• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

64bit Article

Well, I wrote the second one, so this could be an interesting thread...
 
I posted it over at Strictly AMD not to leave out anyone there that doesnt step into this forum.
 
Well, in response to the rebuttal.

Leonard Tsai, former chief technologist at NEC Solutions, said at a conference in mid-July of 2002 that it would take years for engineers to learn the EPIC (Explicitly Parallel Instruction Computing) instruction set used in the Itanium chips, and that this would delay the adoption of the chip.
Click to expand...
NEC is also a vendor of large Itanium 2 systems, so I guess any trouble they might have had are gone.
Chipzilla’s ability to keep the effort afloat due to its large war chest and ‘bullying’ tactics do not lend credibility to any supposed benefits of moving to EPIC.
Click to expand...
Though they do help, as does Itanium 2's excellent performance.
Tim Sweeny of Epic had this to say about x86-64:
“Since our code is pure C++ and already ran on 32-bit Windows and Linux, the only work required was to make the code 64-bit safe. No Hammer-specific work was necessary to get the port up and running; what we did for Hammer is the same thing that would be needed to run on 64-bit PowerPC or 64-bit Itanium.
In the case of the Unreal codebase, about 99.9% of the code was already 64-bit safe and didn't need touching. Of course, with a million-line codebase, the remaining 0.1% left a hundred or so places in the code that needed updating because of assumptions we made years ago before we'd thought about 64-bit. It was a relatively straightforward process, and took Ryan Gordon about 10 days of hard work.”
He also feels that with aggressive content tools, the need for 64-bit computing is coming…and soon.
Click to expand...
For game content creators, which is not a large market, nor does it have much correlation with the amount of memory that the game will use. Nor do I expect games to be a major part of the market for Itanium. And just because you port something to 64-bit, doesn't mean it will be any faster.
“Because of our new content tools, we're already feeling a very strong need for 64-bit internally right now, and by year's end I expect we'll look at 64-bit as something that we couldn't possibly do our jobs without. We expect this sentiment to carry over to other game developers in the next 12 months, to high-end consumers over the next 24 months, and the wide mainstream all the way down to the lowest end of the market within 36 months.”
Click to expand...
Shrug, so if memory limitations become a problem, the current mainstream Socket 745 A64 platform will have not help it since it has difficulty support 3x1GB DIMMs and many MBs only have official support for 2x1GB DIMMs. Peoplw would be better off with the dual-channel P4 platform which can support 4GB of physical memory.
From a technological standpoint, the differences between x86-64 and EPIC philosophy and design are very staggering.
AMD has removed many bottlenecks associated with traditional x86 hardware and has created a truly evolutionary platform that scales tremendously better than the competition, regardless of product or intended market.
The use of Hyper Transport ‘hooks’ on AMD64-based processors has reduced the latency and bandwidth bottleneck tremendously. The on-die memory controller only gets faster with CPU frequency,
Click to expand...
The ability of the on-die memory controller to scale with clockspeed is overrated, since it is already so fast, the memory latency becomes dominated by the latency of the memory module, which does not change with CPU frequency. Meanwhile, Intel platforms still have performance gains available due as the FSB and memory controller speed is increased, or integrated onto the CPU.
and its dynamic ability to utilize other processors in the system and their bandwidth exponentially increase available memory bandwidth. Unlike existing NetBurst and EPIC architecture processors from Intel, AMD64 processors do not have to share memory bandwidth, nor utilize a shared front side bus. A four-way Opteron machine has 25GB/s of bandwidth available to it. An equivalent Itanium 2 platform only delivers 6.4GB/s of bandwidth.
Click to expand...
Sure, in theory. However, increasing Opterons also increase cache coherency traffic and latency on the HyperTransport links. Itanium2 typically have large caches which help alleviate memory bus contention and most commercial workloads aren't so memory bandwidth hungry.
If Intel is truly going to market the 64-bit extensions of its consumer and workstation products as nothing more than an addressing issue, what incentive does it have to make it a competitive product to AMD’s offering? James makes the point that the uneducated consumer looks at numbers only. Well, Intel’s track record is reducing IPC while increasing frequency, giving a false view of performance. What in Intel’s track record points to the possibility that its x86-64 offering is nothing more than a marketing ploy?
The one good thing to come of Intel’s knee jerk reaction to AMD is that it will accelerate the adoption of 64-bit computing. Microsoft has stated that it will offer only one 64-bit version of Windows for the consumer. Intel cannot use its 10 year investment in Itanium in the large and lucrative consumer market. You cannot scale down EPIC for the desktop. The only alternative, lest Intel is to be seen as falling behind, is to slap 64-bit extensions to their existing 32-bit architecture…inheriting all the flaws and shortcomings without offering any innovation.
Click to expand...
Intel's track record has been to improve performance. The fastest 0.18um P4 was clearly faster than the fastest 0.18um P3 and it was also the case at 0.13um. Northwood was clearly faster than Williamette at the same clockspeeds, and Northwood has received further increases in performance from FSB increases and Hyperthreading. While Prescott loses performance per clock, not all of its capabilities look to have been revealed. There are 30-40 million unaccounted transistors in the transistor budget even accounting for the 64-bit capabilities, and should the rumors of a workable dynamic multithreading engine be true and later enabled, then Prescott will become significantly faster than Northwood. Anyways, Intel's 64-bit implementation is very similar to AMD64 with 8 additional GPRS and 128-bit SSE registers.
Adoption by Sun and IBM of the Opteron, behemoths in their own right, gives incredible credibility and exposure for the Opteron. One cannot ignore the needs of a customer base for long. It is even rumored that HP, the incubator of Itanium, will be offering Opteron servers.
Click to expand...
Adoption is far too strong of a word. Sun already sells x86 servers so selling Opterons is a natural extension for their low-end market. They have not yet given up their investments in R&D on SPARC so until that time, Opteron will never become a major part of their offerings. And Sun is only a shadow of what it once was. IBM selling a model of Opteron servers doesn't mean so much given their large variety of server systems, with Power still being their major top-end product. AMD has had more success in the server market, but it still isn't much.
James conjectures that x86-64 will do nothing for the top end of computing. With the Opteron available in 8-way configurations, and Itanium only in 4-way, how can one make that claim? If x86-64, specifically AMD’s offerings, offer higher computational power, lower initial investment, and lower cost of ownership, why can it not change the face of ‘top-end’ computing?
Click to expand...
Ever heard of SGI Altix, or HP Superdome? These are large single image systems consisting of up to 100s of Itanium 2 processors and require expensive custom chipsets and interconnects, something that no vendor has done for Opteron. And there are to date, not a single 8-way Opteron system that has been implemented, with only Sun declaring it will build one at some uncertain time frame. Itanium also has an assured market at the top-end of computing, giving that it is the intended replacement of three of the five major RISC processors, SGI MIPS, HP PA-RISC and Alpha, leaving only weak SPARC and Power left to compete in the several tens of billion of dollars big-iron market.
James brushes aside the advantages of the Opteron over Itanium by stating that they are to be used by totally different markets. This is an interesting statement, and I think that two points can be made it regards to ‘markets’. One, Intel has created a niche market with its EPIC architecture, and is relying on exponentially superior performance in a few applications to justify the purchase of a very expensive and limited platform.
Click to expand...
Intel has entered the market for large 64-bit systems with Itanium 2 and has growing momentum and excellent products which perform well on industry standard benchmarks and workloads.
Unfortunately for them, the AMD Opteron spans markets and beats Intel at their own game, all the while providing a more versatile and lower cost solution.
Click to expand...
But Opteron does not compete with Itanium, it competes with Xeon. It's only available at 1P-4P and neither Sun or IBM are replacing their high-end 64-bit RISC processor with it.
He also states that change is also rarely swift in this market. I think that this is fundamentally wrong. While companies do want to maximize their investment, even five years is a long time in this industry, which also lends to the fact that while hardware often faster than the applications they run, when software comes down the pipeline in the future the need for greater power is often ugly and painful. Adoptions usually come in paradigm shifts. This is the greatest paradigm shift in years. Companies will adopt it. They will need to adopt it, and they will benefit tremendously from it.
Click to expand...
And why will businesses running HP-UX, or OpenVMS adopt Opteron when Opteron is not compatibile, while Itanium is? There's a large market that does not need x86 compatiblity. And you're wrong, the world of big-iron is slow to change, require long-term assurances about hardware and software availability and support.
 
Originally posted by SLee
Well, in response to the rebuttal.


NEC is also a vendor of large Itanium 2 systems, so I guess any trouble they might have had are gone.

Though they do help, as does Itanium 2's excellent performance.

For game content creators, which is not a large market, nor does it have much correlation with the amount of memory that the game will use. Nor do I expect games to be a major part of the market for Itanium. And just because you port something to 64-bit, doesn't mean it will be any faster.

Shrug, so if memory limitations become a problem, the current mainstream Socket 745 A64 platform will have not help it since it has difficulty support 3x1GB DIMMs and many MBs only have official support for 2x1GB DIMMs. Peoplw would be better off with the dual-channel P4 platform which can support 4GB of physical memory.

The ability of the on-die memory controller to scale with clockspeed is overrated, since it is already so fast, the memory latency becomes dominated by the latency of the memory module, which does not change with CPU frequency. Meanwhile, Intel platforms still have performance gains available due as the FSB and memory controller speed is increased, or integrated onto the CPU.

Sure, in theory. However, increasing Opterons also increase cache coherency traffic and latency on the HyperTransport links. Itanium2 typically have large caches which help alleviate memory bus contention and most commercial workloads aren't so memory bandwidth hungry.

Intel's track record has been to improve performance. The fastest 0.18um P4 was clearly faster than the fastest 0.18um P3 and it was also the case at 0.13um. Northwood was clearly faster than Williamette at the same clockspeeds, and Northwood has received further increases in performance from FSB increases and Hyperthreading. While Prescott loses performance per clock, not all of its capabilities look to have been revealed. There are 30-40 million unaccounted transistors in the transistor budget even accounting for the 64-bit capabilities, and should the rumors of a workable dynamic multithreading engine be true and later enabled, then Prescott will become significantly faster than Northwood. Anyways, Intel's 64-bit implementation is very similar to AMD64 with 8 additional GPRS and 128-bit SSE registers.

Adoption is far too strong of a word. Sun already sells x86 servers so selling Opterons is a natural extension for their low-end market. They have not yet given up their investments in R&D on SPARC so until that time, Opteron will never become a major part of their offerings. And Sun is only a shadow of what it once was. IBM selling a model of Opteron servers doesn't mean so much given their large variety of server systems, with Power still being their major top-end product. AMD has had more success in the server market, but it still isn't much.

Ever heard of SGI Altix, or HP Superdome? These are large single image systems consisting of up to 100s of Itanium 2 processors and require expensive custom chipsets and interconnects, something that no vendor has done for Opteron. And there are to date, not a single 8-way Opteron system that has been implemented, with only Sun declaring it will build one at some uncertain time frame. Itanium also has an assured market at the top-end of computing, giving that it is the intended replacement of three of the five major RISC processors, SGI MIPS, HP PA-RISC and Alpha, leaving only weak SPARC and Power left to compete in the several tens of billion of dollars big-iron market.

Intel has entered the market for large 64-bit systems with Itanium 2 and has growing momentum and excellent products which perform well on industry standard benchmarks and workloads.

But Opteron does not compete with Itanium, it competes with Xeon. It's only available at 1P-4P and neither Sun or IBM are replacing their high-end 64-bit RISC processor with it.

And why will businesses running HP-UX, or OpenVMS adopt Opteron when Opteron is not compatibile, while Itanium is? There's a large market that does not need x86 compatiblity. And you're wrong, the world of big-iron is slow to change, require long-term assurances about hardware and software availability and support.
Click to expand...

HP-UX is designed for EPIC, there are equally good if not better versions of Unix available in 64-bit for Opteron (IBM's AIX).

Opteron can easily compete with Itanium, even if the intention is not there. If it's faster, it's faster. If it's cheaper, it's cheaper.

You're grasping with straws with the ability or inabilities of the Socket 754 platform. Chipsets and implementations will vary and to say that going with a P4 is a better solution because of addressable memory is a better solution, I think that's fairly subjective.

And you obviously didn't understand my point about Intel's performance. They have historically lowered IPC with each new generation. I did not talk about taking the mature NetBurst architecture and compare it to the previous PIII generations. Each new Intel offering has historically be slower than their previously offering, being offset only by frequency increases.

This goes all the way back to the Pentium MMX processor.

And when business changes, it's pretty quick. They may be slow to get there, but when the decision is made, it's generally a one-fell-swoop adoption of a standard. Ie Texaco standardizing on Compaq in the early 90's. It's a dramatic change.

The fact is that Opteron blind-sided Intel, and has offered a cheaper solution to the convoluted EPIC architecture that is overpriced, underperforming, and if it wasn't for Intel's name, and Intel's name alone, would have been abandoned long ago.

Opteron is competing with the Xeon, Itanium, and the Pentium 4. You may not like it, Intel may not like it, but that's the way of it.

Oh, and with half the processors Opteron still beats HP's SuperDome Itanium...so why can't it compete again?

http://zdnet.com.com/2100-1103_2-5160704.html
 
Very nice. The Opteron left that previous score in the dust. So yes I would like to know now just how it cannot compete SLee :confused:
 
Hey guys, first post ever in Strictly Intel forum. Did I miss something by reading that article?

http://zdnet.com.com/2100-1103_2-5160704.html

It appears that the chip IBM was using was a Power PC chip, not Opteron. If I'm wrong, then I stand corrected.

Anyhow, if you were in charge of investing and upgrading your current computer servers, would you run Xeon's or Opterons? Knowing that the Opteron's are faster, cheaper and less expensive to operate? I've seen several 4 and 8 way systems, where the Opterons were $5-$6,000 dollars cheaper.

[FS]Puts flame shield on![/FS]
 
And you obviously didn't understand my point about Intel's performance. They have historically lowered IPC with each new generation. I did not talk about taking the mature NetBurst architecture and compare it to the previous PIII generations. Each new Intel offering has historically be slower than their previously offering, being offset only by frequency increases.

This goes all the way back to the Pentium MMX processor.
Click to expand...
The P5 MMX was faster than the P5. The PII was faster than the P5 MMX.

And when business changes, it's pretty quick. They may be slow to get there, but when the decision is made, it's generally a one-fell-swoop adoption of a standard. Ie Texaco standardizing on Compaq in the early 90's. It's a dramatic change.
Click to expand...
Those adoptions happened a long time ago, and once set, are unlikely to change. Its why Sun can survive on income from support and upgrades despite inferior performance, or why HP is still releasing one last PA-RISC chip, or why they still released the Alpha EV7, to service slow changing businesses.

The fact is that Opteron blind-sided Intel, and has offered a cheaper solution to the convoluted EPIC architecture that is overpriced, underperforming, and if it wasn't for Intel's name, and Intel's name alone, would have been abandoned long ago.
Click to expand...
Yes it would have failed had it been backed by anybody else, but it did have the backing of Intel and now it has reached a point where its initial problems have been mainly solved and now has the capability to stand on its own two feet and be self-sufficient.

Opteron is competing with the Xeon, Itanium, and the Pentium 4. You may not like it, Intel may not like it, but that's the way of it.
Click to expand...
How does it compete with Itanium when there are no >4P Opteron systems available. It competes with Xeon, the A64 competes with the P4, and as long as Intel maintains rough parity with AMD, then Intel wins.

Oh, and with half the processors Opteron still beats HP's SuperDome Itanium...so why can't it compete again?

http://zdnet.com.com/2100-1103_2-5160704.html [/B]
Click to expand...
Maybe you should read the article more clearly. The p690 used IBM's Power4+ dual-core processor, which is the only real competitor left to Itanium at the high-end. The p690 is also not a production system, scheduled to ship sometime in the summer. Its score will likely be eclipsed when the Madison 1.7GHz ships.
 
My mistake on the link...misread from AMD's site. Responding to the rest now.
 
Here's a review of the PII Klamath against the Pentium MMX and Pentium Pro.

http://www20.tomshardware.com/cpu/19970301/klamath-05.html

And you speak like the status quo will not change. PA-RISC, Alpha...how long do you see these processors staying a viable solution?

The time will come with either EPIC or x86-64 is going to dominate. x86-64 is less painfull, lower cost, and provides arguably the same performance or better. Be it a solution from Intel or AMD, 64-bit computing will not standardize on IA64.
 
Originally posted by Morley
Here's a review of the PII Klamath against the Pentium MMX and Pentium Pro.

http://www20.tomshardware.com/cpu/19970301/klamath-05.html
Click to expand...

That was the preview of the PII. Here's the actual review a couple of months later in which the PII showed much better.
http://www20.tomshardware.com/cpu/1...#the_windows_95_performance_of_the_pentium_ii


And you speak like the status quo will not change. PA-RISC, Alpha...how long do you see these processors staying a viable solution?
Click to expand...
There'll be support for several more years at least, at which point the natural successor to PA-RISC and Alpha is Itanium, which can run HP-UX and OpenVMS, and will be the easiest upgrade path.

The time will come with either EPIC or x86-64 is going to dominate. x86-64 is less painfull, lower cost, and provides arguably the same performance or better. Be it a solution from Intel or AMD, 64-bit computing will not standardize on IA64.
Click to expand...
Why does it need to be only one solution? ARM dominates the embedded chip market, and big-iron is large enough to sustain a at least two processor families for years to come.
 
There are those out there that disagree with you on where Opteron can compete:

http://www.fcw.com/fcw/articles/2003/0428/tec-chip-04-28-03.asp

Opteron destroys PA-RISC, SPARC, and Alpha in SPEC scores, so now it's down to Opteron and Itanium.

Intel's solution, while better known by ignorant MIS guys, is more expensive, difficult to code for, and is needed by a small niche of applications.

AMD's (and Intel's x86-64) will have a larger install base and be more versatile in their application.
 
What are the difficulties of switching from a non-Itanium install base to an Itanium platform, even if HP-UX is involved? It would be just as difficult to switch to say AIX. Hardware is hardware, however, the versatility of x86-64 cannot be overstated.
 
Whole lot of words there, not a lot of substance.
Click to expand...

You lend yourself quite the respectable image when your first rebuttal happens to be one sentance long and an insult to boot. Also, calling Intel Chipzilla stopped being witty right around the last season of Happy Days when Fonzy's hair fell out from all that crap he put in it.

Also, for someone who seems to know so much about the server market, you're apparently clueless about the functionality of the edit feature. And the quote function. And the close bold tag. And the...

People like you are why I'm proud of my williamete. *pets willy*
 
Originally posted by Vagrant Zero
People like you are why I'm proud of my williamete. *pets willy*
Click to expand...

Be proud of your willy, just don't pet it in public, you'll get arrested.
I'm proud of my northwood. *Pets wood*
 
Originally posted by Vagrant Zero
You lend yourself quite the respectable image when your first rebuttal happens to be one sentance long and an insult to boot.
Click to expand...

Actually, it was the truth. A lot of words and figures, but it didn't really say much and didn't address the issue of my article, which was that x86-64 means more than the what James was talking about.

Also, calling Intel Chipzilla stopped being witty right around the last season of Happy Days when Fonzy's hair fell out from all that crap he put in it.
Click to expand...

That's great, Dennis Miller, but there's nothing wrong with that term, it's a tongue in cheek thing.

Also, for someone who seems to know so much about the server market, you're apparently clueless about the functionality of the edit feature. And the quote function. And the close bold tag. And the...
Click to expand...

Excuse me, I couldn't really care less what tags I do or do not use on a BB in the middle of the internet.


People like you are why I'm proud of my williamete. *pets willy*
Click to expand...

You know, it's fan boys like you that perpetuate the image of the dorky computer geek. Get over it, Intel, AMD, who gives a shit, it's whichever delivers the better product at the time.
 
Originally posted by Morley
Actually, it was the truth. A lot of words and figures, but it didn't really say much and didn't address the issue of my article, which was that x86-64 means more than the what James was talking about.
Click to expand...

Most people in the world are horribley overweight. Telling that to their faces and proclaiming it as truth would still not justify your actions and certainly would destroy your crediblity as a decent human being.

That's great, Dennis Miller, but there's nothing wrong with that term, it's a tongue in cheek thing.
Click to expand...

You're correct, there is nothing wrong with it...on message boards. But when you're writing a persuasive essay or a rebuttal to a viewpoint its always best to come off looking completely unbiased. It's called professionalism and tact, the first essay had it.

Excuse me, I couldn't really care less what tags I do or do not use on a BB in the middle of the internet.
Click to expand...

Obviously, and herein lies the problem. These tags are rather intuitive and not at all hard to use. I didn't ask you to do hard labor for 20 years. But I see you did take the time out to learn about the size tag. Good for you.


You know, it's fan boys like you that perpetuate the image of the dorky computer geek.
Click to expand...

Really? And here I thought it was all due to the Simpson's comic book guy.

Get over it, Intel, AMD, who gives a shit, it's whichever delivers the better product at the time.
Click to expand...

AMD doesn't deliver a product that competes directly with the Itanuim.
 
It's obvious that rather than address my article, you'd rather nitpick the way I do or do not use tags. I know how to use them, and the very fact that I have to justify that to you is silly. You also have no clue what an editorial is. Biased, unbiased, it doesn't matter. An opinion is inherently biased. This was not a review, this was not an industry analysis paper, it was an editorial. Your petty semantics are boring me.

Frankly, you're just pissed that I wrote what you perceive to be a pro-AMD article when it was a pro-x86-64 article. You are ignoring the performance and the possibilities. Your mind is closed that this company that you have such disdain for could possibly compete with your precious company. Let's say for a minute that it can not, nor is positioned against Itanium. That leaves the Xeon market wide open for attack, and the Opteron competes very well against the Xeon.

In a pure numbers game, Opteron can compete against all but 2 of HP's Itanium offerings when looking at available processor configurations.

Keep petting your Willamette, fan boy.

It was an editorial, get over it. All your posturing and ruffled feathers mean very little to me.
 
It's obvious that rather than address my article, you'd rather nitpick the way I do or do not use tags.
Click to expand...

There is nothing to address. The Opteron does not compete with the Itanuim. They cater to completely seperate markets. Nothing further needs to be said on the issue.

Frankly, you're just pissed that I wrote what you perceive to be a pro-AMD article when it was a pro-x86-64 article.
Click to expand...

Oh yes. I HAVE FURY. :rolleyes:

If you weren't so busy patting yourself on the back you might have noticed that I have yet to say anything pro-Intel or anti-AMD. In fact I'm considering getting an FX-53. Also, it's fanboy; putting a space between it doesn't make you look anymore intelligent. At this point, an act of god wouldn't acheive that.

Your mind is closed that this company that you have such disdain for could possibly compete with your precious company.
Click to expand...

I see. So not only are you a raving looney, but you're telepathic too? Wow, can you tell me what I'm thinking now. If you guessed "boy that chap sure likes his panties knotted" give yourself a nice big hug.


Let's say for a minute that it can not, nor is positioned against Itanium. That leaves the Xeon market wide open for attack, and the Opteron destroys the Xeon.
Click to expand...

Yes and no. The Opteron is clearly the superior solution [in regards to the Xeon, not the Itanuim], however that doesn't mean corportate servers have been quick to jump on the bandwagon. Not only has the transition been sluggish, but AMD actuatlly lost market share this quarter. Thirdly, with Intel's statement on upcoming x86-64 servers, this has effectively grounded corporations that were considering the switch. If they're anything like us regular joe consumers, they're going to wait it out and see who wins the round. Personaly, I hope the LGA 3.6E beats out the 3700+ not because I like Intel over AMD, but just becuase nothing would bring me greater joy than to see you and your rabble get shafted. On the high end market, I hope the FX-53 beats out the LGA 3.6EE as it's the chip I plan to go with. As for the server market, I couldn't give less of a crap. I don't buy, build, or use them. But alas, I cannot lie, I own stock in Intel. I blame those blue bald headed...things. :eek:

It was an editorial, get over it. All your posturing and ruffled feathers mean very little to me.
Click to expand...

You asked for comments on your article. SLee already did an admirable job of putting you in your place so there's very little left for me to add except my stinging commentary in regards to your behavior to him. I mean why should I comment on your article when all you'd do is flame back instead of addressing the points made? Second of all, apparently all my 'posturing' has been enough to elicit mutiple responses even though it 'means very little to you'. You're quite the comical lad. Might I suggest you ditch the whole editorial thing and give stand-up a go. You look like a goer.
 
Again, I stand by every word I said.

The point of the article was that x86-64 has won out, the comments on Itanium were secondary. And in 2, 4, or eventually 8 way servers the Opteron can and will compete quite nicely. You are ignoring the fact that all but two HP offerings are 2-8 way Itaniums. You are also ignoring Opteron offerings by Cray, and you are ignoring the benchmarks above.

According to Anandtech:

"While we do believe that the Itanium 2 in its 128-way configurations is definitely out of the Opteron’s league, in the 2-way and 4-way configurations that we are interested in comparing, the two are absolutely competitors. "

I have an Itanium 2 from HP here in the office. We also have many Opteron systems that we manufacture. This is what I do.

You are quite witty, and can seem to talk rings around me when it comes to personal attacks, I commend you.
 
Vagrant you seemed to be a bit lost. I love Intel and AMD products. Whichever one gives me the best performance for my buck and really gentlemen, thats what it boils down to. AMD has just been slapping around Intel in that respect for quite a while now.

How on earth do you say the Opteron doesnt compete with the Itanium. Could be that its not its intention but to say it doesnt is preposterous. It competes with the Xeon, Itanium and the P4....competitively I might add.

If you havent seen the numbers...Roll on over to any tech site. Like this one. You have read articles on the Opterons performance havent you? Or are you really too busy with that willy? :p
 
Originally posted by Vagrant Zero
Most people in the world are horribley overweight. Telling that to their faces and proclaiming it as truth would still not justify your actions and certainly would destroy your crediblity as a decent human being.
Click to expand...

I would tell them to their face. Who cares?

If they don't like being fat, then do something about it.
 
Originally posted by 2phastPRO
If you havent seen the numbers...Roll on over to any tech site. Like this one. You have read articles on the Opterons performance havent you? Or are you really too busy with that willy? :p
Click to expand...

Look again, by competition I mean sales, not performance. If you'd bothered to read you'd have noticed that I say the Opteron blows the Xeon away performance wise but that corporations still won't take it over the Itanium. The fact that AMD lost market share this quarter is proof of that.

And Josh, I wouldn't reccomend riling up fat people...they might sit on you...or *shudders* eat you. :D
 
So...you're talking about sales performance? I was talking about actual performance. After reading Morleys posts, I thought he was too. Why are we talking about the sales side of things? It seems as if you thought the Opteron couldnt compete with the Itanium and wont. Period.

I thought it was quite obvious that Intel has the sales side of things locked down. Arguably the Opteron is a much better solution from what i've seen and read. Intels manipulation of the big companies lends AMD sales a drop.
 
You must log in or register to reply here.
Back
Top