Ace's new AthlonMP benchmark

[Kelledin]

<A HREF="http://www.aceshardware.com/read.jsp?id=45000255" target="_new">http://www.aceshardware.com/read.jsp?id=45000255</A>

A lot of people were waiting for this one...well, it's here.

No 2GHz Xeon, probably because of the extreme scarcity of the part.

Kelledin
[dave@discovery ~] kill -9 1
init: Just what do you think you're doing, Dave?

 

[mr_gobbledegook]

Athlon MP is a strong contender it beats the Xeon almost everytime ! Can't wait till Athlons support SSE2 what will Intel do then ?...they will probably have to resort to Hyperthreading.

<font color=purple>~* K6-2 @ 333MHz *~
I don't need a 'Gigahertz' chip to surf the web just yet ;-)</font color=purple>

 

[Guest]

The dual Xeon might not be a match for the dual MP in terms of performance, it still outsells it by a very, very wide margin. Not sure intel is so worried. They sell more, of a far more expensive cpu. What do they care if its slower, if the consumer doesnt care or know ?

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[bront]

Personaly, I would have liked to see how a dual XP performs compaired to a Dual MP. The t-bird 1.4 was pretty good, and seemed about right just on clockspeed to the MP 1200 most of the time, but it lacked the SSE-1 instructions. A dual XP system is still a cheep alternitive to a dual MP system.

Chesnuts roasting on an open CPU
Bill Gates nipping at your wallet

 

[Guest]

>I would have liked to see how a dual XP performs >compaired to a Dual MP

Easy.. it would perform exactly the same.. it just might not have worked (though doubtfull.. havent seen any reports yet of XPs not working in SMP, but you'd never know). Remember these chips are identical, except the MP has closed L1 bridges (hence easy overclock) and are validated by AMD for use in SMP boards. Other than that, there is nothing that distinguishes an XP from an MP. So, same performance, period.

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[FUGGER]

"Each quarter about 400,000 workstations are shipped, and about 70-80% of them are Intel based workstations, while the remaining 20-30% are mostly Sun and HP RISC workstations. So AMD has a long way to go before it can compete with Intel in the workstation market like it does in the desktop market."

Most professionals will throw the machine back at you if they find out AMD is indide. Strong rejection for a long time to come. so yes Intel has a strong hold on the high end workstation and server market.

XEON is more targeted for servers and databases, but it can render and stuff fine. One thing none of these benchmarks show nor they tell you is that XEON has superior memory bandwidth and very large cache configurations (seems someone is cluess on how to exploit these features or declined, to make benchmarks AMD favorable it seems). But thats what they decided to do.

Ive have noticed across the net that all the Intel sites don't have a huge bone lodged in the ass about trying to prove superiority of the other platforms like AMD does. sure AMD has something to prove but the should start with thermal protection and go from there. everyone would benifit from thermal protection that buys AMD.

Simple analogy.

If Fords break down because they overheat (even brand new ones off the showroom floor), and the engines blow because of overheating. would you still buy a Ford?

Im sure the dumbasses would!

 

[Guest]

>If Fords break down because they overheat (even brand new
>ones off the showroom floor), and the engines blow
>because of overheating. would you still buy a Ford?

Yes I would, if they overheat because I built my own car, and forget to put oil in the engine, I dont think I'd blame Ford. If GM would have a car that would perform worse, cost more, but automatically shuts or slows down when it detects no oil in the carter, I couldnt care less. Also, to stick with the analogy, the Ford would have a system that shuts it down when it looses oil or water. It just wouldnt stand running without oil at all. Big deal.

>Most professionals will throw the machine back at you if
>they find out AMD is indide.

If the machine performs flawlessly, and performs better; how stupid would that be ? The thing is, its hard if not impossible to find a first tier OEM AMD workstation. Its not like they have a lot of choice, do they ? Its also obvious that intels strong arm tactics make OEM think twice about introducing an Athlon MP workstation. I really dont think that has a lot to do with thermal protection; I have yet to see a 3D pro open his cas, and remove a heatsink while rendering 3D max.

Also, I've seen a few rendering farms here switch over to dual MPs. I tell you one thing, they will throw their old P3 Xeons at you if you come too close. If the price /performance ratio of the cpu is not too important when buying just one or a few workstations (where things like graphic cards and monitors cost much more), believe me it sure makes a difference when you have 200+ rendering machines.

>One thing none of these benchmarks show nor they tell you
>is that XEON has superior memory bandwidth and very large
>cache configurations (seems someone is cluess on how to
>exploit these features or declined, to make benchmarks
>AMD favorable it seems).

Aces' test was about low cost workstations. Sub $3000 machines. That excludes big cache Xeon machines, and most Sun workstations. AMD doesnt have anything yet to compete in that area either, at least not price wise. I'd be as curious as you to find out how much these huge caches help in specific apps. Im guessing the Athlon would perform at least very competitive in +95% of the apps. Remember Anands database test ? A single 1200 MP smoked a dual Xeon 1.7. I dont think a bigger cache is going to make up for that.

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[AmdMELTDOWN]

LOL! what a joke, first they have some old misconfigured P4 systems then the biggest AMDmongrel on Aces was too scared to post real results, for fear of being eaten alive by the rabid mongrelpups running around sniffing each others butts.

regarding the skull head:

<i>"we found the Dual Athlon MP
1800+ to be almost 30% slower than the Dual Xeon 1.7 GHz in this test."</i>

Affirmative Action For AMD NOW! is their motto.

"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"

 

[lhgpoobaa]

ok then.
what should they use?

AMD - Anti Monopoly Devices
AMD - At Massive Discounts!

 

[Guest]

Since you like quoting, here is another one:

>The fastest AMD configuration in our test beats the Dual
>Xeon 1.7 GHz by 50%. Considering that the Xeon 1.7 GHz
>costs about 15% more, it is not an unfair comparison


= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[Guest]

I saw a comparison several weeks ago (at least) about exactly that. Some guys compared XP with MP on a Tyan mobo and there was no difference. They even went as far as trying one of each, and reported that it went ok.

I have to apologise for not having linkage, but after a search on google I've come up empty... It's out there somewhere though...

 

[Guest]

oh yeah, there it is...

<A HREF="http://www.gamepc.com/reviews/hardware_review.asp?review=xpmp&page=1&mscssid=&tp=" target="_new">http://www.gamepc.com/reviews/hardware_review.asp?review=xpmp&page=1&mscssid=&tp=</A>

 

[Kelledin]

LOL! what a joke, first they have some old misconfigured P4 systems

Hmmm, so pre-builts from Dell and Compaq come misconfigured?

Second place must really suck for you Intel lapdogs.

Kelledin
[dave@discovery ~] kill -9 1
init: Just what do you think you're doing, Dave?

 

[Intel_inside]

Here is why AMD still has alot of work to do in the workstation market:

Intel chips are trusted, stable, and well supported by all the major professional programs out there. Amd chips are notorious for being unstable with many motherboards, running very hot without special cooling, and burning to a crisp in fractions of a second. If you were going to purchase workstations for a bussiness and want it to stay productive, what chip would you choose?

<i>Mommy that dog is trying to jump over that other dog but he keeps not making it</i>

 

[Kelledin]

XEON is more targeted for servers and databases, but it canrender and stuff fine. One thing none of these benchmarks show nor they tell you is that XEON has superior memory bandwidth and very large cache configurations (seems someone is cluess on how to exploit these features or declined, to make benchmarks AMD favorable it seems).

Funny, the tests are actual real-world tests.

Your quote about the "large cache configurations" is especially hilarious, seeing as P4 Xeons don't have any more cache than standard P4's. :tongue:

Ive have noticed across the net that all the Intel sites don't have a huge bone lodged in the ass about trying to prove superiority of the other platforms like AMD does.

That's because they can't prove Intel's superiority anymore. Like you, they can't prove that the AMD platform is unstable, especially not with sites like Ace's and LinuxHardware.org giving it perfect stability ratings. There's also far too much proof of AMD trouncing Intel for performance. Intel used to be best for everything, but that's not the case anymore.

Intel has better memory bandwidth, but still takes second place in performance in most real-world situations. Apparently bandwidth isn't everything.

Intel has better thermal protection, but that only really helps people who can't install proper cooling.

Kelledin
[dave@discovery ~] kill -9 1
init: Just what do you think you're doing, Dave?

 

[balzi]

If only it were somehow possible to get through to guys like you

"Amd chips are notorious for being unstable with many motherboards, running very hot without special cooling, and burning to a crisp in fractions of a second."

NO they are NOT!!
absolutely Any place you go, and any objective people who KNOW what they're talking about will tell you otherwise.

the heat thing -> everyone, Intel and AMD owners, has a chance of running hot just because of room temps. It is not the fault of either manufacturer if users don't follow standard rules, buy approved HSF arrangements etc.

as for the burning in few seconds, and this is to Kelledin aswell, the lack of protection is a real issue. It doesn't matter who's fault it is.. but whoever is losing money because of it will be the ones to fix it. There's no point in blaming dumb users who install a heat-sink wrong... Intel guarded against that.. so if someone's wants to install their own stuff they might just go for the safety of a P4. Why try and change everybody else, by education or telling them they're stupid - let me do it properly.. AMD or their MB makers might be better off following Intel's road.

as to the trolls.
Why do you comment without backing up anything??
Do you really enjoy people abusing your blatant stupidity??
How much fun can you have stirring people??
Is it people like me that wonder about you that keep you going??
You're scary!!



I spilled coffee all over my wife's nighty... ...serves me right for wearing it?!?

 

[mr_gobbledegook]

Oh god where do I start ?! The biggest downfall of the Athlon MP is that it doesn't have thermal protection, I agree. Companies want solutions that are 'solid' and are impossible to burn up.
As for AMD processors / platforms not being stable that is just pure bull*hit. ATHLON MP are VERY stable.

<font color=blue>
"As we have been able to test both boards for months now, we can safely say that Tyan's dual Athlon boards are the most stable Athlon platforms on the earth. We have tested the samples that both AMD and Tyan have sent us, and several retail boards (Tyan Tiger MP) for more than three months now, and we have never seen one workstation benchmark fail or crash. That is remarkable, because we have not only benchmarked a wide variety of high-end software, but each test has also been repeated tens of times by now as we have been testing with 6 different OpenGL cards (more about OpenGL cards in the next article). That is why we believe that the current AMD760MP boards still deserve your attention."
</font color=blue>

As for performance Athlon MP are very competitive, database transactions rely on memory bandwidth (which the Athlon don't have) but the nature of SQL transactions/queries also rely on good branch prediction. In my opinion the branch prediction in Xeons really sucks.

And yes AMD will also be second place in the server market because it is a new comer in this area, and has not recieved the backing of industry, for god sakes they have only been in the MP business for what ?...2-3 years ?? thats nothing ! don't expect miracles. However what the Athlon MP does show is that it can compete performance wise. For a first attempt that is pretty damn good !

The sever market is really tough to get into with large players like SUN, Intel. However do not under estimate the potential AMD have we could see a little boy VS Goliath type battle next year.

AMD pleeeease sort out your thermal protection now ! Hammer will fail if it don't have decent thermal protection, its as simple as that.
AAHHHHH !

<font color=purple>~* K6-2 @ 333MHz *~
I don't need a 'Gigahertz' chip to surf the web just yet ;-)</font color=purple>

 

[Matisaro]

Funny, the tests are actual real-world tests.

Your quote about the "large cache configurations" is especially hilarious, seeing as P4 Xeons don't have any more cache than standard P4's.

Hey kelledin, remember when I said you post enlightened flames....that was one of the best ones I have seen yet.



"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!

 

[Matisaro]

Balzi

as for the burning in few seconds, and this is to Kelledin aswell, the lack of protection is a real issue. It doesn't matter who's fault it is.. but whoever is losing money because of it will be the ones to fix it. There's no point in blaming dumb users who install a heat-sink wrong... Intel guarded against that.. so if someone's wants to install their own stuff they might just go for the safety of a P4. Why try and change everybody else, by education or telling them they're stupid - let me do it properly.. AMD or their MB makers might be better off following Intel's road.

If a user is too dumb to install a heatsink, he will either a, have a burnt chip, or b have a p4 throttled all the time and not know it. Because he is a dumb computer user, how would he know his p4 is throttling, I dont see EITHER, result as being favorable. While its true the p4 wont burn up, it will be useless without a properly installed heatsink, if the thermal protection is to protect computer idiots, and p4's implementation will basically render a computer idiots p4 system a pos, who is the protection for?

The fact of the matter is,
A: When a heatsink is improperly installed there are consequences on BOTH sides.
B: the consequence on the amd side is worse than the intel side, BUT BOTH computers are pretty much useless, so it is a rather moot point.


However, amd's thermal diode on a motherboard with AMD SPECCED THERMAL CONTROL SYSTEMS, which will be coming soon, will survive EVERYTHING, and furthermore, the system will not run impared to spare a computer idiot, the chip will be fine, but the fact the system is not running will clue the person into the fact that somethings not right. As opposed to the system running fine, just incredibly slow.

The end result is both intel and amd's most current thermal protection systems are perfectly fine for everyone and it is NOT an issue to 99.99% of computer users.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!

 

[Kelledin]

Heh thanx, I thought that was pretty good myself.

Kinda sad that he has four (or was it eight?) xeons in his desktop and doesn't take the time to know the product...

Kelledin
[dave@discovery ~] kill -9 1
init: Just what do you think you're doing, Dave?

 

[AmdMELTDOWN]

"If a user is too dumb to install a heatsink, he will either a, have a burnt chip, or b have a p4 throttled all the time and not know it. Because he is a dumb computer user, how would he know his p4 is throttling, I dont see EITHER, result as being favorable. While its true the p4 wont burn up, it will be useless without a properly installed heatsink, if the thermal protection is to protect computer idiots, and p4's implementation will basically render a computer idiots p4 system a pos, who is the protection for?"

LOL! you're an idiot! with a P4 you will notice when it's throttled down it's like pressing the brake on a car, moron!

and depending on how stupid you really are you can get back up and runing within minutes!

with an AMD procesor you will wait weeks to get you RMA'd cpu back and if you're a real moron then you'll probably make a keychain out of it(wow! that's 80% of thgc) or order another one.

exactly how much did you save on your tbird?


"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"

 

[FUGGER]

Yep, its a P3 xeon kellidin.

Mad bad, I was confused with P3 XEON obviously (run it into the ground I expect nothing less from the gimps), if you will notice we removed all P4 XEON from product line months ago. The only Intel based machine we build atm is the Itanium.

I think this a veto to your trounce theory kelidin:

"Each quarter about 400,000 workstations are shipped, and about 70-80% of them are Intel based workstations, while the remaining 20-30% are mostly Sun and HP RISC workstations. So AMD has a long way to go before it can compete with Intel in the workstation market like it does in the desktop market."

AMD will not be accepted into many major OEMS due to the thermal protection. has nothing to do with installing a HSF.

"However, amd's thermal diode on a motherboard with AMD SPECCED THERMAL CONTROL SYSTEMS, which will be coming soon, will survive EVERYTHING, and furthermore, the system will not run impared to spare a computer idiot, the chip will be fine, but the fact the system is not running will clue the person into the fact that somethings not right. As opposed to the system running fine, just incredibly slow."

Matisaro where is your proof?

AMD cannot rely on halfass motherboard makers for thermal protection. This has always failed miserably.

Until AMD CPU can survive with a HSF removed, I dont think the acceptance ratio will budge.

"That's because they can't prove Intel's superiority anymore. Like you, they can't prove that the AMD platform is unstable, especially not with sites like Ace's and LinuxHardware.org giving it perfect stability ratings. There's also far too much proof of AMD trouncing Intel for performance. Intel used to be best for everything, but that's not the case anymore." - once again, read thru these forums every page has a AMD horror story/testimonial. so that makes your statement FALSE! If your machine doesnt run, how can you claim it has high performance?

 

[FatBurger]

"That's because they can't prove Intel's superiority anymore. Like you, they can't prove that the AMD platform is unstable, especially not with sites like Ace's and LinuxHardware.org giving it perfect stability ratings. There's also far too much proof of AMD trouncing Intel for performance. Intel used to be best for everything, but that's not the case anymore." - once again, read thru these forums every page has a AMD horror story/testimonial. so that makes your statement FALSE! If your machine doesnt run, how can you claim it has high performance?

I love how you change the subject when you know you're beaten

<font color=orange>Quarter</font color=orange> <font color=blue>Pounder</font color=blue> <font color=orange>Inside</font color=orange>

 

[Kelledin]

Because my machine <i>does</i> run. :tongue: Runs very fast, doesn't crash, and doesn't overheat.

Funny how out of all this, you still can't find one horror story about an AMD760MP. And when professionals like Ace's, Anandtech, and (ahem) myself can run an AMD system stable for months with zero maintenance, that really proves that the platform is solid as a rock.

Oh, btw, I'm still waiting for you to prove that official Intel/AMD datasheets are in error. :tongue:

Kelledin
[dave@discovery ~] kill -9 1
init: Just what do you think you're doing, Dave?

 

[FUGGER]

beaten? not hardly.

Those quotes are not mine, they are directly from link provided above. Not changing the subject, replying to posts/pointing out the obvious.

Bottom line, AMD has problems and with those said problems they will never get the respect they deserve. Maybe one day AMD will adopt a thermal protection that is on chip, not depending on the board mfg. but until then they will be rejected by professionals and OEM's. Stability is another cause for alarm, Im sure you ignore all the "help posts" as if they never existed. but this forum is like too many others crammed full of "help me my AMD doesnt boot, locks up, reboots randomly, overheats, crashed and died, ...". But I guess you cannot seem to find those daily posts as problems associated with AMD processors.