Sun To Build Opteron Servers

geekee writes "According to an article at CNET, Sun is planning on creating Opteron-based servers. These are expected to include 2-processor and 4-processor models running either Solaris or Linux. This move isn't surprising, given the performance and cost gaps between the Opteron and UltraSPARC processors. A move to Opteron would allow them to be more competitve in cost and focus more on what they're good at, designing systems, not processors."

Comments

another dell/HP

[stonebeat.org]

so Sun will become Dell or HP???

Re:another dell/HP

[the_2nd_coming]

last time I checked, DELL and HP did not have all the other Ancillary Business units that SUN has going. and comparing a Dell system to a SUN is just wrong headed. I am sure SUN will have al the best interconnects and such on a board that hey build for performance, where DELL just takes crap off the heap and slaps it together.

Re:another dell/HP

[the_2nd_coming]

my point is that Dell uses the stuff available to everyone and is no more innovative in hardware than any other PC maker, SUN adds value to their hardware. SUN makes the slow parts faster.

I guess you can summarize it like this:

Dell is reactive, SUN is proactive. ( in hardware technology, not business sense :-P)

Re:another dell/HP

[Anonymous Coward]

> comparing a Dell system to a SUN is just wrong headed.

Not so much in the 1 and 2 processor space. Sun's lower end has been your basic PCI Mobo grade stuff for many years now, save Sparc and Solaris. IDE drives, limited expansion, "the works" we've come to know and love from PC land.

Somewhat hard to tell, unless you look inside, because Sun controls Solaris and Solaris controls you.

As for the Sparc CPU, it is nothing particular to write home about, not bad, not great either.

Solaris has it's points, good and bad. Linux is clearly catching up, real fast, tho. Soon, if not already, they will be "exactly the same, only different" (So, maybe you're locked into Sun, accept it for what it is rather than preaching).

Sun has always been "overpriced", unless you needed some specific Solaris feature or a good dose of Corporate CYA. Remember, we're talking 4 CPU's now.

Now, IF your application set has a snowball's chance of growing into the 4+ CPU boxes, Sun or IBM are on the short list. You overpay for the little boxes just to make life, in general, easier.

Re:another dell/HP

[Pharmboy]

As much as I hate to admit it Dell is not bad for a x86 server...

I have a few Dell servers. All and all, I am pretty happy considering what I spent, but they still do not compare in quality and reliability to my IBM servers (pc325's from 1997, running 24/7, still running). Its the difference in 99% vs. 99.9%. It may only seem like .9%, but its all the difference in the world. My Dell's cost 1/4th the price of the IBMs, but that .9% reliability has paid for itself a few times over.

I will still use my Dells for non-critical web serving, routing, backup dns, etc. but the average Sun box would blow these lower/mid level IBMs away, and there really IS no comparison to Sun boxes. Apple is also much more reliable and robust than the average Dell. Same for HP, and others.

I like my Dells, but realistically, they are decent boxes thrown together from off the shelf desktop grade parts.

Re:another dell/HP

[bofkentucky]

ummm, hypertransport is hypertransport, the fast memory or PCI bus interconnects will be the same as a dell, hp, or homebrew box, and sun's already use (semi)-standard pci ethernet, scsi and video controllers, so you aren't going to see performance improvements there. If Sparc can't stack up against the opteron, so be it, let it die but 4 and 8 way boxes are not the be all and end all of computing.

Re:another dell/HP

[Doomdark]

so Sun will become Dell or HP???

I don't think article said anything about Sun dumping sparc-based line on short term. So just like IBM builds all kinds of system (from Power - chip based servers to PCs), perhaps Sun is just expanding their product line, taking advantage of cheapness of commodity x86 processor lines. Dell does not sell non-x86 systems, and HP only sells those systems for legacy purposes (HP-PA, Alpha).

Re:another dell/HP

[drsmithy]

Sun takes a generic Pricewatch Intel-manufactured rackmount system and "adds value" by putting a purple faceplate on it.

Actually, Sun takes a Dell-OEMed 1U rackmount system and sticks a purple faceplate on it.

Which makes the people saying Sun machines are so much better than Dell machines even funnier.

Re:another dell/HP

[IM6100]

Also, I doubt if the fact that Sun designs an x86 processor into one of their systems, that it means they're dragging in all the croft and architectural compromises that Dell does. Sun's box doesn't have to run NT or anything at all Microsoft to be successful. Dell's boxes almost certainly must.

Heck, I don't think a Sun 386i box, from back when Sun made their first try at selling an x86 box back in the early 90's, will run DOS or Windows or anything. It was a Unix box designed to run Sun's Unix software.

I don't know for sure, though, as I'm not lucky enough to have a Sun 386i box in my collection of Sun hardware.

Re:another dell/HP

[IM6100]

Somewhat hard to tell, unless you look inside, because Sun controls Solaris and Solaris controls you.

I just checked, and yes, NetBSD/sparc64 is supported on a lot of the architectures that you're talking about. (I had to check because I don't run any of that new 'icky' Sun hardware with PCI and IDE drives- my newest stuff is Ultra 1 boxes I got at auction for $12.50 each- still not a bad price to get an entry into 64 bit Unix...)

So where's this deep-dark-hidden stuff that you're fretting about? Download the NetBSD kernal source and dig around some.

Re:another dell/HP

[IM6100]

Wow.

And Dell orders the UltraSparc processor for said machines from whom?

My point is, you're talking about one low-end Sun box in a market segment where Sun doesn't even consider it worth their time to design the box.

Re:another dell/HP

[christophersaul]

Don't know what the plans will be with the Opteron machines, but the V60 and V65 are standard Intel boxes and as such happily run Windows.

Re:another dell/HP

[upside]

Yeah, bought a dual cpu Dell PE2600 this spring and the disk subsystem just blew up two days ago. Apparently a known issue. I'll only by HP/IBM/Sun from now on.

Re:another dell/HP

[Lumpy]

Amen brother, the IBM servers here we have are killer.

I have 2 of each. IBM, DELL, and Compaq.

The Compaq ML530's seem to be the most robust of the three though. Hotplug Ram,pci and SCSI drives no standard PC parts inside except for the video card and network card. everything else is build for stability,longivity and squeezing the last drop of power out of the 4 processors. and I still have 2 years of warrenty on the drives and whole servers.(standard 5 year warrenty on them)and you CANT beat Compaq's smartstart setup and install.

The Dell Servers on the other hand are nothing more than PC's inside. the features I have on the Compaq and IBM servers are not available for purchase on the DELL servers. the poweredge 6600 we got a month ago just feels cheaper.

I certianly dont have the same respect for the dell servers as the older one keeps having troubles, while my oldest IBM server is 7 years old and has not had a lick of trouble in that timeframe except for one DLT tape unspooling in the drive.

Re:another dell/HP

[Space cowboy]

Not sure I agree with you on that. The engineering inside the servers reminded me very much of an SGI or SUN box, when you open it up. There's a lot of thought gone into airflow, positioning, etc. It just "looks" reliable - certainly it's a custom motherboard, not just an off-the-shelf one. I don't mind them using off-the-shelf drives/processors/ram etc ;-)

I've got ~20 or so Dell rackmount boxen running websites, and only one has ever shown problems, with the gbit nic card (supplied by dell) failing after a year. This was an intel card though, so perhaps it's a bit harsh to unduly criticise Dell...

Simon.

Re:another dell/HP

[turgid]

so Sun will become Dell or HP???No way. Sun is still doing massive amounts of R&D. Sun still has SPARC going forward c.f. Dell who just ship commodity intel stuff and HP who are abandoning everything for itanic.

Sun is introducing these Opteron servers to compete at the low end.

Sun also develops its own OS - Solaris - which it is able to offer cheaper than Linux on the same hardware.

Because of Dell and HP's close alliance with intel, do not expect to see them shipping any Opteron (or any other AMD for that matter) processors in the near future.

Maybe in 2005 when AMD has taken the lion's share of the x86 market, we will see a much smaller intel, Yamhill processors and the end of the 32-bit intel architecture. itanic will have sunk.

Hey, what do I know? I'm just a mad AMD and Linux zealot. So you can take what I say with a very large pinch of salt.

Re:another dell/HP

[yomegaman]

If Sun only puts Opterons in their low-end servers and sticks with UltraSparc III in their high-end machines, they'll be in the odd position of having their low-end boxes blow away the high-end ones CPU-wise. That would be some strange market segmentation to say the least.

Re:another dell/HP

[Pharmboy]

I have had pretty good luck with two of my dual cpu 1400sc for what I expected, but the drives DO bother me, Fujis... so far no blow up in 2 years/2.5 years

Re:another dell/HP

[Pharmboy]

so perhaps it's a bit harsh to unduly criticise Dell...

Wasn't. Was comparing them to my IBM servers. As I said, they are fine for non-critical systems, and are good for the money, but my *experience* has shown that they are not as robust as IBM. This is based on years of using both brands, not only in the server room, but on the desktop as well. In the server room, they are adequate for many tasks, but there is NO comparison to IBM in quality. Its not bashing Dell, its just real world experience.

Rented boxes are not critical systems, not by IT standards, although uptime is important. I know because I rent a rack for offsite backups. Dells are designed to be CHEAPER than IBM, not as powerful/robust/redundant. For most people, this is adequate, and even for lots of my uses it is, just not all. My primary DNS server, for instance, is an old IBM pentium pro 200 (dual) box with IBM drives that I purchased new for around $4500 (no drives, no os, 1 cpu, 32mb ram). Over 40,000 hours on it and I still consider it more reliable than a new Dell. I consider primary DNS to be a critical system, TOO critical to use inexpensive parts, regardless of brand.

You speak of Intel NIC cards, but Intel is who makes many of the parts for Dell. They all use intel chips, chipsets, etc. That is part of the problem. IBM uses some Intel chips, but uses their own for server management. Personally, I'm just buying time until IBM finally releases their Quad 2.0ghz 970 server (970 = G5) for $3500 mid 04. That will replace both of my Dell servers, and two of my IBMs as well, and I will be rid of Intel chips completely, on the server side. Intel on the client side is just fine, although AMD gives more bang for the buck. Dell doesn't offer that option anymore tho.

Re:another dell/HP

[Space cowboy]

The comment

so perhaps it's a bit harsh to unduly criticise Dell...

... was meant to read that it would be unfair in my case to criticise Dell for a fault with an Intel NIC

I'm glad you've had good experience with IBM machines - personally I've never used them. I've used VAXen, DEC, HP snake, SGI and SUN boxes so I'm not sure why, it's just never happened. I can't see much difference in the construction between the Dell and the bigger boys is all I was saying. (until you start getting rack-size boxes. There's a level of difference between an Onyx2 and just about anything. No Dell machine I know of (though this is /., so I'm bound to be educated :-) runs off a 3-phase power feed)...

Rented systems are only not-critical if your business does not depend on them. Mine does to a fair degree. They are therefore critical to me.

I don't necessarily think there's anything systematically wrong with Intel hardware either - from your post we disagree on that too :-) but it's served its' time with me. I've been running unix (and VMS, previously) boxes for the last 15 years or so, so I have some experience in the matter too...

Simon.

Re:another dell/HP

[jo42]

Ummm... Ever look inside an Ultra 60? Or Sun E450? 64-bit PCI, d00d...

Re:another dell/HP

[flawed]

Don't forget that the Sun 386i was a dismal failure, though.

I certainly hope Sun will do better with their Opteron boxes, and will make them real Sun boxes. So far the x86-based Sun boxes are just OEM machines with a Sun logo slapped on it.

Re:another dell/HP

[turgid]

and sticks with UltraSparc III

UltraSPARC IV will be out RSN. Anyway, I doubt that a 2-way Opteron could beat a 72-way E15k with UltraSPARC III processors. In fact, I'm sure that my old 8-way E4500 (UltraSPARC II, 400MHz) from many years ago could keep pace with a 2-way Opteron box on multithreaded applications. Remember though, 5 years is an order of magnitude in the computer world.

Re:another dell/HP

[flawed]

We have boatloads of E450s. You can still do better with a 4 way Intel/AMD.

Well, nowadays you have to compare with a V440.

Wicked

[alpha713]

I'm looking forward to seeing it on the "shelves", any move that reduces the cost of higher end computers is fine by me.

Sun is simply adapting to survive

[RedHat_Linux_Man]

They know that linux is the future-- Sun is simply adapting to survive. Both it and Opteron are more cost-effective than UNIX and SPARC, respectively.

Re:Sun is simply adapting to survive

[kwerle]

Sun is simply adapting to survive

hahaha. Sorry, I just had to laugh.

Isn't this the same ship that SGI is sailing out on?

And by out, I mean off the edge of the world, into the abyss.

Re:Sun is simply adapting to survive

[turgid]

Isn't this the same ship that SGI is sailing out on?

No, that would be the good ship itanic.

Re:Sun is simply adapting to survive

[samhalliday]

Both it and Opteron are more cost-effective than UNIX and SPARC, respectively.

ok, opteron is cheaper than sparc (not as good, mind)... but, what do you mean by Solaris is cheaper than UNIX? surely Solaris IS one of the few registered true UNIX operating systems: a feat that GNU/Linux is not aiming towards. i'm not sure if the *BSDs are exluded for not hitting the standards compliance, or if they just didnt want to fork out the cash to "sit the exam".

Competition is great

[Zelet]

I know this is the wrong thread - but I am so happy to see healthy competition in the market place. Check out what is happening
- G5 vs. Opteron
- OS X vs. Windows
- Linux vs. Windows
- Mozilla/Firebird/Thunderbird vs. IE/Outlook

It is a good time for computing. Although, with Longhorn so far out (and no further IE improvements until then) I think the competition is going to be a little bit one sided.

Re:Competition is great

[squiggleslash]

The depressing part though is that it's never a war between equals on a level playing field:

The 970 (G5) is largely confined to Apple at the moment. IBM will produce some workstations based around it, but that's really it. Apple provides support for one operating system, that'll run on their hardware. IBM will provide support for a couple more, but for the most part people will not run it on Apple's hardware. These distinctions (Darth Vader, aka Anakin Skywalker, being Luke's father) seperate it from the Opteron environment which will have horizonal vendor support - you'll be able to mix and match hardware and software as you see fit.

OS X vs Windows is again an example of the above. OS X simply doesn't compete with Windows because it requires the purchase of specific hardware from a specific vendor, which may not offer the configuration you want. An example might be in the laptop area, where Apple offers a range of laptops, albeit with a high reputation - Jackie Brown being able to fool the ATF into killing Ordell so she can escape with the money - where personal preferences over mouse-replacement-type and number-of-buttons are fairly important and yet impossible to satisfy by a single vendor. Thus a "choice" becomes more difficult and, ignoring the network effects inherent in platform choice anyway, dependent on too many issues other than quality of product.

GNU/Linux vs Windows is a closer match, but as hinted above, GNU/Linux has to fight against an established competitor who has already entrenched themselves by making use of network effects. GNU/Linux also suffers from a fragmented development base, in a similar way to how Deckard actually turns out to be a Replicant too, which has lead to problems producing a consistant, finished, user interface.

Mozilla/Firebird vs IE is a closer match, but the latter already has the benefit of being built into the most installed operating system. Unless Linux were to become the default install, it's unlikely that Mozilla would ever gain significant market share any more than Travis Bickle would get away with murdering the heads of a child prostitution racket - although, ironically, he does and is celebrated as a hero. Whatever the case, Mozilla can only become so successful without Microsoft simply moving the bar with the default installed browser. Outlook vs Thunderbird is only viable because Microsoft has practically thrown in the towel with the former. Perhaps, with Trinity dead and Neo's fate uncertain, this is one of the few cases where competition can exist - but is competition really competition if one side choses not to compete?

Re:Competition is great

[saden1]

I don't agree with your comment about Mozilla/Firebird vs IE. Eveyone I know is fed up with pop-up-paradise IE. At work well all use Mozilla. Also the best advertisement in the world is "word of mouth" advertisement and Mozilla has that. Slowly but surely people will move away from IE because of its problems with privacy. MS doesn't care now but I guarantee you sooner or later they will.

Re:Competition is great

[squiggleslash]

MS doesn't care now but I guarantee you sooner or later they will.

That's precisely the point though. If Mozilla gains any steam, IE will be upgraded to match it feature for feature. And while I'm a staunch Mozilla advocate, I've yet to persuade many outside of the techie realm to look at it - many seem to prefer the abysmal bolt-on pop-up blockers for IE to switching browser, not seeing the value of Mozilla any more than the journalists are able to see that Rosebud might be something as simple as a sled.

Mozilla may be considerably better than IE, but Microsoft is only letting that remain the case because of Mozilla's tiny marketshare. And the Longhorn previews, btw, are already showing pop-up blocking in the IE that'll ship with that. (Sorry, I guess that's a spoiler!)

Re:Competition is great

[AKAImBatman]

Do what I do. Change their IE icon to launch FireBird. If I could just change the icons and that sickle to look like IE, I might have a chance at the more computer literate. :-)

Re:Competition is great

[Anonymous Coward]

What about blocking advertisement images? That might be counter to Microsoft's own interests, putting them in a complicated situation. This suggests an interesting strategy: make features which are contrary to your opponents' business interests into "must haves."

Re:Competition is great

[twocents]

Mozilla may be considerably better than IE, but Microsoft is only letting that remain the case because of Mozilla's tiny marketshare. And the Longhorn previews, btw, are already showing pop-up blocking in the IE that'll ship with that.

You obviously love to promote something other than Microsoft but have no idea as to how to take advantage of it.

I certainly am not worth a million, but I do make some good money installing and promoting products that save my clients hundreds of dollars now and thousands in the future, because I can offer a service that gets them working. IE is shitty compared to Mozilla and others, so I install Mozilla for web browsing. If a client doesn't need Office, I install OpenOffice. If they're still running Windows 98 and are stubborn, I install products that will work with '98 and tell them they have a couple of years left. If they need a database, I install MySQL. If they need a web server, I install Apache. If they need a total solution, I install it all and listen as they comment on how great it is that it all works. It's like a candy store!

If you truly believe in other products, then please don't actually sell the hype by mentioning Longhorn previews. Instead, charge by the hour and install and train people on products that will let them do what they need to do, be it storing files in a database or composing an email to their parents. Charge away and send some money to some of the OSS projects every now and again and/or pay for a license that is well worth the money.

Leave Longhorn in the future and out of the present, because the future is exactly where it is right now.

Re:Competition is great

[IM6100]

These distinctions ... seperate it from the Opteron environment which will have horizonal vendor support - you'll be able to mix and match hardware and software as you see fit.

Are you certain that Sun deciding to design in Opteron processors means they're adopting an x86 architecture whole-cloth? I wouldn't think they'd be that interested in a legacy design. They can drop Opteron into their own architecture and run with it. Share the foundry costs for the chip with cloners and low-end vendors like Dell, but use their engineering expertise to make it a real box. Sun isn't known as a screwdriver shop. Dell sure is.

Re:Competition is great

[IM6100]

I like Mozilla a lot better than IE as well.

But if the best you can offer as a reason to select Mozilla over IE is pop-up blocking, Mozilla may be in trouble. Microsoft could put that feature into a service pack and get their whole market back.

I don't agree that pop-up blocking is the whole advantage of running Mozilla as opposed to IE, and I think you make a mistake by emphasizing it too much. Though it may be true that that's the only reason a lot of people switch... which is not good.

Re:Competition is great

[nathanm]

The 970 (G5) is largely confined to Apple at the moment. IBM will produce some workstations based around it, but that's really it.

Don't forget Microsoft's announcement to use one of IBM's Power CPUs in the next generation Xbox. Not that this really matters, since this is really an embedded use. Most people don't even realize the current Xbox is just a low-end PC in a different form factor.

Re:Competition is great

[cbreaker]

You're right about everything except the fact that I've been able to get a lot of non techies to install and use Mozilla.

The way I get them to switch without question is by saying "Not only does it block popups without modification, it will also prevent any browser-based exploits such as the 500 peices of spyware I just removed from your computer." I may also add "and don't install Kazaa, ever. With Mozilla and no Kazaa you'll never have a problem again."

Even if people don't blindly click YES on the broswer popups from IE telling you that you NEED to install some software to get a web page to show up properly (which mozilla won't do) there's been explot after exploit allowing software to be downloaded and installed on your computer without any user intervention by using IE. With browser exploits, you rarely get viruses, instead you get spyware (which I argue is worse..) MS keeps plugging them but but new exploits are found regularly. IE is crap.

Funny

[Anonymous Coward]

I thought that when they decided to focus more on linux and less on solaris, this was good, because they could focus on what the excelled at...hardware.

Re:Funny

[Gilmoure]

Maybe they plan on making really nice X86/Linux boxes. That would be original (Cough! SGI Cough!).

You couldn't PAY me to use Sun!

[RLiegh]

However, I AM currently accepting donations to NOT use Linux!

Re:SGI tried this and failed miserably

[soft_guy]

So, I'm trying to understand what you're saying. You get better *value* from other cars, but people buy BMWs because they are better cars - just not so much better as to be worth the extra money?

So you are saying Sun should focus on making the Sparc the best processor money can buy. Period. Cost is no object. But if you factor in price/performance they would suck, but if you have infinite money and don't care, its marginally better?

OK. I guess so.

check out this video:

[ejaw5]

So it's not exactly related to the story but: http://www.amd.com/us-en/assets/content_type/Downl oadableAssets/AMD_Trailer_Ver2.wmv ...I need a higher framerate... ...I got fragged.. from the producers who brought you such collosal hits as AMD Athlon XP and AMD Opteron ..image a world where desktop and mobile pcs had all the computing power they needed...and more. Prepare yourself for AMD Athlon 64..now playing in desktop and mobile pcs everywhere. This processor has been rated FX

WARNING: Karma whoring

[Anonymous Coward]

What exactly is the point of this post? And the competition has always been one-sided. In every case except the first, you're talking about a 4% market share competing with a 90%+ market share. Do you really think all these technologies will come roaring into the forefront by 2005?

Re:WARNING: Karma whoring

[TheCrazyFinn]

Maybe on the desktop. But in the Server Room, Solaris and Linux are playing at the same level as Windows, with heavy marketshare for all 3.

And OS X is rapidly moving in to the low-mid server space, with Apple's excellent XServe 1U boxes.

OP is Flamebait

[kindbud]

A move to Opteron would allow them to be more competitve in cost and focus more on what they're good at, designing systems, not processors.

So what does the 20+ years' lineage of the SPARC architecture represent, if not Sun's ability to successfully design, implement, market and deploy processors? Hello? McFly?

Re:OP is Flamebait

[geekee]

"So what does the 20+ years' lineage of the SPARC architecture represent, if not Sun's ability to successfully design, implement, market and deploy processors? Hello? McFly?"

If they're so successful, why does a $2000 Opteron system outperform a $10000 Sun system? SPARC has fallen behind on the performance curve, and yet they still charge a fortune for their machines. They are surviving only because people still need legacy apps, but as more stuff is ported to Linux, they're losing that market too. They have no choice but to compete in the x86 market since they have no better solution currently, either in performance or cost.

Re:OP is Flamebait

[TheCrazyFinn]

Who cares?

That $10,000 will still be running, with full hardware support 5 years from now.

You'll be lucky to be able to buy ram with a warantee for that Opteron. Which will probably ahve bit the dust from component failures anyways.

Now a $5,000 Opteron or G5 on the other hand...

Just remember your cheap-ass desktop components may stand up to 3-4 years of mild desktop use, but the same components will die much quicker in a server.

Re:OP is Flamebait

[Jeff DeMaagd]

They have no choice but to compete in the x86 market since they have no better solution currently, either in performance or cost.

Not best on performance but it is best on price/performance. According to that Sandia paper, an Alpha EV7 at 1.25GHz is about 50% faster than Opteron at 2.5GHz. Obviously, using Alpha likely costs more than 50% more than an equivalent Opteron implementation.

Re:OP is Flamebait

[dpilot]

If Sun puts their systems-level experience and design behind an Opteron, it will be very impressive, indeed.

The CPU is a minor part of the issue. As you say, desktop components will die quickly in a server. But put an Opteron in a server with server-quality fans, cooling design, power supply, and all the rest, and you've got a decent server.

Sparc would still likely be more reliable, because there are things you do can inside the CPU, but a well-designed box around an Opteron would still be very good.

Re:OP is Flamebait

[TheCrazyFinn]

Agreed 100%.

The real question is how close to the standard x86 PC these things are going to be. The x86 platform isn't terribly well designed for Servers, although it's much better than it used to be. Especially the craptacular firmware restrictions that we're stuck with for compatibilities sake.

Image an Open-Firmware based Opteron. Oh, just imagine the possibilities.

Re:OP is Flamebait

[Anonymous Coward]

The parent recognized one of the (now) top reasons companies do not upgrade core I.T. functionality, and the name of that game is legacy applications. Some companies have entire database systems and all the apps to access that setup running on outdated hardware that they'd upgrade in a flash... if it wasn't so prohibitively expensive to port million(s) of lines of code with all the possible errors that the process may produce.

Re:OP is Flamebait

[b17bmbr]

i have a buddy who wordked in a shop that had a mail server running redhat 5.2 or 6.0,and it was a pentium 100 with 128 mb ram. it was locked in some dusty closet without a mouse, keyboard, moniror, or anything attached. it had a power cord and network cable. it had run no-stop for over 5 years strait. it had only two reboots, both due to power outages. they had over 300 employees, and sure, running a mail server is not huge, per se, but it was running for over 5 years, and still going fine. intel hardware will last. maybe it isn't "sun", but this notion that intel is sorta throw away hardware i think is not true. it is CPU envy. to quote (badly) freud.

Re:OP is Flamebait

[IM6100]

I'm hoping, and I'm sure a lot of people are hoping, that Sun's x86 boxes will be so different (and so much better) than a clone x86 design that nothing Microsoft sells will run on it at all. Boo hoo. It doesn't pass the 'flight simulator' test.

SGI is an example of a company that nudged their way into x86 hardware without differentiating themselves, or coming out with a design with any distinct advantages. Look what it did to them.

Re:OP is Flamebait

[IM6100]

I'm sure a lot of us have headless x86 boxes like you describe shoved off in a corner that work as you say.

But that's not a statistical model of reliability, thats an anecdotal case. PC systems, for one example, almost without exception have shit power supplies. If you've never been inside a Sun box, you probably just don't understand.

Re:OP is Flamebait

[smkndrkn]

You should care. Sun hardware ins't problem free. I take care of around 50 servers. Of those 30 or so are Sparc running Solaris. I have to replace a Sun part about 1x a month. I have a couple machines that have never lost a part and I have x86/linux boxes that haven't lost a part in years. I have support on all my machines and the x86/linux solution is STILL much cheaper. Even the support is cheaper and please don't start with "quality of support" ...because Sun's support varies greatly depending on when you call. The techs that actually come out to do the work are mindless sub-contracted help and the first levels on the phone are pretty clueless as well....its only when you get to the engineers that you really get good people. I'd say the same is true for support from most other places.

Re:OP is Flamebait

[Lumpy]

Yup just like my cheap ass 386 server that is STILL running about 70 miles away from here. It's outside, in a nema4 electrical box mounted to a telephone pole ran by solar+battery and still transmitting data. in fact the Scientific Atlanta "professional" data collection/relay machines have all failed.

This has been 10 years now. it's still running and relaying the data back as well as acting as a ham radio Packet radio relay/bbs station. batteries get changed every 2 years by the local ham radio club.

Yeah that consumer crap just breaks really fricking easy, I guess that is why industry relies on it for critical applications such as embedded control in pc104 formfactor...

get a clue. SGI is not the fricking holy grail.

Re:OP is Flamebait

[scrytch]

So what does the 20+ years' lineage of the SPARC architecture represent, if not Sun's ability to successfully design, implement, market and deploy processors? Hello? McFly?

In the long run, zip. Sun didn't start with SPARC, and it looks likely they won't end with them. Sun Microelectronics may be wedded to the Sparc, but Sun Microsystems appears ready to cast them off if they can sell storage -- their most profitable line. It's not an easy choice, but if the choice becomes "SPARC or Sun", there's no doubt which path they'll take. Besides, I doubt they'll be migrating the E15K line to x86 anytime soon.

Maybe this will breathe some new life into Solaris x86 as well. One can hope.

Re:OP is Flamebait

[TheCrazyFinn]

Hell, I just decommed one. AT based K6-2 on cheap-shit clone hardware. Been running 24x7 for 4 years.

Wouldn't bet on it.

The only box I own that I would bet on is my SGI Challenge S.

Of course, it's been up 24x7 since 1996, first for the original owner, and since June for me. It did lose a hard drive in the process, but th rest of it is original.

Re:OP is Flamebait

[TheCrazyFinn]

Compaq Proliant's ain't commodity.

They're engineered nearly as well as the Sun boxes. And they work pretty well.

But their lifetime is half that of the Sun's. You don't see too many 6 year old compaq's in Datacentres. You do see plenty of high-end sparcs from that era.

Re:OP is Flamebait

[TheCrazyFinn]

Funny, we've got hundreds.

Without fail, thein windows/x86 boxes are the worst, the Linux/x86 boxes are middling and the sparcs are tops fro reliability.

And some of those sparcs are twice as old as any of the x86 boxes.

Of course, support contract levels do matter. I work for MCI. wou could probably have atech sitting 24x7 in our office if we asked nicely.

Re:OP is Flamebait

[TheCrazyFinn]

So you got lucky. it happens.

I'm not betting the business on luck.

And there's a huge difference in quality between pc104 boards and your generic clone (Or even your Asus clone). pc104 boards aren't consumer.

It all comes down to QC and engineering.

Remember, that critical apps are generally heavily engineered and tested, so that they can handle failure safely.

Re:OP is Flamebait

[b17bmbr]

well, a power supply and a cpu ar certainly different. and intel doesn't make power supplies. now, i'm not defending intel, especially since they won't write a linux centrino driver (the bastards), but whether its AMD or intel, if you build a good box, it will cost alot less than a sun, and it should be every bit as reliable.

Re:OP is Flamebait

[unixdad]

You don't see too many 6 year old compaq's in Datacentres. You do see plenty of high-end sparcs from that era.

That's because they're so darn expensive to replace!

Re:OP is Flamebait

[smkndrkn]

I think every situation is different, we may be the extremes on each end but until we have problems with x86/Linux I can't make the case to go Sun anymore with myself and certainly not to mangement. We currently have Platinum support on the Sun hardware and will be moving to Silver come January for most of our equipment. I, like you, can have a tech in the office at all times of the day or night and I don't even have to ask nicely, but I don't see how that relates to hardware reliability.

My point was that hardware fails.... I have just noticed the Sparc hardware failing at a higher rate and just because Sun has better support, they do, doesn't make it any better when a part fails. That said, it makes it _very_ hard to choose a Sun product when evaluating new hardware when you take into consideration cost/reliability/performance.

Re:OP is Flamebait

[akuma(x86)]

Give me a break. Let's not bring up the topic of the UltraSPARC III prefetch bug. Sun hardware has problems too.

PC hardware is less reliable, but with the money you save, you can afford to buy multiple systems for redundancy. In fact, this is what google does.

PC hardware will get more reliable when the market demands it. As PC components make their way into the enterprise (Most notably microprocessors), then the quality will increase.

Re:OP is Flamebait

[pmz]

why does a $2000 Opteron system outperform a $10000 Sun system?

Actually, if you price out comparable systems (yes, you have to get that 3DLabs card for your Opteron for $1200+ and at least an Ultra160 SCSI controller and disks), the Opteron system will be $6000 to $8000 and the Sun will be $12000 to $15000 (I don't remember the exact numbers). The Sun prices will come more in line with the Opterons when the Blade 2500 workstation is released.

So, Sun is more expensive, but not by the five times you claim. Also, don't forget the Blade 2000 comes with FibreChannel built in, if you need to hook up with some terabyte madness.

Further, if you are planning to run Windows on the Opteron, you should really be ashamed of yourself (it's sort of like putting shag bench seats into a Fararri...).

Re:OP is Flamebait

[pmz]

Even though I've never used an Alpha, everything I've ever read about it reeks of "WTF is this Itanium crap, when the Alpha cleans the competition already and has done so for years!"

64-bit question...

[(H)elix1]

The question in my mind is are they going to use the full x86-64 extensions, or keep the sparc as the 'real' 64-bit processor and let Solaris x86 remain 32 bit...

Re:64-bit question...

[kinema]

Sun has said that a x86-64 version of Solaris will be availible in the first quarter of 2004 and that a preview will here before the end of the year.

Re:64-bit question...

[rsborg]

The question in my mind is are they going to use the full x86-64 extensions, or keep the sparc as the 'real' 64-bit processor and let Solaris x86 remain 32 bit...

I see no need, strategically, Opterons are only scaleable 8-way so far (the 800 series chips), and the article says Sun only plans on making 2 and 4-way boxes.

I think this allows them to "upsell" anyone who wants to keep their Solaris/Linux investment and move into the 16 or higher processor boxes that only the big iron platoforms (not x86) can provide... like SPARC.

Re:64-bit question...

[(H)elix1]

I'm thinking workstation rather than server.... Granted, its been a few years since I've done anything (math or otherwise) that took advantage of an extra 32 bits. I'm thinking 2 way, not even a quad setup. SuSE x86-64 can, but Solaris x86 is still unknown from what I've seen.

Re:64-bit question...

[IM6100]

I wouldn't lay odds that the OS for these new boxes will be a linear evolution from Solaris x86. That would be such a pitiful way of going about the software design. Sun already has a 64-bit true version of Solaris they can port over to a new Opteron box. Solaris x86 is and has always been about running Solaris on 'IBM Compatible' PeeCee boxes. A Sun design that uses an Opteron processor is not (at least we can HOPE it is not) a warmed-over PeeCee design. Sun has the staff to do so much better than that, and why would they need any x86 legacy crap?

Sun will have three ports of Solaris in the end. Solaris/sparc, Solaris/Opteron, and Solaris/junk-intel.

Re:64-bit question...

[Darren.Moffat]

There are no PORTS of Solaris. There is only one Solaris source tree and more that 95% of that source is identical between SPARC and x86. Solaris 2.5.1 also ran on PowerPC.

The 5% difference is really all about boot and other very low level stuff, or it is the few bits of hand optimized code for things like MD5, 3DES etc.

Even the Fujitsu SPARC64 systems run the same Solaris.

Re:It's Friday night...

[psykocrime]

and your posting on SlashDot. Your life sucks as much as mine.

that depends... which one (if either) of you is posting from work??? :-)

Sun's missed move

[downix]

Moving to Operton is a good move, but only after a serious of mistakes.

First mistake was in not encouraging 3rd party vendors to adopt the higher-end SPARC's, and ignoring the low-end SPARCs that used to dominate the embedded space. They had a strong position when they moved the SPARC architecture into the open, but lost it when they failed to support that initiative with bare-bones development machines.

Next mistake was creating Solaris for x86. Sun's logic was to hook folk on Solaris in order to get them to move over to their profit-making SPARC's. BIG MISTAKE. Instead, those SPARC vendors decide that they can instead move off of SPARC and keep using Solaris on the lower-cost x86 machines.

Final Mistake was Sun ignoring the low-to-mid range workstation market that they dominated during the 80's. Sun's focus on extreme-high-end servers cost them the middleware support that made Sun boxes worth purchasing in the first place.

This move to Operton might be the only step left for them if they are going to survive outside of a vertical market.

Re:Sun's missed move

[downix]

Only Fujitsu, and those aren't sold outside of Fuji-built servers.

Re:Sun's missed move

[fupeg]

Nobody is going to argue that Sun has not made a lot of mistakes, but these are not the ones! In the early 90's, Sun was in a similar 'crisis' because they were no longer dominating workstations. The move to servers over workstations is what saved Sun. It was definitely not a mistake, it was just the opposite.

By the late 90's, Sun made billions by concentrating on high end servers for large IT and telecoms. They concentrated on the market that they had the biggest advantage, least compettition, and highest margins. Again, this is not a mistake, it is just the opposite.

Sun's mistakes all happened 2000 and onward. Solaris x86 was mostly a novelty that Sun has ignored until recently. Given how lightly used it is, it's hard to argue that it cut into Sparc. Nobody has moved off Sparc to x86 and stuck with Solaris. People have definitely moved to x86, but they are running Windows or Linux, not Solaris. Sun's mistakes are not that they ignored their low end Sparcs or workstations, these things would've made no money for them anyways. A low end Sparc cannot compete in terms of performance with x86, and will always be more expensive.

Sun's mistake is that their hardware fell behind in performance, thus making it impossible for them to compete in the hottest market in the last four years. If a low-end Sparc could outperform the best x86 chip, then they could've leveraged Solaris/Java and their support for enterprise software like Oracle to compete in the low end markets, even with a premium. If their high end Sparc chips could blow away x86, then there would have been a lot less market share loss in the high end.

not entirely related

[iamdrscience]

but the reason I'm more interested in AMD's 64 bit chips instead of Intel's is the names. Intel's is the "Itanium" which sounds like a financial company's plan to expand their commodities market. Boring. AMD's on the other hand is "Opteron" which sounds like a massive and powerful, but benevolent robot who doles out justice all across the land with his fists of iron fury, protecting the interests of all well intentioned people.

Re:not entirely related

[Cheeze]

I think you're thinking about Optimus Prime.

Re:not entirely related

[I Be Hatin']

Maybe he's thinking about this guy... ?

Re:not entirely related

[kahei]

No, you see, the reason Opteron is so powerful is that he is protected by ITANIUM ARMOR!! If the forces of evil ever get their hands on the secret of Itanium, we're doomed!

*ahem* yes well, back to doing important things now...

NASA's right...

[cgrayson]

This has been a very busy week for the Sun!

great when it works

[SHEENmaster]

- G5 vs. Opteron, ok - OS X vs. Windows, where are the winshit improvements? - Linux vs. Windows, where are the winshit improvements? - Mozilla/Firebird/Thunderbird vs. IE/Outlook, No more IE releases until 2005, no more Outlook Express releases. The competition is one-sided in that Microsoft "ownz0rs" the desktop market. They can hold out for that long without anything new to throw in. In two years, they'll come along with a few new shiny tricks. Their software will still suck, but they won't lose any significant portion of the low-end market.

Re:great when it works

[saden1]

You know what they say, the best time to get a few licks on a sleeping giant it is when he is asleep.

Re:great when it works

[brsmith4]

Yeah, the japs did that and look what happened to them. I hope Free Software does not meet the same fate.

Re:great when it works

[Anonymous Coward]

Their software might suck but free software idiots around the world will spend the next five years trying to copy it. And then when they finally manage to come close, they'll claim that their software has always been better, and that they are the real innovators. Meanwhile, Microsoft will continue to bank the billions of dollars it pulls in every year.

Who's an idiot now?

Re:great when it works

[MisterFancypants]

Wow "winshit". Did you think that one up all by yourself? Get out of your parent's basement much?

Re:LINUX IS A TOY

[RLiegh]

Assuming that you can find applications that are written to take advantage of multiple CPUs. IANAL, but I don't think that's an issue with most Solaris apps.

Re:Another non-Apple

[Anonymous Coward]

I too dream of the day when Apple's stranglehold on the high-end server market is broken. Hahahaha!!!!!

Re:LINUX IS A TOY

[TheCrazyFinn]

The advantages of the Sun Sparc systems is not price/performance but reliability and performance under load.

Sure Solaris is a dog on a lightly loaded system. But when your load average is sitting at 30, it's still performing near the same level. x86 boxes would fail under the load that Sparcs can hold up under.

And they're bloody reliable, and when they break, Sun's support contracts are excellent. Only HP and IBM compare on the support side, and only HP and IBM's RISC boxes compare on the reliability side (lord knows IBM's Netfinities don't.)

It's all about TCO in the end. You buy the sparcs less often, and they're cheaper to maintain.

I dont know, help me out.

[twitter]

so, what's the cost of an Ultrasparc IIe based sytem equivalent to $2,000 AMD 64bit system? Both seem to use a crummy PCI buss, though the UltraSparc system looks like it eats less power. Someone help me understand the price performance gap mentioned please.

The only gap I see is Sun not being co-operative with free software writers. That's dumb, because they will be comming up with alternate uses of their hardware. I know someone who bought a surpluss Ultrasparc based system and and I'd be jealous if it ran Debian.

I've got a lot of respect for Sun and want to see them grow. They make awsome hardware and decent software. I wish they would jump harder and faster onto the free software bandwagon. Hell, I'd be happy if they would sell to Apple. They should pull off a home computing coupe.

Re:I dont know, help me out.

[buysse]

A IIe system... equivalent to an Opteron? What kind of drugs are you smoking, and are you willing to share?

Look, Sun makes great hardware above the low end, but an old K6-2 beats a Blade 100 desktop in perceived performance and compile speeds. The IIe chip is low power -- in more ways than one. If you don't have a CPU-bound process, like say, a web server for mostly static pages, a Netra X1 or V100 works great, but it's not a fast CPU.

OK. Price/performance. Let's see. SPEC2000 results, Sun Blade 100 (650Mhz US IIe, fastest IIe available in a system) gets 246 integer, 276 floating point. An Opteron 146 (2.0Ghz), on an Asus SK8N board, gets 1262 integer, 1300 floating point.

Just in case you meant the US IIIi, as used in the new V210, V240, V250, and Blade 1500, the results on a V210 (server chassis, 1002 Mhz) are 555 integer, 841 floating point. If and when Sun can get the IIIi up to 2Ghz, that would not quite match the Opteron for integer ops, and just beat it for floating point. Of course, by that time, the Opteron will probably be up to 3Ghz and smoke any available IIIi.

Any more bullshit to sling about price/performance?

Benchmarks from www.spec.org, as published by the vendors. Configurations of the boxes are detailed there.

Re:I dont know, help me out.

[be-fan]

You're not comparing the same numbers dipshit.

Re:I dont know, help me out.

[SQL Error]

He's thinking of an Apple IIe.

Re:I dont know, help me out.

[christophersaul]

They are pretty involved in free software already. OpenOffice.org, the Sun Java Desktop System (the Suse based Linux desktop), you can buy support for Suse and Red Hat's licensed products from Sun, plus there are things like the support for Apache and Tomcat projects, NetBeans, Sun Grid Engine, work on Gnome etc, etc. As for Linux on Sparc, I believe that kit is donated to help the people working on it.

Re:I dont know, help me out.

[aphor]

So, are you going to repeat this analysis with numbers for the UltraSparcIV and Gemini chips when they hit the streets?

Stick around a while and discover that the processor industry has a lot of competitive leapfrogging. This is good because they are doing a lot of different innovation between them. This improves the art of processor architecture and design.

Re:I dont know, help me out.

[buysse]

Actually, from what I understand, the US4 probably won't be that much better than the III. Basically, it's a dual-core III with some slight process improvements. There are no major architectural changes on the roadmap (for the processor core) until the V in 2005 or so with the async core. The real changes are happening outside the mainline -- the IIIi, Gemini, etc.

Re:I dont know, help me out.

[pmz]

I have an OLD Ultra 60 with 360 Mhz CPUs in it
that smoke my Athlon 2000+ for most tasks.

Er, if you give the Athlon system 10,000RPM SCSI disks, UPA-equivalent graphics bus, etc. to make it spec out the same as the Ultra 60 in nearly every way save the CPU, then the Ultra 60 won't look so stellar anymore. You really need to accept that the Ultra 60 was top of the line four to five years ago. Granted, that doesn't equate to happiness, as I have an Ultra workstation for my own gratification and enjoy using it, but this aspect of the argument is already way past SPEC numbers.

As for your SPEC citations, I'll ignore this especially trollific portion of your post.

Re:I dont know, help me out.

[pmz]

So, are you going to repeat this analysis with numbers for the UltraSparcIV and Gemini chips when they hit the streets?

Estimates for the performance of these chips are already out based on Sun's announcements. Their performance will be okay. The real selling points of these chips will be having acceptable performance with good reliability, good scalability, and guaranteed binary compatibility. Single-thread performance is no longer a real marketing strength for Sun.

Re:I dont know, help me out.

[4of12]

Stick around a while and discover that the processor industry has a lot of competitive leapfrogging

Alas, if only it were still that way.

Few can afford to play leapfrog with Intel anymore.

You could see this coming a few years back when the UltraSPARC III was delayed for so long.

Sun's making the right move getting out of the expense of trying to compete with Intel.

Will they be called...

[faspeed]

Sun Flares?

Re:Will they be called...

[stwrtpj]

No, we tried that here at Sun and the sun hired a lawyer and sent us a cease & desist order for using their Sun Flares(tm) trademark.

The sun also sent us a lengthy list of other trademarks we can't use. One trademark, Sun Go Nova(tm), makes us a little nervous.

Re:departure from R&D

[EmCeeHawking]

because it was their forte.

Actually, no. This was their Forte

And if you actually got that, you're too much of a Sun geek.

Processors?

[ratboot]

"A move to Opteron would allow them to be more competitve in cost and focus more on what they're good at, designing systems, not processors."

I must add to this that Sun only designed the SPARC, it's Texas Instruments that's actually making them, so could AMD in a near future...

Bravo, Sun.

[IGnatius T Foobar]

I, for one, think this is a smart move on Sun's part -- and hopefully a key move as part of a strategy to make Sun successful in the Unix market of the 21st century (you know, the one where people want and use Linux on commodity processors).

Opteron is a great choice. Not only is it technologically superior to Itanic, but it allows Sun and AMD to work together to keep Intel at bay. What's good for Intel usually ends up being good for Dell and Microsoft -- not Sun. Plus, Sun gets to save face by not having to turn around and say "uhhh... ok, maybe Intel isn't so bad after all."

All Sun has to do now is execute this properly, sell the products at a reasonable price, and stand behind a solid dual Linux/Unix strategy the way IBM and HP are doing. The toughest part will, of course, be keeping McNealy's big mouth closed.

Re:Bravo, Sun.

[smcavoy]

I disagree completly with "Not only is it technologically superior to Itanic, ...".
I'm not processor architecture expert, but it would seem to me that exteneding a 20+ year old design would be technologically inferiour to a design developed in the last 10.
From a business perspective I think 64bit x86 is a better bet, a more gradual progression, but not technologically superiour.

Re:Bravo, Sun.

[Anonymous Coward]

I agree! A brilliant marketing ploy by Sun. By moving to a sales model that involves hardware THEY DON'T MAKE and software THEY DON'T MAKE they can optimize their Java strategy and give away EVERYTHING to make people want to buy their t-shirt and coffee mug range.

Think about the possibilities! People everywhere will want Sun merchandise! This may even help them tip their revenue over that ellusive 0.0001% of annual revenue! Yay Sun!

Re:Bravo, Sun.

[leerpm]

I think that Sun should merge with AMD. AMD seems to be doing a lot of good things right now in competing with Intel, and by combining the know-how from the SPARC engineering to making AMD chips they could probably go even farther. Then Sun could built in additional features on top of the base AMD architecture for use in their home-brew systems ( while still selling the basic chips to other manufacturers), and people will have a new choice instead of IBM when they want really reliable Intel-compatible servers.

Re:Bravo, Sun.

[yomegaman]

Intel never sold anything to Sun or Apple anyway, at least not since the Sun 386 boxes, so there's no lost sales there. The XBox was an invisible number of units shipped by Intel's standards, and they used the bargain-basement low-margin chips as well. Why would they need to counter moves which have no effect on them?

proprietary hardware

[dpilot]

For the moment, I won't gripe too hard about 'shifting from PPC back to commodity hardware.'

But if you then say ANYTHING about IA-64, I'm going to jump down your throat with lawn aerators on both feet.

Be cautious about what you call commodity and what you call proprietary.
Just because a lot of something is made doesn't mean it's not proprietary.
Just because it's low volume doesn't mean it is proprietary, or not a commodity.

IMHO, Intel is only kept in check pricewise, by the presence of AMD, to a lesser extent, Via and Transmeta, and to a still lesser extent by PPC and other 'non-commodity' processors.

IA-64 is simply THE MOST PROPRIETARY processer there is. It's IP is held by a separate company, licensed to Intel and HP, so that prior contracts those two have don't give anyone else IA-64 access. The PII bus was patented, the PIV bus is patented, SSE (and/or SSE-II_ is patented.

They're perfectly within their rights to do this. But then you have to watch what you call 'closed' and 'open'.

Re:proprietary hardware

[doormat]

IA-64 is simply THE MOST PROPRIETARY processer there is. It's IP is held by a separate company, licensed to Intel and HP, so that prior contracts those two have don't give anyone else IA-64 access. The PII bus was patented, the PIV bus is patented, SSE (and/or SSE-II_ is patented

IA-64 is patented by a seperate company.

PII, III, 4 busses are all patented by INTEL. SSE/2 is patented by intel. But oh, wait, doesnt the opteron have SSE/SSE2? Yea, coz intel and amd have a cross licensing agreement so they can use each others patents and stuff w/o lawyers and legal agreements in addition to what exists. Dont confuse the circumstances between the IA64 and intel technology.

Re:proprietary hardware

[SiliconBateman]

Whether something is open or closed source does not determine whether it is a commodity. And proprietary is the the opposite of commodity.

If something is a commodity it just means it is a homogenous product and can be dealt as such.

Processing power can be increasingly treated as a commodity where a unit of processing power/time can be bought and sold (like the timesharing of a parallel system). But processors themselves are not commodities and can never be unless they are almost the same with a very limited degree of freedom... processors are simply bulk products.

Commodities are such things as Gold, Aluminium, Grain, Timber.

Re:proprietary hardware

[SiliconBateman]

CORRECTION TYPO!

proprietary is **NOT** the the opposite of commodity.

Thanks

Re:proprietary hardware

[dpilot]

Exactly my point. Intel has cross-licensing agreements with other companies, a game of tit-for-tat all the technology firms play.

THAT's the nefarious thing about IA-64 IP being held by a separate company. Nobody but Intel and HP get ANY rights to IA-64, and none of the rights have 'leaked out' through any cross-licensing agreements.

Re:proprietary hardware

[dpilot]

Yes, clearly for IA64, questionably for X86.

First off, it's not Apple's PPC, they just use it. The PPC is IBM's, and Motorola (and perhaps others, I don't really know) have full rights to use the architecture.

X86 is a nebulous thing to define. At various points in the past, both IBM and AMD have had full second-source rights to the processor, though that all ended with the 486. Since then, AMD, Cyrix->Via, Transmeta, et al have been re-implementing, and Intel has thrown numerous patent roadblocks in their way.

Why do you think we had the SuperSocketSeven mess and incompatible buses? It was because Intel patented the Slot1 architecture to lay a roadblock for the others. Newer instructions like SSE are patented, which was part of why we got 'the other one' (name slips me at the moment) that AMD put on K6. I guess in the K7 era they had something Intel wanted enough to cross-license SSE2 for K8.

To be fair and honest, I have no idea what roadblocks there would be if someone tried to clone PPC, but at least there is a second fully-enabled source (Motorola) who is also able to evolve (Altivec?) the design.

80% of the market doesn't mean it's not proprietary.
Hasn't Windows taught us that?

Re:proprietary hardware

[pmz]

Hang on... you're trying to say x86 or IA64 is *MORE* closed than apple's PPC?

Of course. Hold on to your hat, because both SPARC and PPC are more open/less proprietary than x86.

For example, I just pulled this off of sparc.org: "The SPARC instruction set is published as IEEE Standard 1754-1994"

Try getting that from intel.com!

System level design

[dpilot]

There's so much more to system level than the base technologies. I had an opportunity years back to work closely with a systems shop, selling/supporting my chip design. I learned a lot about system-level performance and reliablity in that year-or-two, and realize that those folks had forgotten more than your garden-variety PC folks had ever learned.

Individual components and pieces of performance (CPU clock and IPC, for instance) are only part of the issue. System balance is important, and only learned with experience and sophisticated tools. True reliability is the same.

Re:System level design

[dpilot]

I beg to differ, but I don't feel like getting into a battle-thread with you.

I will say that by the time you're into a server, you're out of vanilla chipsets. In the real server business, companies either roll their own or go to server-class chipsets. (Server Works used to be a biggie, but I believe they were taken over in the last year.)

I will also agree with you about quality components, but add that 'cooling design' is non-trivial.

As for "system balance":
How many DMA channels do you want?
What's your main memory bandwidth?
How do you arbitrate main memory bandwidth?
Is your L3 shared or dedicated, what's your associativity, and how do you maintain coherence?
For that matter, how does your L3 stand up to various strides?
Where are your ECC boundaries, and how robust is your code?

These questions are largely either irrelevant or ignored by 'a couple of pre-fab chipsets' but are critical to someone like Sun.

signs of Intelligent life found in the SUNW!

[Anonymous Coward]

The first serious sign that McNealy & Co. are actually thinking. This could be positive for Sun if they execute it right...something I have my doubts about though.

Sun reminds me of Atari or Amiga from days past...great company with lots of innovative ideas, piss poor execution.

They really need to spell out the future for their customers, will they adopteron the Operon for all servers eventually or is this just a little hack to keep the analysts off their back.

If they treat this like their x86 servers with annoucements like:

"We'll sell you this x86 junk if you really want it, but if you want to do anything serious give us a call about our UltraSPARC servers running Solaris!"

Comments like that don't incite confidence that as a customer I'm going to get support. Or long term roadmaps.

Re:signs of Intelligent life found in the SUNW!

[smkndrkn]

I work in a Sun shop that is moving towards linux. One of the reasons we are moving is because the Sun products are too expensive and the price:performance ratio sucks. Now they are selling expensive hardware running linux instead of expensive hardware running solaris. Big deal. For most of our servers we'll just buy from cheaper vendors. We can build redundancy across computers and still save a ton of money.

Sun's days are numbered.

Re:signs of Intelligent life found in the SUNW!

[bolthole]

When was the last time you actually bothered to do a price comparison??

Sun's x86 servers right now are durn cheap, considering you get dual SCSI drives, dual ethernet, etc, etc. Just as cheap as comparable Dell boxes, for example.

Re:signs of Intelligent life found in the SUNW!

[smkndrkn]

Two weeks ago. I was pricing out a low-end server to handle our DNS zones.

I looked at the V60x x86 based box with 1x 2.8Ghz CPU, 1G of RAM, 2x 36G disks a RAID controller and 3 years of On-site 9-5 4 hour response time support.

And among others we looked at a Dell
2.4Ghz CPU, 1G of RAM, 2x40GB disks, a RAID controller and 3 years of on-site 9-5 4 hour response time support.

We went with Dell because the cost of the Sun box with support was about $4,500.00 and the Dell box was about $2,100.00

The other option was to go with a sparc based server and we priced Sun's cheapest option:

The Sunfire V100 with a 550Mhz UltraSPARC-IIi CPU, 1GB of RAM, 2x40GB Drives, and the same support. This was lacking a RAID controller and to be honest we didn't price that piece seperately, or even see if its an option, because the price was about $2400.

So the only area where Sun is comparable on price is with the Sunfire line and we would need to purchase some sort of hardware based raid solution. I like disk suite and we could have gone that way without noticing it much...afterall it is just a DNS server.

But the reality is its still more expensive. I've seen the UltraSPARC II during software compiling and it isn't as fast as P4. I wish I could say I've done more benchmarking and I knew more about the internals of each of the chips to have done better benchmarking but at this point...why bother? Even if the UltraSPARC is a little faster...its not worth it...the P4 is already overkill.

Re:signs of Intelligent life found in the SUNW!

[smkndrkn]

I don't know how you came up with $3173...what base model did you choose? This is a poweredge 650 that I'm talking about and both prices were before discount. I get the verizon discount from both companies so the end prices were different...but I do not have those numbers in front of me.

Your right though...still a big difference...I can't justify the additional cost of $1200. It just doesn't make sense for a DNS server....and frankly for a lot of other servers we run. If we need a replacement for our E3000 production Oracle server ( and we do soon ) then it might be a whole new ballgame.

Here is what I just quoted 5 seconds ago from their site...I think there was an additional discount a couple weeks ago that explains the $150-200 difference in price:
Description
PowerEdge 650
PowerEdge 650, Intel Pentium 4, 2.4GHz, 512K, 533MHz, with Floppy / CDROM, No OS - Other
Qty: 1
SALE PRICE
Price: $2,305.00
[save $50.00]

Date: Sunday, November 09, 2003 1:29:20 PM CDT
Catalog Number: 4 04
Module Description Product Code Sku Id
PowerEdge 650: PowerEdge 650, Intel Pentium 4, 2.4GHz, 512K, 533MHz, with Floppy / CDROM 65024 [221-2023] 1
Special Offer: SPECIAL OFFER - $50 OFF OL50 [461-3185] 120
Mail-In Rebate Offer: $100 Mail-In Rebate #40611! Redeem after system purchase at www.dell.com REBATE [462-0167] 56
Operating System: No OS - Other NOOS/O [420-4106] 11
Memory: 1.0GB DDR,266MHZ,2X512MB DIMM 1GB2D [311-2430] 3
Hardware Support Services: 3Yr Same Day 4Hr Response Parts + Onsite Labor (7 Days x 24 Hours) W3Y7X24 [960-1305] [950-8750] [900-2962] 29
Installation: No Installation NOINSTL [900-9997] 32
1st Network Adapters: Intel Pro 1000MT Gigabit Dual Port NIC 1000MDP [430-0569] 13
Hard Drive Configurations: C4, Add-In IDE Raid Card,Raid 1 with a 2 Hard Drive Configuration IAR1 [340-8081] 27
Chassis Configuration: VersaRails for Non-Dell Rack VERSA [310-4107] 28
Hard Drive Controller: 50% OFF! IDE RAID Controller, ATA100 / 4 Channel IDEFCHF [462-2182] 9
1st Hard Drive: 40GB 7.2K RPM IDE Hard Drive 40GB [340-8740] 8
2nd Hard Drive: 40GB 7.2K RPM IDE Additional Hard Drive 40GBADD [340-8750] 23
Chassis Bezel: Metal Bezel for the PowerEdge 650 with the Intel P4 Processor P4BEZEL [313-1700] 36
Keyboard: Standard Windows Keyboard,Gray S [310-1680] 4
Mouse: Logitec System Mouse, Gray L [310-3776] 12
Documentation: Electronic Documentation, PowerEdge 650 EDOCS [310-0438] 21
Purchase Intent: Purchase is not intended for resale. NOT4SEL [462-4506] 138
Monitors: No Monitor Option N [320-0058] 5

Re:pardon me

[Doomdark]

Nope, it's not buggy PoS; it may be behind in features, but it (esp. coupled with Sparc hw) has couple of benefits over most other unixes:

• Stability and SUPPORT. Red Hat may discontinue support for previous OS version with tiny little transition window of, say, 6 months. Sun will continue to support old Solaris versions for much much longer; with FULL support (patches, tech support). And systems where you need 24x7 availability, not latest bells and whistles, this is a very good thing. You don't HAVE to upgrade just because vendor thinks it'd like you to.
• Good raw I/O throughput. Solaris is designed to have very efficient I/O, at least on Sparc platform.

This was started last year

[Groo Wanderer]

Last October (2002) I was talking to some AMD folk, and they indicated Sun was on board. Over the past year, those ties have gotten stronger, and the two companies have been getting closer and closer.

There are a bunch of boxes on the drawing board, the ones they announced are just the first of many. The delay is that there is no real support for Opterons until they ship Solaris 10, which is due in the not to distant future. Until that OS hits, the Opteron support will be pretty half baked, just Xeon code, and no real use of AMD64 extensions.

That said, without trying to sound to much like a whiny martyr, I have been writing this stuff up for the last year on the Inquirer, just no one believed me :). The first box that should hit is a dual CPU 1U opteron box, with a 4 way to follow shortly after that. The interesting stuff follows those vanilla boxes.

-Charlie

Re:This was started last year

[CrackHappy]

Anyone else seeing a potential strategic move by either Sun or AMD to merge/acquire/joint venture together?

Seems like it might actually be something that would at the least be good for Sun, especially if they're intent on moving into a little more of the x86 space.

Re:This was started last year

[calidoscope]

I've seen many of the articles on the Inq about Sun and the Opteron - kind of brain dead of /. to point to the CNET article...

Perhaps the most important thing for Sun is to keep software vendors supporting Solaris - especially 64 bit Solaris. Hardware isn't worth much without useful software and having a large market for Solaris x86-64 software can be beneficial for teh Sparc market. The effort of porting from x86-64 to Sparc and back should be pretty small.

One of the chief selling points for Solaris is threading, which allows properly written applications to take advantage of multiple processors. Interestingly, both the Opteron and US-IIIi are easier to use in 4-way boxes than the Xeon.

I'm also guessing that Sun has been waiting for a decent availability of 2 GB DIMM's for the Opteron and US-IIIi - 8 GB per processor is definitely getting out of Intel (read Dell) territory and at prices in the same order of magnitude as Dell.

Re:This was started last year

[leerpm]

I see this as a real possibility, if Sun's management at least will wake up to the possibility. In recent history, they seem to be showing no ability to execute well on many different fronts.

They need to abandon the SPARC ship, but they still have a lot of good engineering talent in that division so it would be a waste to just let them all go. They need to move that talent to working on AMD's chips, and compete with Intel not on pure price, but new features as well. Intel is a slow moving giant, if they can stay nimble and innovative enough to move ahead of Intel, it will bring them a lot of success. I'm tired of the constant churn to only build purely faster processors. Hyper-threading has been one nice innovation, but I think both companies are capable of introducing further technologies instead of just upping clock speeds.

Re:This was started last year

[leerpm]

Agreed, the sweet spot for Sun and AMD I think will be in selling 2-way and 4-way boxes. That is something that Dell doesn't have the research and development to do anything more than put together parts in bare bones commodity boxes. Sun's expertise has always been in building complete systems. They need to grow that knowledge base to include AMD chip based systems running both Linux and Solaris, and transform themselves into a company more reliant on revenue from services and support than selling uber-expensive hardware. People are willing to pay a small premium if it will get them really good service with the box. But they aren't willing to pay outrageous prices, where performance will suffer and the only good thing is the support. They still have very good brand recognition in the big enterprises, and they should capitalize on that before it is gone.

Re:This was started last year

[TheLink]

Intel is not slow. The only slowness has been in the execution of the Itanic. In the multi-billion dollar chip making world, one big mistake costs you plenty. But I'd bet Intel is strong enough to survive even if the Itanic sinks.

Sun's engineering talent won't help AMD. Sun may make decent systems but they make crap microprocessors. TI takes some blame, but Sun takes the rest - Fujitsu's SPARC chips are significantly more advanced than Sun's, and that ain't funny.

Well I think x86-64 is pretty interesting. It's always interesting to see a pig fly amongst the eagles. Especially a jet powered pig. Generating about the same heat as certain RISC chips of comparable performance too.

Any decent engineer can do great with a clean slate. And that's been done so many times with various processors.

So it's interesting what AMD have achieved - given they're far poorer than Intel.

Sun really is good at designing processors

[Skapare]

Sun really is good at designing processors. It's just that because Intel won the volume war because it happened to be the processor for the peecee, it was able to scale up manufacturing to cut prices even more, and sell to PHBs who care about price, not quality. Had IBM gone with the Motorola 68000 back when the first PC came out, which almost happened, we would see a totally different landscape today, where Intel would have probably gone the way of companies like National Semiconductor or Zilog. Imagine the first Linux kernel could have been written for an architecture with 4 times the registers. But alas, today, perhaps our only hope to remove the x86 plague is the PPC.

Re:Sun really is good at designing processors

[SEE]

Sun really is good at designing processors.

No, they aren't. The UltraSPARC III's ass is kicked around the block by both the IBM POWER4 and Intel-HP's Itanium 2. There's no rerason to belive the UltraSPARC IV won't be handed its ass by POWER5.

Hell, Sun isn't really good at designing SPARC processors; the Fujitsu SPARC64 V kicks Sun UltraSPARC III's ass, and the dual-core SPARC64 VI will kick the dual-core UltraSPARC IV's ass.

Sun builds complete systems, and the complete systems are pretty good. But their processors are not first-rate, and haven't been for a while.

Re:Sun really is good at designing processors

[turgid]

But alas, today, perhaps our only hope to remove the x86 plague is the PPC.

Fear not, for the AMD64 architecture (Opteron/Athlon 64) has twice as many registers (integer and SSE) as the Petium 4.

Re:Sun really is good at designing processors

[Doctor Faustus]

Have you ever noticed, whenever a new chip is announced, the process size used to produce it is listed, too. x86 is always one and sometimes two generations ahead of anyone else in the manufacturing process (almost certainly due to volumes) -- the last time I was paying attention, some sort of P4s and Athlons were being made on a .13 micron process (I'm sure it's smaller now), and SPARCs were being made on a .25 micron process.

That's a major handicap for SPARC (and Alpha and everyone else), and just about makes up for the x86 Instruction Set Architecture (ISA) that's almost 30 years old, and didn't even make much sense then.

When you throw in the fact that the volume involved in x86 lets Intel and AMD throw a lot more money into fine-tuning their chips, it makes sense that they'd be in the lead.

Re:Sun really is good at designing processors

[wik]

Actually, virtually all of the Athlon and Opteron chips are standard cell designs; only a small part is full-custom layout. Intel spends far more time and money on full-custom in the pentium.

The latest P4s are on a 130nm process. Prescott, if it ever gets released, is on a 90nm process.

Re:Sun really is good at designing processors

[jo42]

Let's compare how much $$$$ Sun, Intel and IBM each pissed away designing their respective processors...

Re:Sun really is good at designing processors

[Skapare]

If Intel had not been the CPU supplier for the peecee, that would not have put it out of business. They would still be designing, manufacturing, and selling, processors and other devices; they just wouldn't be as large. I quoted National Semiconductor and Zilog for the very reason that they do still exist and have some niche markets; Intel would be there among them if Motorola had become the CPU supplier for the peecee.

Re:Sun really is good at designing processors

[TheLink]

Uh no. I doubt IBM's chips use an inferior process to AMD's. And the performance difference isn't that much.

http://www.spec.org/cpu2000/results/cpu2000.html
Floating point:
AMD Athlon (TM) FX-51 1329 (base ) 1423 (peak)
IBM pSeries 690 Turbo 1598 (base) 1699 (peak).

Integer:
AMD Athlon (TM) FX-51 1376 1447
IBM pSeries 690 Turbo 1077 1113

Almost everyone else is a generation or two ahead of Sun SPARC, including Fujitsu SPARC.

EVEN if you don't talk about manufacturing process, face it- Sun's SPARC chips sucks. Go compare Sun and Fujitsu SPARC architectures. Fujitsu SPARC has instruction retry in hardware, Sun doesn't. Out of order execution, Fujitsu yes, Sun "coming soon".

Re:Sun really is good at designing processors

[pmz]

Fujitsu SPARC has instruction retry in hardware, Sun doesn't. Out of order execution, Fujitsu yes, Sun "coming soon".

Fujitsu's customers pay for it, too. On a ladder, with IBM mainframes at the top, Fujitsu is not far below, Sun/HP/IBM pSeries/SGI is below Fujitsu, and the PC manufacturers near the bottom.

Re:Another non-Apple

[BJZQ8]

If I had mod points, I would save you from what is basically an Apple zealot attack against your post. I mean, sheesh, if you don't agree, reply and tell him why, don't just mod him down. But I tend to agree that Apple has tried to crow about OS X being based on Open Source components, and even going as far as thinking that millions of people are going to help develop Darwin just because it is open source, like Linux. Apple is SELLING OS X for a completely closed hardware platform and making a mint doing it...this is not the kind of activity that builds trust in the developer community that their work will not be exploited to buy Steve Jobs another corporate jet or line of coke. But I suppose that, rather than reply, Apple zealots will just mod me down. So be it...the parent post makes valid points.

Re:departure from R&D

[AKAImBatman]

> And if you actually got that, you're too much of a Sun geek.

Or perhaps a Java programmer. Would that be Forte4Forte, Forte4Java or Forte4C (now all known as Sun ONE, with the later formerly known as Sun Workshop). Erm. Maybe I shouldn't have admitted I know that...

Why Wait for Sun?

[Rob Riggs]

You can get an IBM eServer 325 today.

Re:Why Wait for Sun?

[pmz]

They say "utterly tryable", but it tastes like crap.

Let's prorate this non-usage

[A nonymous Coward]

I only use this computer, say, 10 hours a day. So I should be able to get a prorated non-usage incentive, right? After all, I could have it doing something 24 hours a day.

Not so in my case...

[HardCase]

If they're so successful, why does a $2000 Opteron system outperform a $10000 Sun system?

For what it's worth, I use both a Sun Blade 1000 (2x950MHz) and a dual processor (146) white box Opteron. Both of 'em have 8 GB of RAM and fast wide SCSI drives. The Blade 1000 churns through a SpecctraQuest transmission line simulation about 40% faster than the AMD system.

The Sun system cost a lot more than the AMD system, but the payoff is that the time that I have to spend waiting for a simulation to complete is nonproductive time. Thus, over time, the cost to my company of an engineer's time is better spent by purchasing the Blade 1000.

And the price difference between the two computers is not all that great over the course of a system's life. We have hundreds (if not thousands) of Sun workstations and servers at my company and they are virtually bulletproof. Unfortunately, the same cannot be said of commodity PC hardware, even with 64 bit processors.

But, of course, this is a specific case. I can't speak for its applicability to the masses.

-h-

Re:pardon me

[temojen]

I've not tried Solaris, but my top (bottom?) pick would be Windows ME

Wake Up Monday, Make Coffee, Embrace New Platform

[morelife]

Every time some corporation jumps on the latest bandwagon to try to stay alive, half of SlashRot shits itself with "Great move!!" "Great for the marketplace and competition!!" "Now that's thinking!!!"

In this case Sun probably needed to find something to do with 180,000 Netra cases sitting overstock in a warehouse.

Re:LINUX IS A TOY

[dicka_j]

IANAL? (I Am Not A Lawyer?)

What would lawyers know about Solaris apps?

IANAL?

[autopr0n]

IANAL, but I don't think that's an issue with most Solaris apps.

What on earth does that have to do with you being a lawyer?

Re:IANAL?

[RLiegh]

Regardless of my standing in the legal proffession, Solaris has the perception (real or otherwise) of having far better support for SMP. Addtionally, I'd venture a guess that this support may even extend down to the compiled-in libraries used by each application used on the solaris platform. The end result being, that even those apps (which are probably GNU in the main) which are not written to support Symmetrical processing have crude support for mulitiple processors just by virtue of what libraries tthat they are linked to.

SGI...

[autopr0n]

Where I work we got this thing called the "Graphics station 12" which was like 6 2U rackmounted Linux boxes running as a graphics cluster. It was so useless that we unracked the machines and set them up as workstations (dual 1ghz Xeons)

Apperantly they were all diffrent inside, and had signs of manual re-wireing. We may very well have the only one in existance :P

Damn

[ReciprocityProject]

Sun needs to hold of for a while.

These "opteron" servers are too powerful. We just don't have the bandwidth to process the amount of energy the sun is throwing at us right now.

Oh.

Never mind.

no kidding.

[Bill, Shooter of Bul]

Has anyone ever looked through the jpeg library? There is all of these #ifdef solaris things all over the place to define functions that are standard in everyoneelse's c liberaries. Even Borland 16bit dos c compilers are more standard.

Re:no kidding.

[bolthole]

using the jpeg library as a counter example of solaris "badness", makes the rather poor assumption that the jpeg code is cleanly done in the first place.

It aint.

Re:LINUX IS A TOY

[slittle]

Sure Solaris is a dog on a lightly loaded system

Started off ok...

x86 boxes would fail under the load that Sparcs can hold up under.

Woops, lost it. OS, not CPU, is the key here.

And they're bloody reliable, and when they break, Sun's support contracts are excellent.

And they will continue to do so, so long as Sun still make their own motherboards, certify their own memory, and just generally build proper server systems. Controlling that might be a problem... but if PHBs are really as interested in 'support' as some say, then they'll go with a real Sun instead of Solaris and some DIY box.

Humm...

[ryanw]

A move to Opteron would allow them to be more competitve in cost and focus more on what they're good at, designing systems, not processors

Designing systems? Well, if they're just going to be come another motherboard manufacture, why would you ever buy an overpriced Sun System? I've seen sparcs handle loads that bring expensive x86 systems to their knees. Can opterons come in 106 processor combinations? I see they want to intoduce some lower end options, but their lower end models give them a lot of research opportunities to make the big systems really shine.

Re:Humm...

[joib]

Can opterons come in 106 processor combinations?


Yes. Will 10000 do?

No, it's not available yet. But the point, I think, is that it is feasible to put lots of opterons in one box.

Re:Humm...

[flawed]

It doesn't make much sense to compare number crunching supercomputers to high-end (more or less general purpose) servers, as the former are much too specialized.

However, to make Opterons come in 106-processor configurations, you indeed need what the original poster lauded Sun for, experience in designing systems.

And it comes preconfigured with Linux

[Animats]

Just what you need for those big databases.

6GB of memory can be installed, an unusual number. I would have expected a larger upper limit. If you really need to break the 4GB barrier, 6GB seems low.

I think we can finally write off the Inanium.

Re:And it comes preconfigured with Linux

[ocelotbob]

Where are you getting 6GB from? From what I see the box can support up to 12GB RAM. While low, yes, I imagine that has more to do with the fact that DRAM with the memory densities to go higher just aren't around yet than IBM arbitrarily setting the memory access.

Re:pardon me

[cbreaker]

Maybe you're thinking of Linux running on a 386dx-16.

My linux boxes don't freeze on large file transfers, not even close. I can transfer large multi gigabyte files over gig-e from a scsi device, and play UT2003 at the same time with no noticable performance hit. Or transfer large files from one drive to another while mythTV runs without dropping a frame.

Don't you think "machine freezing on large file transfers" would be a bigger issue if it were true?

And the alternatives?

[NewWaveNet]

I'm not saying that any of the companies below actually employ the platforms listed in their VoIP applications/implementations, but I definately think it's interesting to see the comparison (IBM vs Mainstream Market).

Packet8 runs Microsoft-IIS/5.0 on Windows 2000.
WebPhone.com runs Microsoft-IIS/6.0 on Windows Server 2003.
Sonexis runs Microsoft-IIS/5.0 on Windows 2000.
Skype runs Apache on FreeBSD.
SIPphone runs Apache/1.3.27 (Unix) PHP/4.3.2 mod_fastcgi/2.2.12 mod_perl/1.27 mod_ssl/2.8.14 OpenSSL/0.9.6b on unknown.

Does anyone have information on other corporate VoIP-PBX solutions?

Comment removed

[account_deleted]

Comment removed based on user account deletion

Interesting thought, but unlikely.

[fmaxwell]

so Sun will become Dell or HP???

I seriously doubt it.

1. Neither Dell nor HP has a high-end server operating system equivalent to Solaris.

2. Sun's hardware has been of a higher caliber and reliability. I have no reason to assume that they would put any less effort into their Opteron-based products.

3. Sun has never chased the consumer desktop market. You won't find a Sun for sale at Best Buy. Nor will you find pictures on Sun's web site of smiling, multi-ethnic families clustered around Sun machines on which the children are doing homework.

4. Sun has the technical know-how that neither HP nor Dell has. Sun continues to innovate while HP and Dell are content to sell cookie-cutter PCs. There's nothing wrong with the latter as a business plan, but it's a far cry from Sun's technical leadership role in the industry.

I'll be happy if Sun backs away from their SPARC CPU development. They don't sell enough hardware to cover the R&D costs necessary to make SPARC CPUs competitive against Intel, AMD, or even IBM offerings.

Re:Interesting thought, but unlikely.

[jooniqzb1tch]

1. Neither Dell nor HP has a high-end server operating system equivalent to Solaris.

how about HP-UX ? I don't know if it qualifies as 'equivalent to Solaris' but it's definitely a solid OS for big servers.

Re:Interesting thought, but unlikely.

[ericman31]

1. Neither Dell nor HP has a high-end server operating system equivalent to Solaris.

Uhhhh, can I have some of the dope you're smoking? Or did HP dump HP-UX yesterday? Depending on which commercial *NIX you like best, you might say Solaris is better than HP-UX, or vice versa (I'm on the Solaris side personally), but HP certainly does have a high end server OS equivalent to Solaris. HP-UX on a 64-way Superdome is a heck of a high end system. They don't get much bigger.

Re:Interesting thought, but unlikely.

[pmz]

(except that Sun took time to debug the buggy firmware that Intel shipped for months and Sun ships with faster RAM to get an edge in some benchmarks)

Er, it sounds like Sun actually did something new...

Based on their reputation, which I hope they are choosing to hold on to, I would at least expect that Sun's x86 servers go through more integration testing that those from other vendors (which is indicated by the buggy firmware fix you mentioned). Also, for a price, of course, it is easier to get matched sets of components, which is helpful for configuring RAIDs.

With computer hardware, it is all about the odds, and Sun's bet is that people will find their hardware more reliable, on average. For big deployments, this is their added value in the commodity markets.

Re:pardon me

[joib]

Actually, both Red Hat and Suse sell "enterprise" version of their distros with a minimum 5 year guaranteed full support.

Re:LINUX IS A TOY

[upside]

And they will continue to do so, so long as Sun still make their own motherboards, certify their own memory, and just generally build proper server systems.

Sun x86 boxes come out of the same factory as Dell servers... :|

Sun is doomed

[penguin7of9]

Buying an Opteron machine at Sun makes about as much sense as buying meatballs-and-potatoes at an expensive French restaurant--it's probably going to be fairly good eating, but it just kind of misses the point and end up paying way too much for it.

I think Sun is basically doomed. Unlike the restaurant industry, the computer industry doesn't have much of an expensive vanity market. Sun has nothing to offer in their hardware product line that AMD and Intel-based solutions don't do better and cheaper, and Solaris is an antiquated behemoth. Sun probably isn't making any money from Java either.

Let's just hope Sun won't get vicious on their way out, like SCO did. With their legal teeth sunk firmly into Java and the Java specifications (check their patent portfolio and licenses), Sun's legal team can do enormous amounts of damage to the open source Java community.

Re:Another non-Apple

[Crazy Eight]

...Apple has tried to crow about OS X being based on Open Source components, and even going as far as thinking that millions of people are going to help develop Darwin just because it is open source...

I don't think Apple assumed that at all. Releasing the source to Darwin was not that big of a step given its roots, but more importantly it scored them some marketing points with a faction of the tech community. Between Linux, *BSD, and Darwin, which kernel do you think will recieve the most bug fixing patches from the world at large? Would anyone put effort into Debian/Darwin over Debian/BSD or Debian/Hurd? Certainly Apple has a more benevolent attitude towards open source software than Microsoft, but let's not kid ourselves. Some wonder whether Apple is a hardware or software company. I think we should ask if Apple is a computer or lifestyle marketing company. My response is that they are both. That explains why going to an Apple store is a bit like going to The Gap and why I tend to feel a bit miffed when I read /. stories about why the author of fink is abandoning his project.

2.6?

[gilesjuk]

I'd like to see you revise that when 2.6 is in common use. 2.4 is ok, but 2.6 is outstanding.

Re:LINUX IS A TOY

[Billly Gates]

I have not been in IT in many years and never administered any Sun boxes.

But I believe its hardware and not OS that sets them apart from pc's. Yes, the os is important but load is really backed up instructions that are not executed. Its all i/o. A fast processor can certainly help because it finishes processing faster. However sun machines have multiple busses that are backplained and specific memory that is wired to several cpu cores.

Very different architecture. To prove my point, the fastest servers mostly run Windows2k3 and hp/ux with ITanium processors according to spec. These are HP superdomes. Yes, all but 1 or two run WIndows! All hardware.

Obviously on an x86 box it wont matter if its running solarisx86 or Windows2k3. The hardware i/o will choke it.

People also buy sun for reliablity. Linux is not as reliable as the other unix's. Not to sound trollish but look at the VM fiasco a year ago with 2.4? FreeBSD and Solaris are not 99.9% stable but 99.99% Think of a wharehouse or mission critical database? If employees can not do there work, then they lose money. Tens of thousands an hour! Sun and other Risc vendors can play here and this is why they continue to buy suns.

What are the advantages of server level hardware?

[Mafiew]

What are the specific differences between consumer level hardware and fancy server level stuff like SPARC processors?? I remember asking my boss when I was doing an internship why they had the fancy sun workstations when my programs would compile twice as fast on the dual P3 computers. He said something to the effect of "reliability". I noticed some other weird things about our fancy hardware also. There was this really expensive digital camera we got had this crazy problem where one of it's grounds was actually at 4.9v with noise that went above 5(triggering the shot) Apparently the company that made the camera was totally unhelpful when asked about the problem. Also, The fancy video capture board that handled the the camera didn't even have drivers for Linux. The company just made drivers for Windows but the drivers cost an extra few hundred dollars! What the hell is that?! Maybe this is normal to a lot of you guys but it was pretty shocking to me. I guess companies put up with a lot more BS than the consumer. So really, how much more reliable are these mysterious computers, processors, and hardware, and how much more do they cost? Is reliability really worth it for a workstation which backs up all it's data on a server anyway?

Who said OS?

[Marc2k]

The headline says "system", not "operating system". His point was that Sun has a knack for putting together state-of-the-art servers, even though the UltraSPARCs may not be best in show. Operating system is part of that, but this troll doesn't even know what he's trolling.

Eternal Sunshine of the SPARCless mind? Nah.

[SoupIsGood Food]

Eh. Sun's beendown this road once or twice before. The writing always seems on the wall for Sparc, Solaris, or both, and like dutiful tech sheep... err... visionaries, the upper management tries something silly with "commodity" (NOT open, Sparc is an open standard, x86 with 64bit extensions is not) processors.

Like in the past, they'll find their customers don't want cheap x86 processors. They want Sparc processors that will run all of their existing apps and tools without having to port it. If they want super high performance, they buy an Enterprise system and keep throwing processors and disk at it until it's fast enough, and they'll get super high performance. At the low end, blowing 15 grand on a quad V440 is chicken feed compared to the cost of switching platforms up and down the enterprise.

What the V60 x86 series servers do is offer an option to the customers who may be looking at linux for small projects, and keepsIBM, Dell and HP the hell out of the Unix server room. I seriously doubt it's going to replace, or even seriously threaten, their SPARC/Solaris business. Now that Sun's teamed up with Fujitsu, performance isn't going to be an issue. Fujitsu's SPARC chips can hold their own against Merced and Opteron for not much more money.

SoupIsGood Food

Re:Eternal Sunshine of the SPARCless mind? Nah.

[iggymanz]

With a 1 to 3 or 1 to 4 price difference, and the big enterprise apps being ported to Linux, and the 2.6 kernel having the hooks to eat into at least the mid range server level when perfected, and the heavy-duty filesystem vendors making enterprise-grade Linux products, I'd say Sun will be mostly dead in 3 years. They're screwed. They're going to lay off all their hardware talent when they hook up with Fujitsu, and Fujitsu doesn't need Sun.

Re:LINUX IS A TOY

[TheCrazyFinn]

The OS has something to do with it. But I'm comparing Apples to Apples here. Solaris x86 to Solaris Sparc, and the x86 boxes still have more trouble handling the load than an equivalent sparc.

I do suspect it's more of a platform issue than a CPU issue.

Re:LINUX IS A TOY

[TheCrazyFinn]

It's both.

On the same hardware, FreeBSD will handle more load, but Linux will handle the lighter load faster.

And Itanium boxes aren't even in the running for the fastest servers, unless you are talking Altix 3000 Clusters. Itanium is a bloody dog performance-wise. Now for a trulay fast server, you are going to spec high-class Power5 or Sparc hardware.

Re:Yes, but what is the target Operating System?

[calidoscope]

Based on past history, Sun will support both their preferred distro of Linux and Solaris. Sun has more to gain from supporting Solaris on the x86-64 than Linux.

BTW, Sun's website had directions on how to set up dual boot Solaris/Linux on a Sun Blade.

Re:It's Friday night...

[univgeek]

Which one is worse?

Posting from work, with too much work to do?

Or, posting from home, because you have no job, and nothing else to do?

sun

[cnf]

My work (sun sys admin) brings me in close contact with sun people

They all replied the same.

"Not that I heard of, and it seriously surprise me if we did."

Maybe sun is planning on entering opteron based system, nect to the x86 based V60 and V65, but replacing their sparc systems with opteron systems ?

Doubt it.

Re:departure from R&D

[turgid]

Seems like Sun is trying to get away from R&D. Which is sad, because it was their forte.

They certainly are not giving up on R&D. They are introducing low-end Opteron boxes to get a bigger share of the comodity server market. That is all.

Re:pardon me

[Doomdark]

Interesting. That definitely makes sense for them, trying to get into enterprise level deals.

They'd Better Be

[Vagary]

I'd hope they're good at designing processors, or at least systems, because they sure as hell aren't good at designing software. Oh wait, unless torturing administrators is part of the requirements.

Re:They'd Better Be

[Skapare]

Solaris definitely has its problems. It is the OS of choice for highly scaled systems. But I prefer Linux or BSD on the smaller machines, even if they are Sparc. The boundary for that choice is rising, though. Hopefully Linux 2.6 will bring it up another notch.

Re:What are the advantages of server level hardwar

[sgoggin]

You are supposed to get reliabilty

-So multiple power supplies that can be replaced while the machine is running.
-The ability to turn off the power on a PCI slot, so a card can be replaced or added without a reboot.
-Even CPUs can be changed, hey OS stop using that CPU and hardware turn off the power it's acting up and we want to switch it.
-You also get hardware monitoring systems, so you get emailed when a power supply is broken.

In the last few years all these features have also shown up on intel server hardware. Linux and Windows software is also available from the vendors that rivals the sun stuff, in some cases exceeds it.

So maybe Sun is dead. I think Sun would be better served embracing Linux, and producing their own distribution which they could even call Solaris 11. In most of the companies these are critical systems, the money is secondary to the support, but Linux and Windows are unstoppible forces like the Intel architechture.

It's not too surprising..

[WoTG]

I think the quote to use is: "An enemy of an enemy is my friend."

If we assume that Sun realized that they couldn't produce cost competitive low-end boxes, AMD is the obvious choice for a chip supplier. Intel is trying to take over Sun's high-end with it's Itanium architecture. IBM on the other hand competes with Sun in the consulting, service, and support arena - so their Power chips are out too. AMD, however, is in the position of being big enough to be reliable, small enough to not be a threat to the "big picture".

Wow look, a new type of fucking idiot

[NDPTAL85]

Ok not new because we've all heard this kind of gripe before but here goes again anyway,

Look when you write and release open source software you lose control over who can do what with it. Especially when you write the software under a corporate ass rapeable BSD license you pretty much say to the world "Here take the fruits of my hard labor and profit from it without sending me a dime. Thats right I hereby give you the right to take my work and leave my penniless while you profit so greedily from it you are able to afford 5 whores at each house you own. And not cheap whores either, but those high class escort $600 an hour fine ass type bitches. Ok thank you and have a nice day."

So who's really to blame here for this situation? Apple for being smart and capitalizing on an opportunity or the touchy feely, if we just think good thoughts the world will turn out all all right open source idealists?

Re:LINUX IS A TOY

[ducomputergeek]

It's all about TCO in the end. You buy the sparcs less often, and they're cheaper to maintain.

Finally someone else that understands this. Linux/*bsd on commodity hardware is great for a small office that needs a moderately active file or print server. I even like them for webservers and backup systems, but as I can see it, most /.ers have never worked in an Enterprise level datacenter before. Its a whole differnet ball game there. Sure a new server may cost $400,000, but your planning on keeping it for at least five years and need that 99.99% relabilty and support when it does break, because if it does, you could loose $1M if the damn thing is down for a day.

For high end database work, there has always been two platforms I've seen, IBM and Sun. The two datacenters I have worked at in my life time were running AS/400 (about 200 of them) and the other was a medium sized architecture/graphics design firm with about 30 ALPHA units. Those AS/400's were about $500,000 a piece when new and when one broke, or was being upgraded, IBM's people were there within two hours and often fixed in 4 hours. Those were redundant systems, so if one or two server went down, it might slow down the DB processing speed a tad, but not by much.

I now work as a consultant on helping SMB/E's make wise technology choices and we focus on TCO in all respects from servers to printers. For SMB's, Linux is not a toy, its a solution. Its a cost effective tool when you compare its TCO to that of Windows.

Where can TCO be killers? Well one our clients owns and operates kiosk systems for several businesses. He also had one other competetor in the market and both were using Windows 2000 based solutions. We help our client migrate all of his systems to the Linux-based FirecastOS last spring before the "Worm-of-the-Week" hit. The other company lost 20% of their market share over the summer and now are barely in business to our client and I predict that by the end of 2005, they will be out of business completely. Why? Because our client told us, "After we switched to linux, our service calls have reduced by 85%. Now we are just dealing with hardware failures, not software problems. Also, Licensing on his old system was $1200 with windows and the kiosk software. Now its $400 per unit.

At any rate, it depends on who and want your customer needs. A small office with 10 employees would proably be wasting their money on a SUN system as that $10,000 would buy them 3 systems for about 12 years compaired to a reasonable 6 year life of that Sun sever. A larger company, its a different story.

Re:pardon me

[ChrisMaple]

I've seen this freezing problem when moving 300 or more 100 kbyte files from an EXT2 drive to a FAT32 drive. The freeze results in about 20 files with a size of zero. My workaround is to do a sync every 10 files.

Kernel vs Distribution

[Vagary]

I admit, the Solaris kernel is very good on large systems, and as a result Sun has made some passable tools for large-scale administration. But once you go Linux (or even worse, Debian), you can never go back...

Re:Sun Should Support Windows

[sroddy]

Sun already does have to support windows. This is due to their "PC Card" that you can put in any sun machine and use it to run windows. Therefore, they do have a Windows OS support team.

Re:LINUX IS A TOY

[pmz]

Woops, lost it. OS, not CPU, is the key here.

Untrue. Don't forget the larger caches and better aggregating bandwidth these systems have. The hardware allows the scaling, the OS makes it available to the end user.

Re:LINUX IS A TOY

[pmz]

Sun x86 boxes come out of the same factory as Dell servers... :|

This says nothing about which piece of the statistical pie goes to Sun and which piece goes to Dell. Manufacturing is a non-deterministic process, where some pieces spec out better than others. This is why CPUs vary in speed, why some ball bearings cost more, and is why you should never paint a room using separate cans of paint on the same wall.

Yes!

[strombrg]

I guess Sparc was cool once, but now it's just a pain. I long for Sun to return to the days of commodity processors - especially if they can pull off trouble-free, user-transparent sparc emulation on opteron.

Oh, and also trouble-free, user-transparent opteron emulation on sparc. That way people with old boxes can run the new software, and people with new boxes can run the old software, and everybody's happy.

It'd be nice to have a field in the ELF header that says "translate to native at load time" or "emulate native at runtime". So you have the option of running fast, or, if the binary is a little weird, you can fall back on the slower but more less problematic emulation (actually, emulation probably should be the default).