Slashdot Mirror
Projected 'Average' Longhorn System Is A Whopper
greg_barton writes "At first I thought this was a joke, but this article from Microsoft Watch confirms it: 'Microsoft is expected to recommend that the 'average' Longhorn PC feature a dual-core CPU running at 4 to 6GHz; a minimum of 2 gigs of RAM; up to a terabyte of storage; a 1 Gbit, built-in, Ethernet-wired port and an 802.11g wireless link; and a graphics processor that runs three times faster than those on the market today.'"
By the time Longhorn comes out I would imagine that it is a pretty normal requirment. 2 years from now is a long time in the PC world. Keep in mind that the average home users is close to (if not above) 3.0 HT procs today....
but this article from Microsoft Watch confirms it
According to the article it's not a confirmation at all. Microsoft has released no official statments about hardware requirements, these values are just estimates from developers, who may or may not have a clue.
Of course if it is accurate, then wow.
I think they mentioned a dual-core processor, not a dual processor. Thats a big difference. Intel's Hyperthreading is a step towards a dual-core and we have it today. With the work I've seen from Sun and IBM on dual core processors I'd expect we see this sooner than later.
One should not theorize before one has data. -Sherlock Holmes-
forevery bit of data. you need a four bit EULA.
think about it.
My days of not taking you seriously are certainly coming to a middle...
Some people whined when Apple did away with the then-standard floppy disk drive. Occasionally companies need to push forward (even if it is in x years time!).
And this is the suggested system that would have the OS running at its best.
Maybe they just want to give hardware companies something to aim for, and hardware resellers something to look forward to as masses of users upgrade their computers.
'Thats they exact same thing a banana wrench monkey.'
With that kind of hardware it had better at least make toast too... In terms of 'average' hardware requirements I think that we're stretching it as it is even for XP's requirements. Just another case of bloatware keeping the Wintel monopoly going I guess. More eyecandy and everything in a different place but not necessarily any more functionality. Whatever happened to high performance software that pushed the limits of what was possible? I'm thinking BeOS here. It ran crazy-fast even on ancient hardware! Fortunately Linux is good but not quite that fast yet (on the desktop in comparison to BeOS).
I'm afraid I have to agree with this post. I dual-boot and IE beats the crap out of firefox on my machine.
Konqueror seems much faster though, on par with IE.
I think it's indicative of Microsoft's own expectation of Longhorn's release date. Much better than estimates put out by the PR department or MS fanboys.
In Soviet Russia, articles before post read *you*!
This is probably about right: just remember that even though Longhorn may arrive in 2005/2006, it is likely to have an expected product lifetime of (say) 5-10 years (think Windows NT/2000/XP). This means that the average is planted somewhere midway into the envelope, say 2-3 years. I'm guessing that by 2008, these technology characteristics are properly not too far off base.
I'm sure someone could sit down and do the numbers for us by extrapolating on CPU and hard drive rates and moore's law as it has occurred over the past couple of years.
I mean, design is all about tradeoffs: we don't design in assembler any more because the playing field has moved on. We don't design UI's from scratch, we use UI 'builders'. In the same manner, we don't design for todays technology when we expect our design to work with tomorrows.
If Linux didn't design for MP and scalability now, then it would be hosed by the time MP became "default" for the desktop (well, in fact, with HT, it already is!). Yet, designing for MP now causes some performance and related loss even though the technology is not here.
Who am I trying to lecture Engineering and Economics 101 to the
Yes he did.
Okay. Then cite it.
Quoting Wired is the sign of ignorance.
Wired quoted an interview in the Boston Herald. Next!
I don't think thats ridiculous at all!!
I'm already running a machine at 2GHz, with 1 1/4 gigs of ram, and 300 gigs of HDD space.
4 years ago I was running a machine with 300 MHz, 64MB of RAM, and 4Gb of HDD space.
So in that same timespan, we only have to progress about the same amount. Doesn't seem impossible to me!! C
On the other hand, I'd be perfectly happy to have the cast-off of some upgrader for this system.
-- Mein Systemadminstrator hat einen großen schwarzen Moustache.
If Microsoft really thinks that this will be an average system in two years then I doubt we will ever actually see Longhorn. Microsoft will be finished by then.
The vast majority of people today are more than happy with their computer systems as they are, and a significant number of people have too much machine for what they're doing. For many years into the future you will be seeing people with P3 and P4 machines still doing then what they do now.
There's a reason why processor sales are slipping for Intel, and it has little to do with AMD: no one's upgrading because the last upgrade they did made no real improvement. How much faster can you get a program to start? How much faster can you do what you already do (excluding those who are in scientific or graphics fields).
Hardware speed and power has accelerated so quickly up until now because software development could keep up with it. Now that proprietary software has stagnated (the last two software packages released by Microsoft, Corel, Macromedia and Adobe are exactly the same with one or two completely useless features thrown in and a new splash screen and icons) there is no reason to increase the capabilities of the hardware. Nothing you can do to a word processor will require more processing power than a current "average" machine offers. Same with web browsers and email clients. Even games -- game development has slowed to a crawl because it takes so long to make them now. Then there's the fact that game graphics can't get that much more realistic (and really, they don't need to be -- the Doom 3 demo already makes my stomach turn).
The described system will not be anywhere near "average" for the "average" computer user in two years. Bookmark this post and flame me in 2006 if I'm wrong.
-JemJim Allchin showed Longhorn playing six high-resolution videos at the same time, while playing Quake III in the background.
XP on equivalent hardware barely sputtered out four of the videos. Longhorn is definitely a media OS.
I'm looking forward to this new 3D infrastructure display technology.
I worked at Microsoft on all kinds of projects from Win95 Osr2 to WinXP and then Longhorn.
Longhorn is going to be BIG, Gaudy (ugly) and have far too much in the way of trimmings...
Believe it, you will have to have a HUGE video card just to keep up with all the "pulsing and fading" crap...
I do not doubt any of what is reported. It will be able to be "turned off" kind of like WinXP can be "reverted" back to Windows Classic...but what a PITA...
My history of PC clone purchases (I tend to buy a near top of the line machine every 2 years):
1995: 133 Mhz
1998: 400 Mhz (300% faster)
2000: 1500 Mhz (333% faster)
2002: 2800 Mhz (90% faster)
2004: 3400 Mhz (20% faster)
If the present trend that I've observed continues, however, we won't see 6Ghz in 2006.
However, CPU clock speed is only one factor as far as system performance goes, hence Intel's recent announcment about moving away from marketing Pentiums based on clock speed. So maybe we'll see a P5 "7500+" rated CPU...
I'm not so sure. I think hardware makers are going to find it hard to sell this stuff when there's no demand for it.
You definitely seem to be abnormal with your 1.25 GB of RAM; most people I think still have 128MB - 512 MB. I'm doing just fine with 512.
Even with Windows XP, most people have no use for more than 40 GB of disk space, if that. The biggest thing driving disk space demand right now is people wanting to store all their music as MP3s, or downloading a lot of movies online. I only do the music part (with my own CDs), and my 80 GB is still far from full. People who don't do music and movies (such as office workers) have no use for large hard drives. I really don't see how Longhorn could use so much disk space either, unless they're loading it down with useless video clips for some reason. Even MS couldn't write code that bloated, even with the hidden flight simulators.
Intel is already having problems with selling their processors because users are finally figuring out that you don't need 3 GHz to read email and surf the web. Intel's even sponsored video gaming competitions in Vietnam in an attempt to drive demand for faster processors.
All in all, while some home users (mainly gamers) will want equipment with these kinds of performance specs, businesses aren't going to like the idea of having to upgrade so much hardware just because of an operating system upgrade.
That pretty much what my dad said about the 30MB hard drive we got with our 286 computer in 1991. He wanted a 20MB hard drive because he couldn't imagine filling up 30MB. It eventually was filled up, all right!
By 2010 or so I'm sure there will be plenty of ways to fill up a 1TB hard drive, some of which haven't even been conceived of yet. I certainly never would have thought 6 years ago that I could fill up a 40GB hard drive. Storing large collections of movies, music, and images, and bloated Microsoft programs simply wasn't done that long ago. Well, we complained about Microsoft programs and OS's being bloated back then too, but few imagined they would get much, much bigger.
Yeah, I'm optimistic too. Our friend Mr. Gates has said that in the future, hardware will be free or almost free. I'm wondering how or in what world that is reasonable. You'd have to sell a lot of advertising built into the OS to pay for a dual core 6ghz machine.
Clock speed goodness has slowed down lately. About the only thing that's still going nuts is hard drive space, or so it seeems to me.
Why do I have this? I don't smoke.
My "fastest" system here is 900MHZ AMD and runner-up is a 2x300MHZ UltraSPARC. They are my *servers*, my clients are like 500MHZ or less. All my computers, Mac/x86/Sun, all run a flavor of Unix.
My point is, my computers are doing evrything *I* want them to do, and more. Uptime and stability are great. Will they beat those "been up for 10 years" VMS uptime records? Probably not, but whose keeping track? =)
If this is the path Microsoft wants to take with reference-hardware, which clearly is not the case according to that MS-SpokesHole, I'd think that this would be another great selling/advocacy point for OSS and just nix in general.
iGZo
The 6 GHz is a little fishy to me, and here's why:
:-P
6 GHz --> 0.17 ns per cycle. Light travels 5 cm (about two inches) in 0.17 ns, and information cannot travel faster than light. This means that even at the speed of light (electrical signals in typical electronics propogate at ~0.8 c, IIRC) it will take almost the entire clock cycle to get information across the chip, never mind whatever time it takes the transistors to respond.
In the meantime, those nursing dreams of 100 GHz chips had better look beyond nanotech to picotech-- atom-sized transistors.
The problem I see is heat. Intel's latest chip design, the Prescott, puts out ~80 watts of heat at 3.4GHz. A dual-core, 4GHz version would put out around 150 watts. No air cooling system in the world can handle that sort of heat density.
Now, look at graphics cards. Triple the video power, and you can expect to double the heat output -- if the process shrink to 90nm reduces the power output. If, instead, they run into the problem Intel did, the heat output will increase five-fold. There's enough headroom on GPU cooling that you can still air-cool, but these really will be the "vacuum cleaners" that recent nVidia cards were accused of being.
GigE and terabyte storage are reasonable expectations.
"They redundantly repeated themselves over and over again incessantly without end ad infinitum" -- ibid.
The article only specifies the specs for the Longhorn client machines.
Makes me wonder about the specs for the Longhorn servers.....
That's what I've been reading anyway. This is partly why it keeps being delayed, .NET isn't exactly mature and even if it was, there's a lot of old code that's being rewritten for this.
You are not alone. This is not normal. None of this is normal.
Such a system would allow clocks that run faster on larger chips than allowed by speed-of-light calculations.
How is this insightful? this is for the freakin' OS, not for gaming or anything!!! Remember, the average consumer does little more than web+mail+maybe movies/music. 3x faster video cards indeed!
... but these requirements for the OS are ridiculous. Besides, this is not going to be the average system very soon, as the 'average system' is still sold to businesses - and good luck trying to convince those they have to shell out so much money for useless hardware (3d? loads of ram for the secretary's freecell?) just to upgrade the OS! Heck, good luch trying to get a system to this spec from Dell for less than $1000! And if Dell won't sell it ...
yes, hardware will improve, there will be faster CPUs, GPUs, faster and cheaper memory
Also, if this spec turns out true, there will be a lot of noise from all the people who bought the last MS license plan - and it won't be cheering, either!
The only good news is MS will lose a lot of corporate/gov customers with this spec. Maybe Longhorn is not such a threat to opensource as previously thought?
If I said something embarrassing I would want to deny it too.
Problem is it's not his responsibility to deny he said it; it's your (or whoever's accusing him's) responsibility to prove he did. Anybody can just accuse anybody else of saying anything; doesn't mean they did. Show me the proof. And the fact that a bunch of Slashdotters think he said it is not proof, so don't pass it off as such.
Nobody has ever come up with an original cite for this alleged quote, in all the times it's gone around the net. See here for Gates' own response, including his own call for a citation that he knows doesn't exist (and if it did, he'd finally be able to disprove this silly quote once and for all by digging up the original article cited and showing the world that the quote is not in it).
As Gates himself admits, he's said plenty of real stupid and dumb things, so I don't see why he'd choose to deny this particular quote and none of the others if he's lying about it.
This is actually a very common technique in the commercial software world, advocated at least as early as 95 by Alan Cooper in The Essentials of User Interface Design: look at your project schedule, try to project what kind of hardware will be common by the time you ship, and plan for it. It's not rocket science, just common sense. And as others have pointed out, the specs they are targeting should be standard by 2006, let alone by 2008 when the beta program will end.
BTW, as an official Longhorn beta tester, I can confirm that this story is not a hoax: I was given these specs over a year ago at some of the early beta launch meetings, and while they've bumped the RAM up from 1 GB to 2, nothing else has changed.
BTW2, at WinHEC this week the graphics vendors are complaining that Longhorn won't be using enough of the vast amounts of GPU power they will be providing by 2006...
Delivering militantly anti-commercial music to all two people who care!
However, 1TB >> 120G that is standard now, and 120GB is effectively infinite for most people. When disks were ~40MB, every time I got a bigger drive, it became filled in about 2 months. It now takes me longer than three years, which is about my upgrade cycle, so it is "infinite". 1TB would represent the largest growth factor from current standard systems.
You can fit a 2,000 page autobiography for ever man women and child who ever lived on a single terribyte disk.
Now tell how a simple os whose job is only to act as a layer between apps and the hardware fit in?
What the hell is in there? Seriously?
http://saveie6.com/
I call bullshit or troll. First of all, 50 open tabs for days is so unbelievably illogical that I can't even begin to imagine what you're doing. I can't exactly say I'm going to hold it against the Firefox developers that their browser becomes a memory hog when people are using it waaayyyyy beyond its intent. If that's something you legitimately need, offer a patch or use a tool that's actually meant to do that. Otherwise, don't complain that it's not doing things it's not supposed to.
Second, the trite old "it loads slower than IE" is so incredibly irritating that I have to bite my tongue to prevent a slew of obscenities. Boo hoo. So, you have to wait an extra 2 seconds for it to load up because the WEB BROWSER isn't tied to the KERNEL. After all, what sort of moronic dipshits would make a web browser an integral part of a system kernel anyway?
Finally, I call bullshit on the "slow loads". If you've got benchmarks, show 'em. Otherwise, my anectdotal evidence says your anecdotal evidence is full of crap because the only lag I see on my 1.5/Cable connection is from the servers on the other end of the pipe.
Alito: A vote for Alito is a punch in the eye to put that bitch back in her place!
I also read somewhere that the multi-core opteron will be pin compatible with existing opteron processors, so you will be able to drop them into place on existing boards due to the miracle of hypertransport (the REAL HT)
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
Well, they don't test it. When they release the product to the market they let the people test it and then (about a year later) the provide the 1st Service Pack.
So, longhorn is to be released at 2006, with at least one year of delay and another one until the first SP, we're talking about 2008. By that time these hardware requirements will propably be obsolete.
I don't think the estimates are the minimum spec to run the OS, the spec is to run the applications that MS expects will exist by the release of the OS.
When you think about it, longhorn will take AT LEAST a few more years before a near "complete" version will be released for testers and/or testers.
For the video card part, it's somewhat plausible considering new video cards come out every half a year to a year and prices of precedent generation cards go down pretty fast.
As for storage, that's pure bullshit... I stil know people who have a hard time filling up their 20GB (without pr0n). the AVERAGE user won't know what to do with that space (unless he's told to download like a freak)
2 cpu's? Right so I'm gonna use a windows operating system for a multi-cpu system when linux handles smp way beter AFAIK. People won't get a multicpu system to use word/excel and use email.
2GB of ram? and my friend's name is richard simmons. Are they saying that based on their current longhorn versions running in DEBUG-MODE? 2GB could be a mainstream for gamers or developpers but I doubt it will be for the average joe.
1gbit ethernet. oh isn't that nice. Microsoft are predicting the evolution of home networks with the transition from 100->1000. Unless it's a house with crazy exchange of pr0n, then I don't see the use of 1gbit lan (/. talked about transition to gigabit lan I think). Even if there's no home network established, is this a hint given to us that says "our future residential service offered by isp's will offer blazing speed?"
wireless? Can't say much for this one really. Have friends who need wireless, some who totally don't care. Can't really say if later, wireless would be introduced in products other than laptops and pocket pc's
Basically, what I think on this article? nice way to tell pc vendors "sell monster pc's for the people who would want to play 3d minesweeper with AA/AF)
giddy up!
IBM went the other way with OS2 (1994?). I seem to remember salesmen trying to persuade people to buy OS2 machines with 16Mb. OS2 needed 32Mb to run properly.
Everyone just ended up complaining about the OS2 performance which was a shame, since it had a multiprocessing angle which Windows didn't have at the time.
So maybe we should applaud the efforts of developers to be realistic.
[Longhorn will require] a dual-core CPU running at 4 to 6GHz; a minimum of 2 gigs of RAM; up to a terabyte of storage; a 1 Gbit, built-in, Ethernet-wired port and an 802.11g wireless link; and a graphics processor that runs three times faster than those on the market today.
:)
Now, if that's just the OS... well, let's take the minimum requirements for XP.
* 233 MHz minimum required (single or dual processor system);* Intel Pentium/Celeron family, or AMD K6/Athlon/Duron family, or compatible processor recommended
* 64 MB minimum supported; may limit performance and some features
* 1.5 gigabytes (GB) of available hard disk space
* Super VGA (800 × 600) or higher-resolution video adapter and monitor
Now compare that with the requirements for a modern game... I'll use Unreal Tournament 2004 as an example:
PIII 1000
128MB RAM
3.5GB HD
64MB Video Card
Alright... Comparing XP and UT we get:
233:1000 processor speed... (~1:4)
64mb:128mb memory (Which is stretching it, you tried playing UT2004 on 128mb RAM? Slow as hell here.)... (1:2)
1.5gb:3.5gb hard drive space... (~1:2)
SVGA:64mb vid card... Tricky. Let's say (1:5)
After looking at longhorn's reqs, we are left with the conclusion that games of longhorn's time will require...
(4-6ghz*4)16-24ghz... yeah, sure, that'll happen if 4 years...
(2gb*2)4gb ram... that's MODERATLY reasonable, at least in comparison to the processor speed...
(1tb*2)2 terabytes of storage... right.
(3x*5)And a video card roughly 15 times what we have now. Not a chance in hell.
So... who's up for some pong?
You know you're a nerd when you can mathematically prove that you have no life.
A computer of this stature would have an incredible demand in my field. I work with Computational Fluid Dynamics (CFD) software, specifically, microfluidics. Computing the solutions to atrocious systems of partial differential equations is a bear of a problem, even for a cluster. A 50-microsecond simulation takes about 24 hours to compute on my 1.5GHz, 512MB RAM workstation.
I certainly have a demand for this kind of computation power, and then some. If I had my way, I'd be working on a 100+ node cluster, but unfortunately thats cost prohibitive.
We're overlooking that Microsoft has always targeted the average PC. But methinks that Microsoft is selling more OSes on new machines than upgrades.
So perhaps the strategy is to give us "tomorrow's OS on tomorrow's hardware" and really take advantage of it?
They need all that storage and horsepower to power the DRM crypto that will keep you from running "unapproved" OS's on the next-gen hardware.
No foolin. These systems are gonna run hot.
Now, consider that a EPIA 5000 Mini ITX board runs so cool, that it doesn't need a fan. Sure, its a "slow" box (500-600mhz), but bear with me for a sec. Now consider what the average user does with their PC. I just don't get what the average consumer is going to need to do with this horsepower. I've been saying that for years though. Most non-techies I know send mail, browse the web, write letters. Any of those $200 PCs at Wal*Mart can do this.
I see a lot of stuff all the time, Moore's law and what not. But has anybody ever studied WHERE consumer PC's are going and WHY? Won't they reach a point where they are "good enough"? Like a small Briggs and Stratton engine for a mower. Sure, they change a little over the years, but not a whole lot. Because they work and server their purpose.
Now, faster computers obviously have tremendous, unimaginable uses outside the consumer market. But my point is at some point the consumer won't need the same type of hardware that a scientist is using to sequence DNA or an IT shop is using to run web sites, process transactions, etc.
Frequency itself is only one contributor to power. They will likely continue scaling processor voltage down too (power goes with square of voltage). Wider parallel busses at lower freq? My imagination is limited, but these are just engineering problems that will be solved as always.
I've got a starting set of 9 tabs and as i read through articles and posts i'll open referenced pages in new tabs. I'll then go through the tabs later and read the referenced material. Some of those tabs i'll close when i'm done reading, others i'll keep open so i can show them to my girlfriend later in the day.
When i'm researching something i'll often keep several windows open on the subject at once. I currently have seven tabs open on pages about Venus and the effects of a planet's tilt on seasons/climate as reference to an idea for a science fiction novel someone is working on.
Got another three or four tabs open to statistics and a message board for an online game i play. I usually check in on it ever three or four hours, so i tend to just leave them open.
Once you start to multitask it all really adds up quick.
As for opening time, after first rebooting my computer Netscape usually takes between 10 and 30 seconds to open, depending on how grumpy my laptop is feeling. Of course part of that is because i told it not to pre-load its components. IE opens in about five seconds. Once i've used them both the both re-load in a second or two. Or are you talking about the individual loading of pages? The only serious problem i've noticed is that sometimes i need to load a page twice because the first time it will time out, but on the second attempt it loads right away. I have no idea if that's a problem with Netscape or a problem with my ISP however.
This Space Intentionally Left Blank
An OS manufacturer suggesting such a hard drive storage requirement would suggest in turn that not only the OS is going to grow significantly, but also all the other applications that MS want you to isntall are also going to grow significantly. Presumably the terabyte storage is so that they can still say "our stuff is only using 1% of your diskspace". But anyway. People keep going on about the OS requirements as if it is such a terrible thing that the OS would need so much hardware engine to make it go. My question is "Why not?". I have my PC here and the only thing that even begins to drive it is the latest 3D shooter. I am going to have all that power sitting there for that reason, so why shouldn't the OS utilise it? I for one am sick of these 2D windows with slow redraw times. This is my working environment. I want it to look good. I want it to give at least as good a visual impression as a game. Now I don't mean that I want to run down a corridor to find my finance spreadsheet, but some flexibility in the front end would be such a boon. Whatever the validity of that mostly suspect spec - it is just a prediction at best - I would like to see a breed of OSs that allow you to commit the resources you have available to you toward the process of getting stuff done, rather than have them sitting idle waiting for the next time you pull up Doom III. koan
This signature intentionally left blank
"No air cooling system in the world can handle that sort of heat density."
Rubbish. To keep a processor at a reasonable working temperature (lets say 45 degrees Centigrade), assuming an ambient temperature of 25 degrees C, you would need a heatsink rated better than 0.2 degrees/Watt (without fan cooling). Such heatsinks are common, are used in audio amplifiers and large linear power supplies, and can be bought for less than $25 from good electronic stores.
That's not to say that a heatsink that size isn't a problem for a neat tower case, but dissipating that kind of heat is not unheard of, and certainly not impossible. However, without improvements in the chip technologies the days of processors being a seperate module to the mainboard may be numbered.
Actually while you were sleeping DDR2 and DDR3 came out.
Also much of the RAM manufacturing was moved back to the west from Asia as prices there were being fixed.
After several lawsuits and DDR2 being scrapped as power hungry and inefficient DDR3 was drawn up with the major advantages being scalability to 2.x gighz (DDR) and a price of production several orders of magnitude cheaper than existing DDR.
there are some interesting articles over on CNET about the big box computer manufacturers fighting with memory makers.
Well, I have 410gb disk and 3.5GB RAM on my dual 2ghz PowerMac G5, which sounds like it might just scrape past the minimium Longhorn requirements if it was the PC equivalent.
But I spent about $4,000 making it that way, which hardly sounds realistic even in 2008.
Why did I spend $4,000 on a computer? Because I do CPU-sapping Final Cut Pro and After Effects work. Nobody spends this kind of money just to run the newest operating system. Well, not unless they're insanely wealthy anyway.
I think these requirements will be a tough sell even in 2008. I wonder if the hardware requirements are a major reason for the delays, since there's no way consumers today would accept anything near them.
D
One of the major reasons we are moving towards a transition to linux from windows is:
a) XP is expensive, even by volume licensing an organization with 1000+ machines is a costly thing to licence
b) Most of our machines won't run XP. They won't run win2k very well
c) Upgrading/replacing all our machines to run a new OS is more expensive than the OS. Moreover, with the MS track record, by the time it was done there would be a new OS.
Cue in Longhorn, I think this will be even moreso. It's not just the cost of the OS businesses can't afford, it's the hardware required to run the damn thing... not to mention the dependability/security issues. If not for our linux servers offering protection from the outside world, we'd be sasser'ed nicely too if we ran a lot of winXP machines.
Actually, I'm fairly certain that he did say it. I know it was reported in the popular press of the time, AND considered a reasonable statement by most people, including most technical people.
Please remember that 640KB of RAM was more than the typical IBM 360 (370?) of the time had. And that the Apple ][, which IBM was attempting to replace, only went up to 64KB (and that required using bank switching).
Still, I'm not surprised that he denies it now. Now it sounds silly. As silly as the IBM chairman's forecast that there might be a market for (I want to say 5, but all I really remember for certain is that it was less than 20) computers in the country. Of course, he said it back in the 1940's, and he didn't want to consider any competitors. (At the time that he said it there were already more than 20 computers active...mostly, admitedly, in university EE departments.)
I think we've pushed this "anyone can grow up to be president" thing too far.
A 200-stage pipeline will only realize a performance gain on instructions that take 200 cycles to execute. The bulk of instructions that a CPU executes tend to end up being pushing words around. Even on a bloated Pentium, that does not take 200 cycles.
Today's Sesame Street was brought to you by the number e.
Ok, but to this I say - "Apple Mac".
Seriously, if there was ever a strong argument for buying today's PowerMac system, it's got to be for video editing.
The G5 systems support up to 8GB of RAM, and it's not at all uncommon to find people configuring them with at least 2GB - 4GB right now. (Because quite frankly, it's not really that costly to do so using 512MB PC3200 DIMMs. They have 8 slots on their board.)
I've done video editing from a DV camcorder on my Pentium 4, and believe me, I get *much* more accomplished without crashes and hassles using Final Cut Express or even iMovie with some 3rd. party plug-ins. Don't forget, Steve Jobs owns Pixar, along with being Apple's C.E.O. That means he's VERY attuned to the needs of movie producers and editors. His systems practically revolve around it. So I'm only concerned with what Apple does, when it comes to a need for more CPU, RAM or disk space + video editing, and I suspect I'm in the majority in that particular niche of the market.
Well, I'm sure I've got out of bed and posted this thought way to late for any good responses... however...
It would be interesting to look at global market boom-bust cycles and use these to evaluate when MS is going to release longhorn and its' next armarda of operating systems for maximum profitability. I believe we are starting to move into the next stockmarket boom period, here in Australia anyway. The rest of the western world can't be all that far away from us. As the equity markets boom and money flows into them, productivity needs to increase and ms-windows is an excellent short cut to helping increase productivity, forget linux for the moment, linux is a longterm investment. Windows-longhorn and its decendents will create more profit in a boyant marketplace. 2008 might be a good time to start pushing/selling longhorn when the equity market starts to get ahead... similar things happened with NT4 release dates IIRC and if my theory is tracking along sane lines...
Any thoughts? or did I really sleep in and miss the boat...
Does it go on forever?
I am not trying to troll here, but when someone says a recommended system, then they mean it will run semi-fast. That usually means a 486 with Windows 3.1 type of fast.
I must say, that if MS rolls out their search engine on this machine with this software (the specs are probably not right), then they will need like a million servers for the search engine. Google will just need like 1000 with these specs!
I would be seriously interested in hearing what other people would use a 64-bit 6GHz processor with a terabyte harddisk and gigabyte of RAM for?
I could see having the drive size for storage of 10000 CDs or a few hundred of your favorite movies. I asked the same question about five years ago about a 10 Gig hard disk and the most common reply was to store all your music recordings on your PC. That's what I'm doing now.
I would like to see high quality language translation cheaply available. Language translation seems to have five levels. Level one is a word by word dictionary look-up. Level two is phrase translation inside sentences. Level three would translate whole sentences and compare them to other sentences in the paragraph. Level four would catch most idioms and ensure that the paragraphs made sense in the destination language and level five would be equalivent to a modern professional translation.
This is just my WAG on the subject. But it seems that the web translators like SysTranCom and Babelfish are working on level two. I wonder if a 6 GigHz CPU and 1 GigRAM box would be able to do OCR on Arabic and also translate to English. I would think that Arabic to be the hardest language to do Optical Char Recognition on because the syntatic elements are linked together.
I wonder if 6 GigCPU with 1GigRAM would be able to do speech-to-text better than today's Dragon Systems and IBM. A $50 hand held box that does level four translation from speech in one language to synthetic speech in a second language would be a great goal to hope for. But I don't think these devices will be around for another 15 years, at least at $50 US.
Another wish-for would be audio remixing of commercial music. Hate that stupid guitar solo or dumb background vocal? Then just phase-lock onto it and remove it.
How about a comment compilier? Toss the source and do linguistic analysis on the code's comments. Then have the comment compilier create the source according to what the designer wants.
If it's not right, then do another interation until it gets closer. C language is so primitive: it's a legacy from the days when RAM was tiny little metal beads woven into a grid that doubled as a spaghetti strainer and CPUs acted as room heaters.
What are your thoughts? What would you do with a 6GigHz CPU, a gig or two of RAM, and a terabyte or two of storage?
Let me guess....
Ultra Porn
and Games
So, even though these are the specs for the "average" computer, it's possible to have it today. And bottom line, if it can be done today, then there is no reason to think it wouldn't be average in 2.5-3.0 years.
But to me, 'average' computer specs implies that I can have all of this for $2000 or less, including a decent monitor. When that happens for these specs I'll stop laughing. Oh it's inevitable sure, but for now it's rediculous.
Sure. After all, 3 GHz --> 4 inches which is certainly smaller than your motherboard, which is one reason, I suppose, the bus runs at a fraction of the processor clock.
I think the trend in CPUs is going to change dramatically soon, much the same way it has with video cards. Except, in a slightly different direction. Instead of getting faster, the direction they need to be going is getting less expensive to produce, less power requirements, and less heat output. ...and more of them.
I wouldn't be suprised if 2-4 processor machines were not the normal in 3 years time.
Why keep going up, when you can go sideways and gain as much ground?
Free Gamer - Free games list and commentary
What the hell machines are they testing alpha builds of Longhorn on anyway? Are they installing it on a cluster and considering that "tomorrow's platform"? No Beowulf jokes... the 640 thing is enough "classic" Slashdot already.
US Democracy:The best person for the job (among These pre-selected choices...)
Everybody has 3 or 4 machines already and a game box. We simply don't need a 6 ghz processor. We certainly don't need another bloated M$ product to surf the web. We (I believe) will soundly reject this upcoming drm and new word/excel format. This cycle needs to stop, and will.
These companies make this stuff because that's what they do. The ultimate proof will be when the consumers actually buy this stuff or not. There have been many "great ideas" that the unwashed masses have already rejected. Anybody remember "PUSH"?
Microsoft also backtracked this year on their intention to end support for win98. Guess they checked and found that 28% of the web was still using win98... probably with no intention to upgrade. Our dollars will decide where the computer industry goes. There is no new Internet to drive sales so I can't really see it getting stronger. BTW, here in Canada, an AMD 2400+ with most goodies is about $475 American.
I'm not anti-social, I'm anti-idiot.
I thought katz was a fine writer, but clearly some folks didnt care what he had to say, only that they where offended he said it eloquently.
I thought Jon Katz was a fine writer in print, but the problem is he just stopped following any sort of journalistic standards when he started to write for slashdot.
Whatever happened to optimizing your code? Are programmers getting lazier, or are we just trying to force everyone (including those with virtually non-existant budgets) to buy the newest PC ever few years?
....but I think anything that important (I'm talking an operating system with lots and lots of...stuff...) should be optimized so that it will run under much lower specs.
I fall in the lazier category because I do not write programs for people other than myself (normally anyway).
If you think I'm wrong, then please...flame away, for this is just my opinion.
Bah. I edit Wikipedia, and if that's not tab-intensive, I don't know what is. :) Who wants to go to RC patrolling? :) there's your 50 tabs right there, plus a few to check potential copyvios, see the talk pages, drop {{subst:test}} on all the newbie let's-test edits, notice page histories, keep an eye on your boilerplate and on Votes for Deletion... :)
The World Wide Web is dying. Soon, we shall have only the Internet.
pfft! "unless you want to look cool or got it for free"
Well hhmm, my 15,000 RPM U160 drive I bought brand new for a whopping $100 bucks off of pricewatch defrag's in 5 minutes, loads games in 1/10th the time the basicly same box next to me does with SATA, has a 5 year warranty and an MTBF 10x greater than your IDE/SATA drive. Also it copies a gigabyte from one place to another (even in windows) in around 5-6 seconds. No, no advantage to U160 at all, nah.
Oh and on a non competitive note, I would use SATA if they came in 15K RPM flavors. RPM == speed.
I can imagine that the system when in use for some time for video editing or something simmillar may need 1TB. But when you install a bare OS - is that what you need 1TB for? And then you start loading the applications? What is the expected ratio of OS/Apps?
I agree. I think there's going to come a point where users simply don't need any more computing power. The fastest machine I own is 1.7GHz, and I imagine that will satisfy my needs for a few years to come -- I use 3D modelling tools, compilers, and a few other "intensive" applications, but the machine still operates at a reasonable speed.
Why would I *want* a bloated operating system that needs all those resources? Surely if I need that processing power, I would be better using a lightweight minimalist OS to squeeze every bit of power out of the system.
Also, with Microsoft's Secure Computing initiative, and all that crap, why would I choose this Operating System over an open, free alternative?
Finally, I think the world will be pretty much converted to Unix alternatives (Linux, etc.) by the time 2006 rolls around. During that time, we'll all be enjoying constant updates and improvements, the Linux kernel will have gone through many versions, and Linux (and BSD, etc.) will almost certainly be "READY FOR THE DESKTOP". And hopefully, Linux & co. will have dominance as the gaming and software platform...
Microsoft is obviosly attempting it's usual trick of tieing everything into the kernel, and building an all-in-one "user-friendly" solution...i.e., eliminating choice.
I would like to see high quality language translation cheaply available.
So would we all, but it's not gonna happen.
Why not? Because even if you manage to come up with a way to make a computer handle context and nuance, a good translation often relies on the imagination and writing skill of the translator.
Look, forty-odd years of top-level academic research has failed so far to come up with a good theoretical approach to machine translation. That is to say, you could provide the best machine translation experts with a cluster of 6 exahertz machines, with petabytes of RAM, and they still wouldn't be able to produce anything that sounded remotely like a human translation.
In the 1940s they thought it would be there by 1960. In the 1960s they thought the computers of 1980 would be able to do it. In the 1980s they thought we'd have it by 2000. Here we are in 2004, and we still don't have it. And only a fool would expect it to be there in 2008.
Babelfish as it stands is laughable - it's still at level 1 in your categorisation. Feed it "My Japanese is bad", translate from English to Japanese. The translation is fine - except that the word it uses for "Japanese" means "an inhabitant of Japan" rather than "the Japanese langauge". Nice one.
this whole thread is a waste of time ... everybody is laughing at MS just because is cool to do so on /.
when longhorn will appear in 2006(optimistically speaking) this would be a LESS_THAN_AVERAGE computer
let's see:
dual-core CPU - average to high end market TODAY .. in 2006 this will be way under-average stuff
CPU running at 4 to 6GHz - today's high-end is 3,4ghz ... 4ghz is supposed to come by the end of THIS year... by 2006 this will be less than average
a minimum of 2 gigs of RAM; average new computers of today come with 512MB ... it's quite sure by 2006 2GB will be the average
a terabyte of storage; high-end today is 300GB ... and the storage doubles(or more) every year => 1TB could easily be average in 2006
1 Gbit, built-in, Ethernet-wired port ; that's high-end today ... by 2006 it will surely be low-end
802.11g wireless link; same as above
a graphics processor that runs three times faster than those on the market today ; the new generation of graphic chips, schedulled for this summer/fall, already runs with almost double speed compared with the today high end ... 3X by 2006 will be less than average!
"There is nothing more frightful than ignorance in action." Johann Wolfgang von Goethe
> How does Microsoft Intend to Survive?
Simple. DRM in BIOSes at the hardware level. Attacks on Linux via SCO etc at the OS level. FUD, loathing, and lock-in at the applications level. Patents, DRM, EULAs and DMCA at the legal level.
Remember the hidden APIs in Windows 3.x? They'll be at it again. Even better, Microsoft could put in "Trusted Computing safeguards" so they can Trust that only Microsoft's applications suite, IDE, etc will run. Bypass these safeguards, and it's charges under the DMCA and 20 years in max security prison as an evil godless communist hippie software pirate terrorist hacker for you, buddy!
Oh, and meanwhile they'll sue you for breaking the clause buried in the Longhorn EULA where you agree to only install Microsoft applications. Good luck in fighting off their army of rabid jackals with law degrees.
This isn't insightful. This is paranoid ranting with no connection to reality.
Microsoft is a monopoly and aims to continue as a monopoly. On this we agree. But there is not a snowball's chance in hell that the court or legislature (no matter how corrupted by money) would allow a monopoly to enforce that computers run only Microsoft software by means of an EULA, the DMCA or shotguns loaded with salt peter. And not even Microsoft is so ignorant and arrogant enough to try.
Microsoft's real life abuses are egregious enough to piss off most folk if we just point them out. We don't have to invent fantasies about the machinations of Evil Bill and his minions.
Phiwum's law: anyone that names an obvious law after himself and then puts it in his own sig is just pathetic.
XP ruined MS and Longhorn will be worse. Bloated. Completely useless garbage code.
Since XP I've moved on to a slackware desktop main workstation. Never looking back.
I was writing a review of Office 97, which shipped on something like 5 CDs, totalled more than a gig when installed. I asked the PR flak if that wasn't a little bloated and he said, "We expect in a few years it will be very common for PCs to have multi-gig hard drives." To which I responded: "So Office 97 is designed for PC 99." He refused to respond ;)
Bark less. Wag more.
And it comes with all the free spyware you could ask for. IE sucks, it's a bug ridden virus vector. Wrapping it in a shell that adds some of the functionality of a real browser doesn't change that (though I do carry a copy of crazybrowser on my USB memory key for when I can't install a real browser on a clients system). When people see me browsing without any flashing ad's or popups the experience is so jolting that they invariably ask me what I use, I tell them that I use Mozilla with a couple of addons to make my browsing experience much better (I have a click for flash plugin, I've turned on popup blocking, turned animated gif's to once, etc). This almost always results in the person asking trying out Mozilla.
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
I read that the [in]famous quote you mention was actually a misquote but regardless, the "computer" that got us to the moon was not a full-blown media center/gaming device/archive/web browser/etc.
That's because computing power necessary is inversely related to the operator of the device. Microsoft has realized that Joe Sixpack is no rocket scientist.
--laz
"Just remember, it takes a village idiot." -- The Motley Fool.
They won't. They'll say "Fuck it. What we got works".
Not Buzzword 2.0 compliant. Please speak english.
Check here:
http://www.snopes.com/radiotv/tv/newlywed.
for details on the "Up the butt, Bob" story.
Recently a Newlywed Game clip (from a Game Show Network rebroadcast of the show) has come to light
So, it DID happen. Despite the YEARS of denials, despite the $10,000 cash reward Eubanks offered (HAs Gates offed a reward? He must be less sure then Eubanks, and Eubanks was WRONG!), it DID happen!
So, denials mean nothing.
Well, I doubt Mitnick did too, but we used to whistle the initial connection sequence into the phones at the University of Delaware in the 1970s.
With the 300 baud acoustic-coupled modems you could do it, and once you got the sequence down (turn on hard-copy terminal, pick up phone, dial number, whistle, slam phone into rubber biscuits, wait for login prompt) it was actually faster than the way we were supposed to initiate connections.
When they converted to 1200 baud we couldn't do it any more.