Opteron Benchmarked Against Xeon

jbmnuke writes "Tom's Hardware has posted a review of AMD's Opteron v. Intels Xeon." Nothing gets the blood pumping like a whole new generation of CPUs to compare numbers to, right? Update: 04/22 12:35 GMT by H : And there's the official benchmarks as well, with more coming - like Linux Magazine and Newsforge

Comments

Bleh!

[matttastic]

It doesn't get my blood pumping, I can't afford such things (or cool them)! *Pats Duron 1000*

Re:Bleh!

[madmarcel]

Neither can I :( <<pats crusty old Celeron 466>>

But but but...given time, these things will be common as muck and we'll all have at least one.

<nostalgia>
Ah, I remember when 386's and 486's where top-notch stuff and hideously expensive..
</nostalgia>

NOW...I have a whole attic full of %^$#@#!* 286's, 386's and 486's and I wouldn't know what to do with them :o

I have a dream...that one day I'll have an attic full of 'old' opterons and xeons....and I won't know what to do with them ;P

Re:Bleh!

[byolinux]

Duron 1000 would make such a good name for an evil droid.

Re:Bleh!

[Perdo]

That Duron is pumping 45w of heat....

Opteron is putting out 41w

Xeon 3.06 is putting out 81.9w

And the real beauty is, an XP 2400 cost $94 because of the opteron price war.

Reaganomics lives in tech land.

All the good stuff trickles down to us eventually.

Re:Bleh!

[TheLink]

"Pats Duron 1000".

Free advice: do not pat more recent microprocessors with remaining hand ;).

Re:Bleh!

[Citizen of Earth]

It doesn't get my blood pumping, I can't afford such things (or cool them)!

1. Open up case
2. Point 12" desk fan at CPU, turn on full blast
3. Duck, as dust is blown out of case

Desk fans: Keep CPUs cool and cases dust-free.

Re:Bleh!

[Bombcar]

Use a paintball gun (with no paintballs, of course!) Works great.

Well, that and...

[subreality]

Nothing gets the adrenaline pumping like the flood of trolls this sort of comparison should inspire.

Re:Well, that and...

[byolinux]

... also, someone's bound to point out that the IBM 970 is going to be a whole lot faster, and we Mac users will once again triumph! Wuh hah hah!

it's all well and good..

[rj-eleven]

but it reminds me of a benchmark performed between desktop x86's and a sun machine. Given the different architectures, it really didn't make sense. However, the benchmark was supposed to show price::performance. Is this what Pabst is trying to convey? I don't take much stock in benchmarks anyway, as I would rather get my hands on it and try to break it.

Re:it's all well and good..

[odaiwai]

No this is not what Pabst is trying to convey.

What Pabst is trying to convery is that he needs more views on his website, even though historically, he's made a point of exagerrating the statistical differences between test results to push Asus motherboards (for example). I remember him making a huge procuction out of a less than one percent difference in the performance between sone dual processor motherboards. I realised then that either he was mathemathically incompetent or he was just a shill for his advertisers.

Either way, he's not worth the bother of checking out anymore.

dave

Current Review: Xeon vs Opteron

[Judg3]

Next Review: Apples vs Oranges, Which has more of an Orange taste?

Re:AMD is dead

[trezor]

• Considering how P4 3.06Ghz actually runs at 3Ghz and does much better at important tasks such as encoding video, instead of synthetic benchmarks, we therefore conclude AMD is officially dead.

So you do not believe that Intel got where they got today becuase of competition and pressure? You sincerely believe that Intel wouldn't sit back on their lazy ass and inflate prices, if there were no copmetition?

Naivity ensues obviosuly.

Old news...

[G3ckoG33k]

Suddenly, I feel this is old news... It came out five hours ago for gossake!

Nonetheless here is the condensed version:

_____________Server_______Workstation
Opteron__ ____Very good____Good
Xeon_________Good_________Very Good

Opteron vs Xeon

[mrgrey]

Didn't I see this in an Evangelion episode?

Blood pumping?

[Anonymous Coward]

"Nothing gets the blood pumping like a whole new generation of CPUs to compare numbers to, right?"

I find sex better, whatever flicks your switch I guess...

Impressive SMP scaling

[Anonymous Coward]

Check the Spec benchmarks here.

SpecFP_rate, 2CPUs:
Itanium2 1GHz: 30.7
Opteron 1.8GHz: 26.7

SpecFP_rate, 4CPUs:
Itanium2 1GHz: 49.3
Opteron 1.8GHz: 49.2

Here we see the beauty of AMDs integrated memory contoller. Despite that 1GHz Itanium2 is a $4000 chip and has 3MB of cache, doubling the number of CPUs increase performance only by 60% because Itanium2 uses shared bus.

Opteron gets impressive 84% improvement because
memory bandwidth increases as more CPUs are added.

In SpecInt Opteron is much more faster than more expensive Itanium2.

Re:Impressive SMP scaling

[Zathrus]

No, if you go purely by SPECmarks then the Opteron roundly trashes the Xeon... and it definitely holds its own against Itanium2.

SPECint2000 for a single CPU system (x44) is 163% of the Xeon result. SPECfp2000 is 111%.

For dual CPU systems, the x44 SPECint is 115% of Xeon and SPECfp is 193%. For quad systems the numbers go to 139% and 243%, respectively. The charts on the AMD website are a bit weird here, since they use the dual Xeon system as a baseline.

Of course, there are lies, damned lies, and benchmarks. SPECmarks hardly show the whole story. But by any measure the Opteron's price::performance is astounding... even without considering the 64-bit capabilities. Consider that this is a tenth of the price of the Itanium2 for 95% (or more) of the performance.

Not quite a fair comparison

[Hektor_Troy]

The dual Xeon has 512 MB RAM.
The dual Opteron has 2 GB RAM.

Pretty sloppy, if you ask me.

Re:Not quite a fair comparison

[dreamchaser]

You expected an impartial comparison from THG? You must be new here.

Re:Not quite a fair comparison

[Zathrus]

So what? MHz isn't everything. AMD is trying to make this abundantly clear with their performance ratings and talk like this is simply counterproductive.

If you want a "fair" benchmark then it should be a 2.8GHz Xeon vs the Opteron x44, both with the same amount of memory. A better benchmark, however, may be Itanium2 vs Opteron, but you can't run standard benchmarks on the I2 -- it's simply not designed for it. Oracle transaction ratings (albeit largely disk I/O dependant) and similar server benchmarks would be useful though.

Excluding the memory mismatch, however, it's a good idea to compare the Xeon 3.06 and the Opteron x44 -- they're the top end chips available and so the most likely for corporate shops to be choosing from. An alternate comparison would be similarly priced chips -- at current prices you'd be looking at the Xeon 2.8GHz.

Re:Not quite a fair comparison

[Perdo]

It gets worse, Dual channel was available for the Opteron, but not enabled. Also, and this one is not Tom's fault, the Opteron supports DDR400, but Tom used DDR333. The problem is the super limited supplies of DDR400 w/ECC,reg.

This is where the opteron with an 800mhz fsb with DDR333 ends up with less memory bandwidth than a Xeon with DDR266. The 533mhz bus Xeon used Dual Channel, giving it an effective 533 bus while the 800mhz bus Opteron was chokeing on 333mhz memory.

That is why the Opteron was falling down in the workstation benchmarks, because they tended to be bandwidth hogs.

Looking again, the opteron used 4 x 256 sticks of ram... 1 Gb not two.

Hrm...

[Anonymous Coward]

Nothing gets the blood pumping like a whole new generation of CPUs to compare numbers to, right?

Seeing a naked girl is really going to blow your mind.

Blocking /. referrers

[nutbar]

Tom seems to be blocking referrals from slashdot, so copy and paste this to make it view the article: http://www.tomshardware.com/cpu/20030422/index.htm l

Comment removed

[account_deleted]

Comment removed based on user account deletion

server vs. workstation?

[Thorgal]

Well, these benchmarks are supposedly suggesting that Opterons excell at server-type operations, while workstation performance is lacking. However, if you check their benchmark setups, there seems to be another way of looking at this: isn't is so that Opterons simply run better on Linux rather than Windows?

Re:server vs. workstation?

[double_u_b]

And more: Opteron needs specifically compiled binaries to show its full power. Not that I think it's a bad thing. But it shows that OpenSource solution will be more easily available for Opterons than comercial specifically compiled and optimised software, thanks gcc. I don't think Intel C++ compiler will support Opteron's new registers.

Too bad their web server isn't running on it

[MacroRex]

Slashdotted after 20 comments, that was quicker than usual.

Re:AMD is dead

[Gossy]

Considering how P4 3.06Ghz actually runs at 3Ghz and does much better at .....

Tell me, why is the fact it actually runs at 3Ghz important?

MHz is not a useful measure of performance.

Jesus. No wonder AMD implemented their 'marketing MHz' rating system - the average guy on the street thinks that's how you measure perfomance of CPUs, and even some /. readers seem to as well. Naive of me perhaps to think that /. readers would be more clued up, but hey.

I'd love to see the MHz rating be completely scrapped from how we rate CPUs in stores. Yes, it's useful to see that an AMD 2000+ is faster than a 1800+, but it's not so great when comparing with Intel chips. The trouble is that since AMDs are better at some things, and Intels better at others, a number of figures would have to be provided to make a fair and useful comparison. Too many numbers though I'm sure might confuse people, so I guess we'll be stuck with the MHz wars for a while yet.

Re:AMD Faster Speed markings?

[sprouty76]

Actually, I read that the numbers were based on the equivalent performance of an old Thunderbird Athlon (non-XP), rather than pentiums.

However, there's little doubt that they are meant to be compared to pentiums, and you raise an interesting point. Even stranger would be - what happened if intel adopted the same scheme? Then they'd both basically be making up numbers!

German version of the review

[dastrike]

The German version of the review seems to be quite a lot faster now than the English one: http://www.de.tomshardware.com/cpu/20030422/index. html

Memory-bandwidth?

[10Ghz]

Why does P4 with 2x64bit memory-bus get so much better results than opteron with 2x64bit memory-bus? One would think that since the mem-controller is integrated on the Opteron, it would get better results. Also, since each CPU has it's own memory-bank but they can still use other CPU's memory as well, the bandwidth should go up as number of CPU's increase. But still, P4 has more bandwidth than 2x Opterons! How can that be? IS there something wrong with the chip Tom benchmarked?

Re:Memory-bandwidth?

[Perdo]

Dual channel was available but not enabled.

The opteron uses an 800mhz memory bus.

But was chokeing on single channel DDR333

The Xeon was running Dual channel DDR266 or 533mhz effective.

Vast oversight (Intentional?) on Tom's part.

The Xbit labs clawhammer article shows the memory controller pushes at 97% of DDR400 theoretical maximum.

Now you know why all the "workstation apps" ran so poorly. They were all bandwidth intensive and Tom's ran the Opteron crippled.

Re:Memory-bandwidth?

[Perdo]

Ace's hardware has an article up. Their benchmarks are showing Opteron beating Xeon by 20% in 3DSmax while Tom's has Xeon beating opteron by 25% in the same test.

Right-O, toss out tom with the rest of the paid for rabble and move on to less biased sites.

Another "Editorial Content Sponsorship" from tom.

Re:Memory-bandwidth?

[10Ghz]

You're still limited by the controller speed and the memory speed, both of which are considerably lower in Opteron than P4/Xeon.

You are wrong. Opteron can use DDR333 or DDR400. Same memories P4/Xeon use (that's what AMD has said, DDR400 is just not officially supported. It does seem that Athlon64 does fully support DDR400 as well). So there's exactly zero difference there. Opteron has 2x64bit memory-bus, same as P4. Again: zero difference between the two.

The fastest that the Opteron can read memory is 333 MHz. Period. End of story.

You can use DDR400 just fine. Period. End of story. And besides, fastest memory you can use on P4 is 400MHz, and the difference between 333Mhz and 400Mhz isn't that big.

The Xeon can read 2x64 533 million times per second.

You are (again) confusing FSB-speed with the speed of the RAM. Yes, the FSB on Xeon if 533Mhz. No, the RAM is NOT 533Mhz. The P4 that had best bandwidth-figures in Toms tests used DDR400.

Figured out yet why the Xeon has more bandwidth?

Yes. Second mem-channel was not enabled on the Opteron, whereas it was on P4. review at Aceshardware shows more realistic bandwidth-numbers.

Please, learn about this stuff before you start to "educate" others, OK?

Re:AMD is dead

[Perdo]

So Centrino running at 1.6 Ghz but outperforming the 2.8 Ghz Pentium 4 is invalid?

How about Itanium at 1.2 Ghz outperforming the Pentium 4 at 3.06?

Or how about the 3.0 Ghz Pentium 4 beating the 3.06 Pentium 4 in every benchmark?

Yeah, you are right, Centrino, Itanium and the 3.0 Ghz Pentium 4 are all P.O.S. They are all officially dead.

40 Watts

[avandesande]

Keep in mind this opteron only uses 40 Watts.

Re:40 Watts

[red_dragon]

40 watts! Just 40 miserable watts! How can I cook an egg with just 40 watts?!?

I won't be upgrading until I can cook a meal while playing UT.

Another review at Ace's Hardware

[lorax]

Ace's hardware as an in-depth review as well, and it isn't slashdotted.
http://www.aceshardware.com/read.jsp?id=55000251

Blood pumping

[Sloppy]

Wow, what an image: an alternative to water-cooled overclocking. Pray to Satan that your overclocked Opteron will work, sacrifice a goat, and cool the chip with the unholy-consecrated blood.

Jarvik brand coolant pumps, Hellfire thermal paste, copper tubing with simulated brimstone anodized finish. And as for the cosmetic aspects of the case-modding, the thematic possibilities are endless. Start with this: Horns!!!

Re:16TB? No.....

[hotchai]

Read the review ..

AMD did not implement the full 64-bit virtual address (neither does Itanium2). The Opteron has *only* 48-bit virtual address and 40-bit physical address. That means it can address upto 256TB of virtual space and 1TB physical space.

And yeah, 256TB ought be enough for everyone ;)