a support contract doesn't cover, like software issues or hardware from other vendors etc. It's a no brainer for me.
Though I'll admit occassionally I do insist on installing the replacement hardware myself... some of the new Dell cases are pretty nifty.
Such hostility in you...
Maxim: People cannot follow directions.
Increases in truth directly with the length of time spent explaining them
At work I am required to use a Win2k Dell Optiplex GX-150, P-III 1GHz 256MB RAM (due to outsourced IT). The machine is very sluggish and really chokes with a few apps running.
I also own for my personal use a Compaq Armada laptop Pentium 166, 64MB RAM running Debian GNU/Linux. That machine will play MP3's and run Office Apps (open office) at the same time with no choking.
I suggest you determine what apps you require, and whether there is a Linux counterpart. Then rollout a few of those P-II 400's to the new employee's (you know, the ones with little computer knowledge) running a nice clean install of Linux with about 256MB RAM and the proper apps and see how it goes. Then plan on rolling out 2-3 a week until everyone is up to speed.
Good Luck!
For a lot of businesses who have a 30 day net account or whatever, buying machines from a Mom and Pop isn't a problem, but having worked for places where it takes 30 days for accounting to pay a vendor, I've had places refuse to give us quotes as they couldn't handle the quantity that we wanted to order. A lot of smaller places can't sit on 50 grand worth of hardware for the 30 days it takes to get paid. We ran into problems with this while trying to order 20 machines for a beowulf cluster. When we finally did get a vendor that would sell to us, they had to ship the parts for 5 machines at a time since there was no way they could afford it otherwise.
I know this isn't a problem for most businesses (I work for an edu that for whatever reason doesn't have such an account), it can become an issue.
the cost of winblows and office are more than the box if you get them seperately
Chances are that you probably are depreciating the cost of the PCs you have now. You are also likely to do the same for new ones. But most accountants will have significant trouble depreciating the machines without serial numbers. As machines built yourself will not have. Plus, you need significant costs of goods to make depreciation feasible, and 60 sound cards, and 60 motherboards, and 60 CPUs, don't really add up to much individually.
Typically this is the main corporate reason for not upgrading machines piecemeal, I is often the reason that most companies buy from the Big Three computer manufactures.
I would defiantly check with accounting.
the really cool thing is the warranty factor when you do this....when you use box versions of the cpu the warranty is 3 years or more..same for all the other componets...the hard drives and video and nics all have longer warranties then if you buy them from places like gateway or dell etc
alex
--- Wherever you go, everyone is always connected...
Hardware quality: How much can I trust a popular Athlon chipset in a business environment?
Even after you've researched a compatible combination of motherboard, cpu, video card and whatever else, I strongly recommend building one prototype and running it as someone's desktop for at least a couple of weeks. Pick someone who uses the more tempermental apps in your office. One of the things you're buying from Dell and the others is a configuration that's been engineered and tested to work together. Win2K can be particularly finicky about hardware. If you haven't researched specific hardware and tested it, you'll need to add that to your time frame.
I used to run a small computer shop -- 30-100 PCs a month kind of place.
We used to do hand-builds, then eventually switched to getting 'mostly configured' systems we could then further customize for the customer.
_If_ you know exactly what you're doing, _and_ you have a number of good contacts with various suppliers, _and_ you get a good batch of parts with no incompatabilities -- in other words, best case scenario you will have:
A: A big pile of boxes to assemble. The days of jumpers and whatnot are mostly gone, but you still need to figure out how everything fits together, do it 10, 20, 30, whatever times in a row, and never break anything or have a DOA part. And even though there probably aren't a lot of jumpers, there are still finicky CMOS settings to set correctly and equivalently on all of the machines.
B: To then load everything. This is generally best done on one 'master' system with that disk image 'ghosted' onto the other hard drives. Sounds simple, but setting up that master image properly can take a while. Perhaps you'd have to do this with a Dell anyway. YMMV.
C: To deal with any integration problems -- hard drive fails? Call the hard drive vendor. Flaky problems? Oops, you couldn't afford a RAM tester or other diagnostic equipment, and so you play the swap-out game -- you pretty much need a complete computer on the side for this kind of troubleshooting. And a _lot_ of time on your hands.
And this is absolute best case. The crackpot idea of upgrading the mobo in place and re-using the hard drive, video, etc. is fine in principle, but in practice doesn't scale beyond the one-off home hobbyist sort of thing.
Worst case is that you buy parts for perhaps 20 systems, get about 14 built, RMA 3-4 hard drives, have some strange driver problems with the video cards, and get 2-3 variations of motherboard --- rev. 1, rev. 2, one yellow one green -- whatever, RAM seems to be flaky, but you're not sure if it's a CMOS setting or a bad MOBO or a bad RAM module, and if the latter, which one it might be. Start chewing through all of the permutations and eventually you figure it out and maybe get 18 of the original 20 built, the other two are constantly rotated with various users as their desktops crap out.
IMHO other than for home hobbyist use, getting a Dell/IBM/Compaq/Gateway/HP/insert favorite brand here/whatever computer beats the heck out of a roll-your-own system.
"But actually trying to use m4 as a general-purpose langage would be deeply perverse" --ESR
DON'T DO IT. I went through the same thing with a small firm. PROBLEMS: Getting the parts together, Users like shinny new boxes, billing, down time on each PC( users will have to wait while you upgrade their PC. Drivers, moving the old data ( they will blame you for any lost files for the next year) any hardware problems will be your fault because you built it. If it is a Dell then everyone can hate Dell(instead of you). Hardware support and warranties(each vender blames the others product) They will quickly forget about the money saved when their video card keeps popping out of its socket. It may sound like fun to go shopping for new hardware, and may impress the cute secretary to say you built it, but I can absolutely guarantee you will regret it as I did.
In a server based product, a development team can get a lot of mileage out of having a couple of reasonably beefy machines to do integration and testing on. What is the complaint that leads you to want to replace machines? Do builds take too long? Are the tools too slow? Instead of spending $600 on each developer, which as you know, won't really fix all their woes, how about spending $100 on memory for each person, as several people have suggested. Then buy three or four $2000 systems, and parcel them out to subprojects.
If you were a lead developer, I would suggest a slightly different tactic for those machines, and that perhaps a review of the tools the developers use might be in order, but as a sysadmin, no one is going to want to hear from you that the build process needs to be seriously overhauled.
I say this, because if your project is small enough to be considered a candidate for being shelved, and developing has gotten painful on 400 Mhz machines, then I suspect that like many projects, the build environment has grown by accretion, and no one has bothered to get off their ass and do anything about how long a build takes. Each change only adds another five to ten seconds to the build process, but over the course of a year, you get nickeled and dimed to death on these little slowdowns. And while some tools might be good enough to produce a working prototype, a lot of the IDEs end up falling to pieces as soon as you introduce source control, and serious project requirements. Sometimes it's better to just pick a good compiler, a good source control system, and a decent build tool, and chuck the rest out the window.
At my company, a coworker and I were recently pulled off of our project and onto the company's main project. For the last few weeks, we've been allocating a good bit of time to tuning the build process. What used to take 35 minutes for a build now takes just over 8 minutes on a 1Ghz machine, and can be trimmed to less than 3 minutes by turning off some more exhaustive building features. The time we spent will pay for itself in a couple of months. The last two place I worked at, I achieved similar results with less effort, but the development teams were smaller.
Currently, I'm working on a change to drop the build world time to 7 minutes, while he's working the kinks out of setting up a continuous integration machine. Once completed, people won't have to run the optional tests on their machines (most already don't). They can just rely on a 5 minute build on one beefy workstation (in this case, a Solaris box, as it's a Java/Ant/Perforce project) to tell them if they screwed up, instead of a disruptively long build cycle on their more meager development boxes.
Just because it works, doesn't mean it isn't broken.
For example, "He shot my friend and me."
I
He shot my friend and I.
autopr0n is like, down and stuff.
That's it. I've only fried a single component ever, an old hard drive, and it was because I drove around with the components rubbing against the cloth seats in my car in the sun in mid-winter, sliding around. I reached out and picked it up and felt the shock. uh-oh. Sure enough, it was toast. But it's still pretty rare.
Any connection between your reality and mine is purely coincidental.
Apart from not being stupid, don't worry too much about static discharge. It is usually pretty easy to tell if you should worry.
For example, if you touch a metal doorknob after walking over the carpet and get 'zapped'. This will blow something for sure.
Microsoft - Where would you like to go today, Maybe Jail?
Linux Terminal Server Project: http://www.lstp.org
Add Citrix MetaFrame and you get another hefty license requirement
Roll your own Linux workstations is cool - just watch for the MS hidden extras.. ;) Also be aware of the potential extra training hit. This is a small shop - so you don't want to be running around all day going through the details of how to use KDE..
Evil ZEN Scientist
Remember the article the other day when we all laughed at the Wilkes Barre IT guy who stopped the IBM maintenance on the AS/400 ? Well, this is the same kind of thinking: penny-wise and pound foolish.
Remember: support contracts are a form of insurance. They insulate you from the risk associated with the issue at hand. When looking at any form of insurance, you have to take into account what the worst-case senario is, and if you can handle it. In your case, the scenario is that you have multiple desktop failures, including critical failures of important machines (ie, severl of your main developers). Do a cost analysis: if I do a roll-my-own machine, what's the cost of it breaking? How much does it cost for that developer to have no (or a seriously inferior) machine for a week or more, vs. the 1 or 2 days a supported machine would be out?
For small companies, (especially those heavy in software development) I can't imagine a situation where the TCO of a fully-supported system is worse than a roll-your-own box. None The downtime and IT personnel time alone will kill that equation. For huge companies, it may pan out, but for a 60-desktop company with 1 IT person? Not a chance.
You need to put this into perspective with Management. Once again, they are looking at only the up-front costs, and none of the hidden costs, which in this case are the majority. Explain to them what the true cost of a desktop is, and how NOT buying a supported machine results in a WORSE return over the next year.
Now, here's a couple of recommendations for getting SUPPORTED desktops into your organization while not breaking the budget and still meeting increased performance needs:
I don't mean to harp on you personnally, but this kind of thing is why IT has a long, long way to go before being really professional. Folks, this isn't a garage. IT folks need to quite thinking like it's an expanded hobby, and also need to remind the Executives of this, too. It's a Profession, not a Trade.
-Erik
Systems/Network Architect and former SysAdmin
There are always four sides to every story: your side, their side, the truth, and what really happened.
Yes. I recycle a lot of my PC's. But one thing is, that my brother last 3 months bought all the seperate parts of his dream PC, and built it together for about a price of $800-900 (without the monitor, since u can recycle them). It includes an AMD Athlon XP 1800+, 256MB ddr ram, Soyo Dragon Plus motherboard, GeForce 3 Ti200, 40 Gb Western Digital hard drive @ 1700rpm with Raid, new TDK CD/DVD/Buner (forgot what speeds..), floppy drive, addtional 2 USB in front, MS internet keyboard, MS intelli optical mouse and a full tower Titanium case by Antec..all that for about just the price of 800-900..(and they are all probably even cheaper now)so i mean if you really can find the best prices on everything, that's great. But if u can't, then go with recycling them.
http://www.palmzone.net
Do not trust the power supply factory assembly line worker with your life! Seriously. I've gotten a good 110v dzzzt when I though I was safe, but someone who put the [temporary] building together swapped out the neutral power line.. Everything switched off at the breaker, but it was hot!
What I do is thus:
Ground the metal frame of the work bench. You'll hit it every couple of minutes with your leg and charge won't get the chance to build up.
Touch the bench / radiator / doorknob whatever's grounded immediately before you open the anti-static bag. Or if you've got you power supply plugged in, touch it, then unplug it.
If it's one of those cold dry days, or if your car door gives sparks, use the ESD strap.
On a side note [aka grain of salt]: don't use the anti-static bag to rest the motherboard on when trying things out! It's anti-static because it conducts electricity! opps, that was dumb. Luckily just a bit of solder got splatted around and it was easily fixed.
[I've been doing this pro since the 80's, only fried two things, both due to momentary lapses of reason like the above. Seen others ESDeath stuff though. d'oh!]
Not to sound too cynical, but my ORG has had this issue come up many times. The best solution is to let the users suffer with the 400's until they make enough noise to get the attention of your CFO. Then you can get the money you need to present them with a complete solution. If you solve this problem "on the cheap", you will need to solve all future IT problems in this way. I am guessing you do not get paid enough to build "home-brew" solutions for everything. Save those resources for your servers and routers! Good Luck, Tony
www.walmart.com. Get a 1ghz duron or celeron with 128 Mbyte of ram, 40gig hard drive, modem, etc. for 399 (S&H about 15). No os or monitor, but you don't need them. If you order en masse you might get a break on S&H.
A lot of people have given a nod to this point already, but I think this is worth the attention. The most important "negative" you listed, IMHO, is I don't get a proven, prepackaged system that works right out of the box. After working for years as a consultant and systems manager, I've found that it's almost always more cost-effective to leave each job in the hands of the "specialists". I don't doubt that you're entirely capable of building every one of those PC's yourself, but it's still likely that some things will go wrong, e.g. bad memory module here, inexplicable OS problems there, etc. Even one or two glitches might cost you in additional hardware, but more importantly, it will cost you in time. It's so much nicer, when a power supply explodes or a NIC goes funny, to just call the manufacturer, get the on-site service, and bang -- you're up and running a day later. The point is, take a look at the true impact on your bottom line, not only of the time investment associated with building these systems, but of maintaining them indefinitely. Good luck to you.
these things rock. Just add water.
The asus terminator get it at www.googlegear.com
if you can't do this w/ the new box for less
than $600, you shouldn't be trying to do the
up keep on the old ones. Godspeed
Small dotcoms need to prepare for the imminent release of Unreal 2 and co. Quake just doesn't cut it anymore.
I'm sorry if I haven't offended anyone
Heh, i do desktop support. I've got a Sparc 5 110Mhz with a 20 inch monitor. even with 256mb ram, the thing still goes at a good clip for what i need it to do... "Right tool for the right job."
Just do as rest of us do, take off your pants, bend over and get ready to recieve some li'l Bill. Don't bother yelling after justice because nowadays DOJ is Microsoft's bitch.
It is not the voltage that is the problem its the current (ie amperage). All the EEs around here must be asleep this morning.
We went through this at my last job. I think that you are better off to go with new machines with a 3 year warranty. Subtract the price that you can likely get for the old machines, maybe $150-$200 and you are really looking at $200-$250 more for entirely new machines.
Those parts that you are going to re-use are also getting old. Here are things that can go wrong:
- The cases have been sitting under somebody's desk collecting dust bunnies. Those fans will start sounding like lawnmowers soon. Then you will have to replace the fan or the power supply. Parts + your time = You just blew your savings there.
- Those CD ROMS have also been collecting dust and getting beat up by users. They'll start to act funny, have trouble reading disks, etc. Inevitably your boss's boss will have trouble with his. After screwing around with it he will call you to replace it. You just blew your savings and lost a little credibility.
Buy from a good manufacturer with good service. If something goes wrong, you can have them mail you a new part. You don't need to stock extras.
At my last job we made a deal with HP. We got the desktops for only about $50 more than we could build them. They came with 3 year warranties. Since we were buying so many of their machines they took care of us. Any parts we needed, we got. Every year we budgeted for 1/3 of our total machines. Now nobody has a machine older then 3 years old, they don't have any oldball machines sitting around, and OS and software deployment only need to be tested on a few types of machines. Users are happy because they have good machines and good response for the occasional hardware problem. Support folks are happy because they are bothered by bad fans, etc.
Don't forget you'll be able to ebay all those components you replaced (mb, cpu, etc.) -- this also helps defray the cost of your upgrade...maybe more than you think...ebayers seem to have an itch for outdated hardware....
remember the old amstrad 512 and 640 well if you were upgrading the 512 to 640kb ior ram you had to uinsert chips manually and they reccomemded putting it on the kitchen sink, grounded through the buildings earthing and touching metail before a chip or PCB.
Some years ago, I worked for a small company that got an even smaller company to build us some 200 PC's in the space of 2 years.
We were trying to keep prices as low as possible, so we had to cut corners here and there. So we were using 486 clones, motherboards by some unknown manufacturer etc.
After some time, we started getting wierd crashes and lockups. This was on Win3.x, and machines that weren't used much interactively (they were control PCs for CD players - long story). We sent the faulty machines back to the supplier. "There's no problem with this PC" they'd say, and give it back to us. "The CPU was loose". "The memory was unseated".
Of course, they were bullshitting, and didn't have a clue. After about 3 months, and having had lots of arguments with my management (the boss of my company was a school friend of the boss of the PC maker), I started looking at the problem. After some fun tracking down a motherboard manual, which was half in Taiwanese, I determined that they'd set one single jumper on the motherboard to the wrong voltage. Worse, it wasn't possible to set the motherboard to the correct voltage - they had just got the closest one and hoped for the best. Most of the CPU's were ok but some didn't like it, so they crashed when the extra voltage just got too much for them. Cue termination of contract...
We got a new guy to build new PC's, and I did all the repairs to existing machines myself - none of which failed again.
Moral of the story - make sure you know exactly what your doing, and watch your suppliers with the eyes of a hawk.
As for all the people talking about Linux. It's a nice OS and all that, but unfortunately many people in business use apps that simply aren't available, or don't have any useful equivalent, that run on Linux (or Unix). That's the way of the world. Unfortunately.
<posted Anonymously because some people might recognise me that I don't want to hear from>
ESD is crap. i have doing builds for years and never have had any problems whatsoever. Just touch the case first and you are good to go. Computer hardware is so much more robust than people think. I dont know how many components I have had rolling around on the floor of my truck for months, that worked flawlessly once installed.
my company mirrors yours.. same user base, same problem two years ago. i was in the same spot as you and I did build the machines myself.
two years later, would i ever do it again? nah. replace what you can replace, it's not worth the headache, and while i don't use openimage i will say that using a single vendor when possible is very nice, i have only 4 images for a company now with ~130 desktops...
if you know what you're doing, it will be a fun project... but i'll tell you building all those machines gets pretty damn tedious.
First off P4's suck. They stack up to be about 1/2 clock for clock with the P3. Unless you are running heavily optimized apps (you aren't) P4's are a waste of money.
Doing upgrades yourself is just stupid. Are you a computer manufacturer? Thats what I thought. Putting together desktops are a onesy-twosy proposition for hobbyist. Doing it over several hundred is plain dumb.
The only desktops to upgrade are machines that have a second processor slot, or machines designed for CAD/CAM with upgradable Oxygen video cards.
Any other upgrades are just code for "I'm going to upgrade my home computer for free".
Hey, just be smart and buy off lease P3's - you can get $800 gigahertz desktops.
Good luck, and may the force be with you
I admit it, I am careless about anti-static precautions. I don't wear a wriststrap, I don't stand on a rubber mat, and I don't have a proper workbench. I do most of my computer work on the dining room table, which is fine as long as I don't set the hard drive on the (open) butter dish and get done before dinner.
However... I live in a very humid climate. Most of the year, I can't build up a noticeable static charge even if I shuffle my sock-clad feet across the shag carpet. Things change in the winter, when it gets dry; I try to avoid working on the internals of my computers then.
I do ground myself to bleed off any static before opening my computers or handling any components. So far, things seem to work, though I do have the occasional brain-damaged motherboard--which I can't tell if it's because of my sloppy anti-static habits or because I buy cheap motherboards. However, the vendor is pretty no-hassle about replacing bad ones, and they are cheap enough in the first place that replacing a slightly flaky one two years down the road when I want to upgrade to the next CPU anyway is no big deal.
YMMV. Frankly, I probably couldn't get away with my current practices in a drier climate.
---dragoness
Ok, posting anonymously so as to not identify the guilty (possibly myself). :-)
I work at a university. Budget is almost non-existent. I'm on a fixed salary, so often my time is considered "free". Bleh. That said, I quit rolling my own boxes because it took too much of my time. Time spent researching, building, troubleshooting, etc... We established a replacement plan to replace 1/3 of our machines a year. We don't always meet it, but we do often enough to get by. Pick a vendor. Our choice was Dell, but you should choose one that YOU like to deal with. We don't buy cutting edge systems. We buy mid-range systems with the least amount of RAM we can get. We then upgrade the RAM from a cheaper vendor (for what it's worth www.crucial.com has been good for us). At least with our vendor of choice, our platforms aren't identical, but they are close. The system vendor has a subset of brands that they seem to stick to so that you can keep a standard set of drivers around so that you don't have to constantly download new drivers or figure out what it takes to get something to work with a particular package (linux for instance). We still do roll a few boxes, but they are mainly for low-cost servers (ata drives + linux + ext3 + software raid + samba = cheap reliable storage).
Licensing is another issue altogether, and I really wish it wasn't such a pain. The entity at the university I work for, does its darnest to keep legal. I don't know how we'd fair on an audit though. We use bulk licensing with our own inventory system. It leaves something to be desired, but is better than many I have seen. Microsoft has gone out of its way, in my opinion, to make software licensing and registration a pain. Thus I avoid it where practical. I won't even attempt advice on license questions, but I had to rant for a moment.
As to the discussions on ESD, I feel a bit (maybe not much) more qualified to offer an OPINION. ESD is an issue, but one that is often exaggerated. I have a degree in electrical engineering. We studied ESD. Yes you need to take some precautions. Touch big metal things often. If you can get a wrist strap that's great. I encourage you to use one if building systems in mass. If you are out "in the field" working on a system: touch big metal things often, and don't wear things that tend to build up static charge. If you are working on a system with a metal case, touch the case, or better yet put the case in your lap / try to keep some part of your body in contact with it. Even if the case is not grounded, keep yourself at the same electrical potential as the case. As to the effects of "latent damage", it can occur. Many devices can take a lot more voltage/current than they are rated for for a BRIEF period of time. Doesn't mean its good for them. You may not (to quote previous posters) "punch a hole in the glass", but you may "chip" it or weaken it so that the next time it is put under stress (be it another ESD hit, power surge, etc) -- it fails. Many devices now have protection diodes in them that prevent currents from flowing the wrong way. This helps, but given sufficient voltage, they too will reverse bias and allow current in. All that said, I've never personally had any problem with computer components failing due to ESD; however I do try to take precautions. However, back when I was in school, I did nuke a loose chip or two walking across carpet (dumb on my part).
If you're looking at a $400 difference, perhaps you can recoup some of that by selling off the old systems at auction. I'm sure they're worth $200-$300. Then you have the benefits of new supported machines from the big guys, without the hassles of supporting your own systems, and the cost factor is minimalized.
Just be sure to erase those drives several times, and make sure you conform to whatever license transfer blah blah M$ wants.
Actually, it is the voltage.
Since we're talking STATIC electricity, there is, by definition, NO current flow.
The problem with static charge buildup and FETs is called punchthrough. Electric field strength is measured in Volts per Meter (V/m) The gate thickness of a typical CMOS FET is on the order of nanometers (1x10^-9 m).
Our 15 kV static voltage produces in the gate region of the FET a field of (15kV)/(40x10^-9 m) = 375x10^9 V/m.
According to this link the dielectric strength of (Pyrex) glass 14x10^6 V/m. Applying a field stronger than this will cause ionization of the material: electrons will be literally knocked off their atoms! This ionization allows a current to flow through the (normally) insulative material, called dielectric breakdown. In a CMOS FET, gate insulator ionization leaves residual conduction paths, ruining the transistor (punchthrough).
CMOS FETS have very thin gate insulators to increase performance, but the side-effect is that they can tolerate only very small static gate voltages without damage.
I posted earlier in support of buying new PC's. Just thought I'd throw out another idea: lease them. Given that PC's are business tools that aren't going away, you might be better off with a 24 month lease. Though terms of the lease itself would indicate whether it was beneficial or not. Your company avoids the initial expenditure of funds to purchase your own PC's. Just a thought.
You're in a company where money is tight, but 400mhz desktops aren't good enough? What are you doing, 3D Rendering?
Or is money not so tight that the company shouldn't spring for you guys to have better Quake machines?
Just do it yourself you pussy.
I run Win2k on my P166/64MB laptop, myself, with fairly minimal swap (it could really use another 64MB, which I'm going to get soon). It runs fine overall, just boots a tad slow. I can only imagine that a 400 Mhz box would be even better...
I sing the doggie electric!
I am commenting as someone who has scratch-built many PCs, both for home and business use. First, I'll assume you know what your needs are, and not try to tell you that your 400 Mhz Pentiums are just fine. You said you need to upgrade, I'll take your word for it. I'm also not going to tell you exactly what you should buy, I assume you know what you need/want. And your Win2K license terms aren't my problem, either. Other people have also commented well on anti-static issues.
First, don't start this job until you are comfortable tearing computers apart and putting them back together. Building and repairing computers is fairly simple these days, when everything is componentized, but you do have to know what you are doing. You need to be able to understand those motherboard manuals and figure out what jumpers and BIOS settings you need for your particular configuration. You need to be able to screw motherboards into place and shove cards into slots without breaking them or slicing yourself open on the chassis (I swear every one of my personal computers is christened with my blood!), and plug cables in right-side up. All simple things to learn, but they can be expensive and frustrating to learn the hard way. If you're not comfortable doing these things, don't plan on building 60 PCs yourself. Farm the job out to a good local vendor or technician who is.
Line up a good vendor, either local or mail order, who can sell you what you need, when you need it (finding out replacement parts are unavailable or back-ordered for a month when you need them NOW is not helpful), at a satisfactory price and with a no-hassle return policy--because you will be returning bad components when you order enough for 60 PCs--unless you pay the higher price for a vendor that does 24-hour burn-in. Even then you may not weed out all the bad components.
Make your PCs as much alike as possible--it's easier to assemble a cookie-cutter configuration, and of course, ghosting a Win installation works a lot better if you're using the same drivers from computer to computer. As others have mentioned, don't cheap out on the components! Good quality, name-brand components are worth paying a few dollars extra for; you get fewer returns and mysterious failures, and name-brand quality components are more likely to actually follow the industry specs for whatever device they are, instead of cutting corners the way cheap components sometimes do. BTW, this is where you win over buying cheap pre-built computers: guys like Gateway and those Wal-Mart computers save money by putting the absolutely cheapest, bottom-of-the-line, no-name commodity parts in their computers. That's how they can sell them so cheap. Sometimes it works; back in the early 90s, the favorite no-name graphics card used in our company's computers had the Cirrus Logic chipset, which was a moderately accellerated, halfway decent graphics card
that actually had OS/2 drivers (which we were using). Usually, you have the problem with discount computers that the cheapest no-name card changes from week to week, so this week's discount computer may have entirely different components and drivers than last week's discount computer, even though they are supposedly the same model. Now that is a major hassle in the support department!
OTOH, some parts are so commodity that it doesn't matter. Who cares what brand floppy drive you buy? It's a mature technology and they all work alike. IDE CD-ROM drives are much the same way. IDE hard drives are NOT. Neither are SCSI drives.
I personally like Western Digital IDE drives and won't touch a Quantum if I can help it; YMMV.
If you're using AMD Athlons or similar chips, invest in a slot fan or bay fan in addition to the CPU fan. If the noise of all those fans is likely to drive people postal in a week, consider spending the extra dollars for low-noise fans.
So, you've got a vendor or three, and you've got a list of parts that meet your criteria for price, performance and quality. To lower your own frustration level, make sure you have plenty of tools; those Phillips-head screwdrivers and nut drivers seem to migrate of their own accord whenever you're not holding them in hand. Also, make sure you have plenty of small screws of various sizes, spare Y-junction internal power cables, and spare IDE cables. Save any leftover small screws that came with cases or whatever; you'll need them sooner or later. Spare mounting rails of various flavors are nice to have around; vendors never seem to ship the right mounting rails for your chassis, if they bother to ship mounting rails at all with the drives. If you are lucky, your chassis's don't need mounting rails at all, but support drives being bolted directly to the chassis. Wish mine did.
If an IDE drive doesn't work, check your master/slave jumper settings first, then the IDE cable (that's why you need spares--I've had a lot more bad cables than I ever had bad drives). Keep a "known good" AGP card around to test out the AGP slot when you think you have a bad graphics card--I've had more bad AGP slots on motherboards than I've had bad graphics cards or bad monitors.
Ditto for memory and memory sockets. (The quality control on certain brands *cough*SOYO*cough* of VIA-chipset motherboards was a bit off...) Also, watch the fun-n-games of putting PCI cards that don't share interrupts happily (NIC & AGP combo, particularly) in the wrong slots.
Being able to ghost the first OS + software installation onto all subsequent PCs is a major time and hassle saver.
As for "support" issues, if you can put together the PCs yourself, you can handle most support issues yourself. PC hardware is commoditized and componentized, and a hell of a lot easier to support than PC software. Keep "known good" components around for troubleshooting, and have spares of everything on hand, including and especially power supplies. (Make sure you get an adequate power supply in the first place).
Anyway, hope this helps.....
---dragoness
A good way to tell whether or not you should use I or me at the end of a sentence is to see what it would be like without the second object.
For example, "He shot my friend and me."
"He shot I" would sound rather silly, wouldn't it? "He shot me" sounds much better, and should be grammatically correct as "He shot my friend and me." "He shot my friend and I." may sound better, but it is wrong...at least, last time I checked.
I sing the doggie electric!
I used to be a big believer in the "more RAM is what you need" mantra, but have recently changed my mind. If you're working with files that are larger than half the amount of RAM you've got, you're going to be swapping to disk for temp space. It's inevitable. I do a lot of graphics work on very large image files (1+ gigs) and no amount of RAM is going to save my 1.7ghz AMD from taking at least 10 minutes to load up the file.
What WILL work is to get a couple of fast HD's and RAID-0 them together. If data redundancy is crutial, you'll have to backup your data often. But the speed difference is ASTOUNDING. After upgrading my single 7200 rpm drive for two 7200's RAIDed together, I can now load these monster files in a little over three minutes. And HD's are cheap nowadays. I purchased a couple of 7200 40 gig Maxtor's for about $75 each, new, and THAT was probably more than I should have paid, but I was simply fed up. I knew the bottleneck wasn't in the processor or RAM. Any moving-part system is going to be orders of magnitude slower than electrons moving around silicon.
Let's see...
New Dell = $1000, Old PC = $-100. Total per PC = $900.
Upgraded AMD = $600, Old PC = $-0. Total per PC = $600.
900 > 600.
Now, how is getting a new dell better than picking your own hardware out (for less) again?
IBM bailed out of the hard drive market.
One guy worked all day without a good strap and everything he touched zapped. He fried about 20,000 cpus that day.
Was this guy licking the sockets or using the chips to fluff up shag carpet?
This is pure BS. Don't touch the pins! Hold chips and PCBs by the edges in a non-conductive area. If you follow this rule, you can run around on shag carpet and never fry a single part.
We just trashed (donated to school) all our PCs and replaced them w/ neostations. Only cost $299 to put a new box on each desktop. "is a Linux-based thin client that lets you run the latest Windows-based applications... connect to Win2k servers or UNIX/linux servers via terminal emulation, use your existing serial and parallel printers...and is highly secure since everything resides on the server." With all our applications on the server we just upgrade/configure/secure one machine instead of a 100+.
A few years later, I had my dissertation laptop on my lap, and a mug of good beer (home made) in my right hand. Without warning, my wife plopped one of the rahter new (3 months?) twins in my left arm. Immediately, she kicked, spilling beer on the keyboard of the new (to me) thinkpad, thus becoming BoKD.
The keyboard wasn't happy. Heck, the bios wouldn't even boot due to the errors. When all else failed, I let the keyboard hang from the case so that the case could support it in the sink, removed the battery, and let the keyboard soak overnight. It took three days to dry, but all the keys worked again. I did lose the rubber springs for two of the keys when my wife spilled the bowl, though . .
Anyway, if you need to do this, most keyboards will survive it (besides, what do you have to lose?). Distilled water is a better idea than tap water, though . .
More recently, a friend sent an older powermac for my kids. She used to be a chain smoker; you could smell the machine from a few feet away. I took the opportunity, and had my disgusted oldest daughter (the one with the apple juice; now 8 years older) help clean it. I'm reasonably sure that after seeing that (along with some comments from me about lungs) she'll never smoke
She had to wash the pieces in the sink
I removed the electronics from the keyboard and ran it through the dishwasher without soap. It didn't survive
hawk
:)
hawk
We have a couple of Athlon systems at our company, with some mixed results. I've found that if you load the programs you need on the system, don't make a lot of changes (other than the needed and frequent OS updates), and don't play around with them (loadiing/unloading freeware, shareware, etc), they are stable.
r0wan
If you're not outraged, you're not paying attention.
Hey! If you're wasting time, commenting on such an insignificant task, shouldn't _you_ bail?
If you're with such a hearty shop, don't you have more productive, consequential tasks at hand?
In my past experience as a small town computer consultant (and working for one of the big names) I've had to struggle with the same question. It's very tempting to build a system yourself. If you're like me, it's fun to do and you think it might give you a chance to "build the best". Don't fall into this trap!
Believe me, noone cares about your technical prowess, they won't be impressed. And when the slightest thing goes wrong, including M$ errors, you are to blame. Also it is really irresponsible from the company standpoint to have one lone guy that knows how their systems were built, how to maintain them, and where to go for help.
But the most immediate point is the cost. With Gateway or Dell you can get a very large discount with a volume of 60 systems. You can purchase the systems(complete) with a fresh copy of Win2k for $500 to $700 a piece depending on your options. On top of this you will have triple the support: from the assembler(Dell, IBM, etc), from you, and also the manufacturer of the hardware. Most people forget this fact but that little Seagate (or any other) HD contains it's own 3 to 5 year warranty and many other parts do too.
I would submit your needs to both Dell and Gateway and see who will take care of your needs the best. Let them fight. I used to work for Gateway and it was a real challenge to take business orders from Dell.
Bottom line: You have a big advantage here...let the bigs fight, lower your cost, raise your support, and find a better use for your company time.
High humidity will prevent static buildup better than any strap or conductive shoes. I can't even get a Van de Graff generator to work here in Florida during the summer! It just shorts out in the air. Dry skin is no problem either. I wonder if a room humidifier (nebulizer) would help in dry climates. You could rub the cat with a rubber glove to check . . .
I'll bite. The reason why I'll flame you for leaving a machine plugged in while you work on it is that the ATX motherboard spec gives power to the motherboard while the machine is off. Even if it didn't, you still have a great ground to the machine, which leaves in the slight possibility of electrocuting yourself if you find a power source (why do you think that the static wrist straps have resistors in them?). The trick should be, ground the chassis (with the resister), then make sure your are touching the chassis while you are working on the machine.
Just my $.02
it was a bad scene. It's because he was wearing a thick wool sweater under his smock
This is why I do all my computer assembly in the nude!
smock.
We partially built machines from scratch. We ordered all the parts from once place with specifc model numbers and specs. We got about 20 in and they were great. We created a ghost, and everything was fine until we recieved another 20. They all had slightly different motherboards and video cards... they were better, which is why the company thought it was okay to replace them, but it caused complete havoc because we didn't know they changed them, and had no idea why things weren't working. Now we have to have three ghosts for the three sets, various driver disks labeled appropriately, and have to go through the building labeling the computers as having specific hardware. It was a mess. Moral is order from someone with large stock of the same model number, or make damn sure the company doesn't try to do you any "favors" by upgrading.
I have the same position for another dotcom. Not too long ago I was in a similar position. It is tempting to try to get the best machines for the $ for your company-Like you are shopping for yourself. Do not do this. Don't put yourself in a spot where you can be personally responsible for the productivity (or lack of) for the entire company.
I would suggest dumping the 3+ year old 400MHz machines before the drives start to crash.
If possible replace them all at once.
1. Buy inexpensive machines from a major manufacturer that offers oem licenses and hardware support and put it on a lease. Management will be much more likely to agree to a monthly fee than a lump sum bill.
2. Format your old machines reinstall the os and sell them on ebay.
Read everything the other posters have written about why you should buy Dell, Gateway, etc. prebuilt systems. Read it again. They're right.
One additional thought that I haven't seen expressed yet in this discussion: everything is negotiable. Here are examples of what I mean:
A couple of posters have touched on warrantee treatment. Let me add one thought: go with the supplier's on site parts and labour three year warrantee. This means one less head ache for you. If you keep the machine configurations identical, this allows you to quickly swap components around to get a particular individual back up and running in an emergency (e.g, on a weekend). Also, if you do negotiate 60 machines in one shot, have the supplier commit to haveing a service technician on site once a day during the first week---that will ensure any DOA situations are fixed quickly (and your users will be happy).
I'd say no. Not because its a bad idea, but because you're asking.
If you can't tally in your head the pros and cons of a homebuilt vs an OEM solution, I have to doubt that you have the knowledge and experience needed to build and maintain 50 machines on your own. From the sound of your article, you seem to be very price driven, and buying cheaper parts is usually hell in the long run. (Well, 'cept the cheap cnet nics for $9 apeice. Won't run under linux, and the performance isn't great, but they are rock stable under windows and I've never seen one go out).
That being said, if I had to homebuild machines for the office, here's what I'd do. First of all, everyone doesn't need the latest and greatest. Break the office into 2 or 3 groups. First group gets the highend stuff, second group gets the average stuff, and the third group gets whatever will run a basic wordproc and email client. Thus, you have a machine rotation path, first->second->third. Those 400 mhz machines sound great for the third group, and maybe even the second, with a memory upgrade. Btw, this is a great lesson for you: a lack of memory and slow hard drives will make even the fastest machine seem slow - thus don't skimp on memory to buy a faster CPU and don't stick a 3 gig HDD into a 1700+ Athlon XP.
For the machines, but a quality motherboard. You don't want to go for the top performer, but for stability. Right now, I've had great experiences with the Gigabyte GA-7VTXH+ (Socket A DDR, 100/133 mhz bus, built in creative sound, realtek nic) and the Tyan S2390B (Socket A, 100/133 mhz bus, no sound/lan). Buy memory from a trusted supplier (mushkin & crucial seem nice), and use memtest86 for a few hours per machine for testing, and burncpu from a floppy (I'd suggest tomsrbt). Since you will support these machines, you want them to be stable. Also, grab yourself a large (locking) file cabinet, give a number to each machine, and store all manuals/cds/floppys/software in the file cabinet in a folder with the machine's number. That way, software audits are easy, users won't be able to install software on unauthorized machines, and you'll always have the documentation. For hard drives, use 40 giggers. They won't need the room (since they'll be putting all essential data on the server), but a 40 gig seems to be the optimal price/size ratio. Throw a cheap 8 or 32 meg vid card in with good 2D support and no history in the usenet archives of having problems with 9x/NT/2k/XP, use a quality floppy/cd drive, and you're set. All windows installations are scriptable, or else you can ghost the drives, and make sure you install a good antivirus client (Norton AV Corp is expensive, but nice).
The advantage is, you get a machine that's built your way, without all the added crap software OEMs throw on them. The disadvantage is that you lose tech support, so you probably need to develope the skills to type in "groups.google.com" in your browser and search for problems. Anyone with an A+ and half a brain in their head can maintain their machines, which doesn't make it easy, since there is a severe lack of people out there with at least half a brain in their head. (And don't get me started about "teach to the test" A+ courses...)
Just my $.02
I've seen the way Compaq builds PC's. I would recommend building them yourself. If your company needs the hand holding, a repair contract might suffice. When the quantity is under 200 building your own right makes sense.
Don't buy compaq computers that have a serial number beginning with the letter U. That's Indianapolis and they persecute their own quality control people into not failing computes because corprate might notice the QC reports.
Don't worry.. I'll whack you with my meta-mod stick!
You want lower cost-per-seat and easier maintenence overall? Spend half that money on a powerful server, and convert the desktop machines for use as thin clients. No more tweaking settings on each system! No more cleaning up after settings screwed up by users on each system!
I'm surprised not to see more references to the stories about other organizations doing this, such as:
Binson's Hospital Supplies
Newspaper Association of America
(vendor) Integrity Networking Systems
City of Largo
Home Depot
. . .
And if you'd like to really cut down on MS licenses, don't forget about Crossover Office.
"...Acceptable P4 systems from the big guys run at least $1000. By recycling the OS (Win2k), case, cdrom, floppy, and K/V/M..."
Nice try, but there is a little known clause in the EULA for M$ OS' that prohibit the re-use. I ran into that before,luckilly not the hard way.
Thats one reason they now put the non-removeable stickers on the cases, so you cant take the license cert with you when ya go.
Somehow I don't believe you, sorry. I run Win2k on a Celeron 366/64 Ram and I really dislike using it. Sure the basic OS runs fine, but just wait till you run any apps....
I've seen that 30V figure mentioned twice now. Where did that come from? All the modern (TTL voltage) CMOS parts I've ever seen had a maximum bias voltage of 5.8 or 6V (OK, with the exception of Vpp on PROMS). No individual pin is spec'ed to tolerate voltage above the Vcc(+5)(+3 or less on RAMs and CPUs nowdays!) supply nor below ground. I once encountered a power supply that had the unusual "feature" (more likely a regulatory failure) of ramping its +5 output to about 8V on startup, for only maybe 15 or 20 mS before it fell back where it belonged, but this burned up two motherboard/disk sets before I figured out WTF was going on. _Very_ few things in a computer cabinet can withstand a 30V bias. Maybe a 30V ESD pulse? This is nothing, simply moving around in fairly dry air will result in >300V static potential, and you'll never feel it, either.
That said, I'll agree that the ESD hazard is overstated.
A few simple rules keep me out of trouble:
Keep a rather nervous frame of mind, which will make your hands sweat, or turn up the AC a bit. No, not so hot that you drip or anything. Sweat is an ionic solution (salt) and a natural antistatic coating for the human body.
Sit down. Sit at the table, keep your feet still, put your elbows on the table. Standing beside a bench shuffling your feet on the floor is a pretty good way to generate a static charge. Sticking your then-charged hand into a box, without first touching the box and discharging the static, is asking for it.
Hold the cabinet in one hand and the component you're dealing with in the other. If your task requires both hands, lean an elbow or forearm on the edge of the cabinet. Try to keep contact with a metal surface (one painted with EMI/RFI suppression paint will do) of the cabinet as much as possible. This eliminates static charge, much as a bracelet with a wire would, without cuffing you to the cabinet. "Grounding" the bench and/or cabinet is irrelevant. Equalizing the static electronic potential between your body and the gear you're working on is vital.
To pass (unbagged) equipment from one person to another, touch the person's hand first, to discharge any potential difference, then you can touch the gear they're holding. Be sure to explain this as a static precaution so they won't think you're gay.
Keep the gear bagged as much as possible. Keep disk drives in their shock packaging, this is usually antistatic as well. The spectacle of stacks of naked motherboards sitting on shelves at the local used hardware dealer makes me queasy. I don't buy that guy's motherboards.
I leave the power supply plugged into a grounded outlet while working on the machine.
IMU a modern ATX PS doesn't ever really fully power down the motherboard. Touching (with your hands) the 3/5/12 volts you'll find there isn't a hazard, but working with metal tools isn't such a good idea. You're of course safe with a (now old-fashioned) AT power supply that _fully_ shuts off.
"_Nothing_ works after you let out the Magic Smoke(tm)."
Exceeding the recommended torque is not recommended.
Check out something cheaper like Compaq Evo D300v. At $569, it's comparable to what you'd spend to build your own and it would be all new components. It even comes with Windows XP and a 1-year on-site warranty (can be extended to 3-years for $99 more). Other OEMs have similar packages available, but I'll use Compaq as an example since that's what I'm familiar with.
Why do I think that this is a better solution? Here's why:
1. It's pre-built and will save you the time and hassle of selecting, assembling, and testing components. Compaq has taken care of the problems of making sure that all the bits are ccompatible, so there's no headaches over wondering if the problem that you are experiencing is a hardware conflict or something else.
2. If something breaks, call 1-800-OK-COMPAQ and have them fix it. It's not your problem. That's much easier than trying to get a vendor or manufacturer to provide warranty service on a mainboard that died. More importantly, a component vendor or manufacturer might require you to ship components back and forth (if you're lucky, just cross-ship) before determining that there needs to be a warranty replacement. You're potentially looking at 2+ weeks to replace a faulty component. With an OEM like Compaq you can have the problem part replaced next business day. That's a big difference in a business setting.
3. If something breaks after you leave the company, your replacement will know who to call for assistance. They won't have to worry about checking warranty status on a widget and then getting a vendor to replace it. This offers peace of mind for the business principals.
4. If you need updated device drivers, they're all in one place at www.compaq.com. Locating and downloading softpaks is far more convenient than scouring the Internet for the latest versions of somebody's reference drivers.
5. If you want it, you'll have access to Compaq Insight Management Tools. You don't use anything like them now probably, and you may not in the future. But at least you'll have the option.
6. As far as system specs go, you'll get a 1.3GHz processor, 20 GB hard disk, an Intel NIC, 128 MB of RAM and decent integrated video and sound. The average office worker doesn't even come close to needing the power of a 1.3 GHz system. Sure you could get faster parts, but unless you're doing 3D modeling (in which case you're already shelling out $3000-$4000 for a high-end video card) or running intensive engineering or financial simulations you'll be OK for several years with what's included. Granted, you wouldn't want this on your desk at home for gaming but it's a pretty decent work machine.
7. Microsoft Windows is included. OK, a lot of people think that's not necessarily a good thing. But if you're going to be running Windows anyway, you might as well get it included so that you can be legal with it. Building your own PCs would prevent you from transferring your OEM licenses from the old machines to a new machine (legally speaking anyways).
In my opinion it's a no-brainer.
The reason that AMDs seem to do "poorly" in business, is that the examples used to prove this point show mass produced systems that are faulty anyway. The same could be said of celerons (if you based your information on Hewlette Packard Celeron boxes.) The athlon systems work great, if you put it together with a quality motherboard, and memory. If your goal is a
"cheap" system, then an athlon (or any system) will run poorly. mhop
----- LoboSoft specializes in Digital Language Lab
remember the old amstrad 512 and 640 well if you were upgrading the 512 to 640kb ior ram you had to uinsert chips manually and they reccomemded putting it on the kitchen sink, grounded through the buildings earthing and touching metail before a chip or PCB.
Yeah. It was a new thing, all this new-fangled CMOS. :) The good old days, when motherboards were full of TTL logic (SN74xx), the Internet connection was a 300 baud acoustic-coupled modem, and UUCP e-mail was too complicated for spammers to have figured out.
[sigh]
Fire and Meat. Yummy.
A platform has an API.
What you have are manufacturers.
That's the $65,536 question.
My kid brother bought an OEM XP license from Fry's for $100. (I sent him to confession. It was Holy Week, and he buys XP!) Installable, not an upgrade. They required him to buy "some hardware" in order to qualify it as computer sale, with which they can sell an OEM license. He bought a kit of cabinet fasteners, you know the ones with disk drive screws and those plastic standoffs that you need to mount a mobo. Let's see the Empire figure out which of those screws did or did not get built into his computer.....
Exceeding the recommended torque is not recommended.
Right now, I'm running Word 97, Trillian (with about a dozen open windows), Outlook Express and 3 instances of IE. No complaints, really. There's a bit of churn, sure. Like I said, I could use more RAM. It's fairly responsive, though.
I sing the doggie electric!
Also, the computer doesn't want any beer, no matter how much you think it does. NEVER, EVER feed your computer beer, it doesn't
need it and is prob[ab]ly too young for it anyways.
My old 5160 (PC/XT) is about there. Of course I've gutted the cabinet, and replaced all its insides with newer stuff. Does that count?
Exceeding the recommended torque is not recommended.
I dont see why companies are refusing to use AMD based machines because of the possible reliability problems, Poor software, from one vendor in particular, causes far more downtime than hardware failures, and no-one`s refusing to buy their products due to their laughable reliability track record.
http://spamdecoy.net - free throwaway anonymous email - avoid spam!
I hear a lot of people saying that you shouldn't; Well, here's three good reasons you should:
:-)
:-). But, I would suggest that you recycle your copies of Windows. If you erase the copies from the old machines, logic suggests that even MS wouldn't have issues with it since you are still in possession of only one functioning copy of Windows per license.
Job security!!! If you built those machines for your company, they're going to want you around to fix them. If you do the job right, you won't ever need to fix them, but they don't need to know that
Quality!!! I've never seen an OEM system that could keep up with my home built systems when it comes to quality.
Speed of repairs!!! OEM's often take weeks to replace or repair systems. When you're local, it'll be no trouble to fix them in 1-2 days. Often, it can be done with less than 2 hours of downtime. Besides, if you roll your own, you won't have to deal with the awful tech support person who treats you like you don't know WTF you're talking about.
Regarding your O/S choice: I won't tell you that you have to go with Linux. Many of your end-users wouldn't know what to do with all that power
# They who can give up essential liberty to obtain a little temporary safety, deserve neither liberty nor safety. --Fran
Maybe because mine is an upgraded version from Win98. However, that shouldn't matter.
By RAM if you need it, but why are you upgrading 400MHz PCs? Seriously, stick some memory in the things and let them go. Who knows for sure what you are using the systems for, but most office type apps do not require serious horsepower to operate. If however, you must build systems then go with a good barebones solution. The ones from http://aberdeeninc.com are decent and way less than the figure you mentioned. Good luck.
Good point well made - you do need to consider running multiple servers for redundany and a blended environment, with both fat and thin client gives you the best of both worlds: run the graphic-intensive stuff on the fat and email etc on the thin ... can all be done on your existing hardware...
I am a leaf on the wind
Try KOffice or AmiPro under FluxBox on a cut-down Linux kernel, it'll fly on those boxes - and keep OpenOffice.org around for compatibility reasons, even if it runs like a 3-legged centipede in 32M. Alternatively, run them LTSP off a single decent server.
Got time? Spend some of it coding or testing
Just to address the 30V thing, I actually have no idea what the tolerance of a CMOS transistor is. I was getting that number from an earlier post (which I am waay too lazy to search for) and to me that was as good a number as any. The point still stands that ESD voltages are out of the ballpark when compared to the tolerances.
alex
--- Wherever you go, everyone is always connected...
there really is a site called gotapex.com
fucking go to it jeff kim you fucking faggot.
Like at http://www.TCWO.com where you can get a amd 1.33 system for around $449. This comes with everything except the monitor. They have other systems, I just used this as ex.
ER
Naw, it shouldn't. I may just have lesser standards of speed than others. Whatever works, I suppose.
I sing the doggie electric!