Zambezi ES performance weirdness

[RussC]

And yet you bought a phenom II?

Why I own and use certain things really doesn't matter, Im looking more at an investment thesis right now.

Superpi is absolutely useless, it's a relic from nearly 30 years ago when we used to install our own co-processors.
AFAIK AMD has deprecated x87 in BD and it will be emulated in the fp pipe. Don't be surprised if 1m times are out over a minute.

Still think it's a great performance indicator?

Fine, Ok, pick the benchmark of your choosing. My statement still holds, the company with the faster SINGLE CORE/THREAD AND MEMORY/CACHE INFRASTRUCTURE WINS MARKET SHARE. So whatever test gives this matrix tells the larger story.

RussC

[SkullCracka]

Btw, it is Llano L-L-A-N-O, not "LIano". Also it is not bulldozer, so it is not appropriate for this topic..
But yes, nevertheless it looks good!

well allrighty then. sorry about the SP.

but it is related to this topic since both of the latest samples to have been benched. have shown to be crippled in some way.

[Sparky]

you are all looking at bowlful of batter and arguing about whether or not it is appetizing.

ha, nice

[Manicdan]

ha, nice

a bowl of cake or brownie batter or cookie doe and many other baked sweets are really good before they go in the oven. however theres also the problem where some moms dont know how to set the timer and ends up ruining everything that started off tasty. so its a perfect analogy, lol

[JF-AMD]

Actually, I do work in the industry, and I believe one of our etchers is in one of your fabs.

With that said, You don't mean that the engineers that designed the processor dont have some idea of performance before final silicon? Um, that makes no sense to me! So AMD has no idea of where the BD cores stack up against competitors before final silicon, I have a hard time believing that.

RussC

We have plenty of data on where we believe the performance will be. That is like having a picture of a cake. But when you are mixing the batter and looking at a picture of cake, you have a ways to go.

People getting all wrapped up about engineering samples are really obsessing on the wrong points.

All of this talk about engineering samples is really annnoying because they are not designed around performance. So trying to measure something for performance when it is designed for validation is really pointless.

[Kobaltrock]

All of this talk about engineering samples is really annnoying because they are not designed around performance. So trying to measure something for performance when it is designed for validation is really pointless.

One could say it is more annoying that AMD isn't doing anything to quash the results of ES, if the ES samples are so bad performance wise.
People want info, and AMD isn't giving it to them, so, they get hard numbers from ES, and think that is the best AMD can do.

Marketing 101, if you see your #1 product getting butchered, it is time to trickle out more info to combat the misinformation on the purpose of ES samples.
It would also be nice to know what the current stepping of BD is, compared to the ES samples.

[dave_graham]

Kobaltrock,

hate to break it to you but 99% of AMD's revenue doesn't come from overclocking folks or categorical "enthusiasts." It comes from integrators (HP, Dell, etc.) who, as JF-AMD noted above, test based on qualification, etc, not on supposed gross performance. edge-cases always emerge (e.g. the development of a particular line of systems) where performance matters but you'll typically note those are introduced AFTER mainstream parts...

so, AMD doesn't need to waste marketing dollars on folks who build use cases around stuff that XS tests....they need to spend the time on revenuing systems and integrators.

cheers,

Dave

[informal]

One could say it is more annoying that AMD isn't doing anything to quash the results of ES, if the ES samples are so bad performance wise.

FYI,gimped B0 ES videos on youtube are gone.So they are doing something about it.

[Solus Corvus]

The SuperPi and single thead tests are not useless. As I and others have posted, the company that wins the single thread/core and internal memory/cache performance, gets the spoils, plain and simple. Again, either company can stack cores till the cows come home and play the leap frog game of performace king for a year till the next higher core count part comes out(the same way AMD and NVDA are fighting it out). But the company with the better core/thread performance wins market share, sorry.

RussC

I don't agree with this. It could very easily be the case that Intel wins at single thread and low thread counts and AMD wins at high thread counts. They can both stack chips, but that doesn't mean that they can both stack chips equally. A shared design will be able to stack significantly more chips before running into thermal, power, or die space limitations.

As for market share, chips with different strengths will capture different segments of the market. A strong multicore design could capture market share in the critical high-margin server market. It would also be quite useful for media/rendering/encoding workstations and scientific computing. Certain power users that use heavy multitasking and heavy mutithreaded apps (like myself) would prefer the extra cores. While Intel would continue to capture more market in the low-threaded segment such as gamers, certain benchers, and office/home users.

I don't really see a problem with this. Having choices is good. No one processor is going to effectively process all types of code.

[Manicdan]

I don't really see a problem with this. Having choices is good. No one processor is going to effectively process all types of code.

xtremely true. back when all you had was one core and a set mhz/ghz, there was only 1 task, be as fast as possible

today we have:
very low thread count with high speeds designed for lowest latency possible
typical home use with medium number of threads which are great for the mainstream applications
power users who go with high thread counts and high speeds
efficiency chips that run slower for 24/7 100% load but offer better perf/watt

thanks to turbo and power gating, you can blur the lines a little, but no chip yet has the ability to offer all of those by adjusting a setting in the bios

[El Gappo]

We have plenty of data on where we believe the performance will be. That is like having a picture of a cake. But when you are mixing the batter and looking at a picture of cake, you have a ways to go.

I think what JF is trying to say is "the cake is a lie" =D Especially the one OBR is regurgitating.

xtremely true. back when all you had was one core and a set mhz/ghz, there was only 1 task, be as fast as possible

today we have:
very low thread count with high speeds designed for lowest latency possible
typical home use with medium number of threads which are great for the mainstream applications
power users who go with high thread counts and high speeds
efficiency chips that run slower for 24/7 100% load but offer better perf/watt

thanks to turbo and power gating, you can blur the lines a little, but no chip yet has the ability to offer all of those by adjusting a setting in the bios

Word.

[RussC]

I don't agree with this. It could very easily be the case that Intel wins at single thread and low thread counts and AMD wins at high thread counts. They can both stack chips, but that doesn't mean that they can both stack chips equally. A shared design will be able to stack significantly more chips before running into thermal, power, or die space limitations.

As for market share, chips with different strengths will capture different segments of the market. A strong multicore design could capture market share in the critical high-margin server market. It would also be quite useful for media/rendering/encoding workstations and scientific computing. Certain power users that use heavy multitasking and heavy mutithreaded apps (like myself) would prefer the extra cores. While Intel would continue to capture more market in the low-threaded segment such as gamers, certain benchers, and office/home users.

I don't really see a problem with this. Having choices is good. No one processor is going to effectively process all types of code.

What, that argument doesn't ring true, and certainly hasn't happened at all in the market. So far in the multi-core world, the fastest wins market share. AMD will never make a dent in INTC 85%+ market share with what your talking about and will be religated to the price/performance end of the curve. AMD wont, hasn't and has lost significant market share for nearly 7 years being second fiddle with slower parts. So far, AMD has lost on all those market segments you talk about. Again, they will never take back any market share with what I see so far.

With that said, we'll wait and see what BD really is to gauge where AMD goes from here.

RussC

[Sn0wm@n]

What, that argument doesn't ring true, and certainly hasn't happened at all in the market. So far in the multi-core world, the fastest wins market share. AMD will never make a dent in INTC 85%+ market share with what your talking about and will be religated to the price/performance end of the curve. AMD wont, hasn't and has lost significant market share for nearly 7 years being second fiddle with slower parts. So far, AMD has lost on all those market segments you talk about. Again, they will never take back any market share with what I see so far.

With that said, we'll wait and see what BD really is to gauge where AMD goes from here.

RussC

please remind me how much money intel makes a year compared to amd ....


and btw if you believe intel's product are so great ... why dont you buy their product instead ... or their stock

[charged3800z24]

What, that argument doesn't ring true, and certainly hasn't happened at all in the market. So far in the multi-core world, the fastest wins market share. AMD will never make a dent in INTC 85%+ market share with what your talking about and will be religated to the price/performance end of the curve. AMD wont, hasn't and has lost significant market share for nearly 7 years being second fiddle with slower parts. So far, AMD has lost on all those market segments you talk about. Again, they will never take back any market share with what I see so far.

With that said, we'll wait and see what BD really is to gauge where AMD goes from here.

RussC

Sad fact is, AMD is having more issues gaining Market share due to other factors. Like Software Company requirements that say "Intel blah blah or higher" on the requiremnet specs. I never see AMD bosted on the top of the Specs sheets. You sometimes see "or Equivalent" which basically means the competitors option (AMD). And companies like mine where we blindly used Intel regarless of the perfomrnces at any price point. Even when AMD had the clearer Choice all around, They couldn't cut the market share due to these factors.

[TESKATLIPOKA]

Guys I don't know If someone posted it or not but a turbo slide is on the net although it doesn't say what turbo frequency will be used for just one core(not module) but still it looks great.

http://www.techpowerup.com/147408/FX...-1.00-GHz.html

[Solus Corvus]

xtremely true. back when all you had was one core and a set hz, there was only 1 task, be as fast as possible

Fixed that for you.

The problem of parallelism has been around a long time. CPU manufacturers had to decide if they were going to keep their chips scalar and just make those execution units faster or if they were going superscalar and add redundant execution units. Obviously the later course has won historically for general purpose cpus. But they can't go wild adding tons of redundant units because there are limitations and you get diminishing returns. Sometimes there simply is only so much instruction level parallelism (ILP) to be extracted from any given segment of code. It also takes quite a bit of extra silicon to calculate dependencies and it increases rapidly with the number of extra units.

Instead of fighting a game of diminishing returns with Intel to extract maximum ILP, it seems that AMD is simply refocusing on thread level parallelism. Even Intel with their renowned branch prediction, OoO engine, hyperthreading, etc still has execution units idle for a significant portion of time. So AMD removed one of the ALUs and all of the expensive logic required to check dependencies. They can use that freed transistor/power budget on branch prediction, cache, etc to keep the remaining units fed. IPC could still increase despite being less superscalar than their previous arch and the upshot is that the previously redundant execution units are freed to work on separate threads.

I think both AMD and Intel are working towards eager execution. In essence, instead of trying to predict which path to take when you come to a branch, take both. IMO, this is the next leap both companies need to take in single threaded performance. Bulldozer would give AMD a significant lead on developing this, IMO.

What, that argument doesn't ring true, and certainly hasn't happened at all in the market. So far in the multi-core world, the fastest wins market share. AMD will never make a dent in INTC 85%+ market share with what your talking about and will be religated to the price/performance end of the curve. AMD wont, hasn't and has lost significant market share for nearly 7 years being second fiddle with slower parts. So far, AMD has lost on all those market segments you talk about. Again, they will never take back any market share with what I see so far.

With that said, we'll wait and see what BD really is to gauge where AMD goes from here.

RussC

For one thing, AMD didn't make much of a dent in Intel market share even when they did have the faster chip in both single and multi threaded apps. So there is indeed more going on than simply who has the faster chip for any given application.

Of course my scenario doesn't ring true, because it hasn't been the case in recent memory. Before conroe AMD had the stronger performance in single and multi-threaded performance. And since conroe AMD hasn't had the faster chip for single or multi-threaded performance. And in that time they lost the small market share gains they did make. But if they do have the better multi-threaded performance I expect that they will regain ground in the segments I mentioned, or at least the server segment. Single thread performance doesn't mean much of anything in the server world. What matters is total throughput and power consumption while doing it. Why do you think the market was salivating at the thought of clustered ARM chips for server applications?

[jazkat]

hi guys i dont post here much but have been sneeking around the BD threads and this delay has got me wondering..... this is just a thought btw, at the moment i dont know what is going on.... its delayed , its not delayed , needs a respin for a higher clock blah! blah! blah!

do you guys think there is a possibilty that amd delayed zambezi because the enhanced/next gen BD cores trinity uses offers better performance? they have clearly been testing trinity for a while so would it possible for this new core/cores to be a replacement for the first gen ones but obviously without the gpu part?

i may be mad but it was just a thought

[XRL8]

Few things.
Some people suggested that BIOS maybe the issue.
Well, boards are final and shipping, nobody in their right mind would ship a board to OEMs and even retail, with a claim about AM3+ compatibility if they would REQUIRE a bios flash after the client gets it.
JF insists that they dont cripple ES.
JF u sure ? Cause that doesnt tell a good story.
Because well, if bios aint broken, and the chip isnt.Its just BAD.
I know, u take as many chips as possible to get ES`s ,so clock wise, they can be very different from final product.
But they should perform clock for clock rather same.
Up until now, all ES`s ive seen had clock for clock very similar performance as the final product.
There were many issues, like low clocks, high volts, high TDP etc etc.But performance wasnt that drastically different.
And you say that AMD has only some basic idea about CPU performance before production ?!
Dont you guys simulate all the logic on these fancy hardware simulators big as hell which can simulate any processor albeit at MUCH lower speeds ?
And shouldnt you know your IPC targets for different workloads when the cpu is still on paper ?
Youre claims are kind of alarming.

@RussC

The thing is, in this day, pure performance doesnt really matter to the 80% of population.
"good enough" is the mantra for todays computing.
And even NOT good enough atom made a killing.
And now AMD makes a killing with zacate.
Simple example, if i offer my girlfriend a new laptop with sandybridge, which will be too slow to play sims 3 without choking, shes not going to be happy.If Llano will do that, shes going to opt for that.
And it doesnt matter that some tasks that she rarely does are going to be slower.Because gaming NEEDS some kind of performance, but with encoding a movie clip she an wait a minute instead of 30secs.
Llano will do good too.Although it needs cpu part to be revamped, but its coming next year.
Only problems may be with desktop/server part as we dont have BD numbers yet.

@Informal

Are there any pics of B1 in action from obr ? i cant find it.

@Jazkat

Trinity uses BD v1 .So the same arch as coming BD for am3+.

[jazkat]

@ XLR8

well from everything i have been reading trinity uses a enhanced , next gen bulldozer core..... wouldnt that mean a different one to zambezi?

"Trinity will utilize x86 cores based on improved Bulldozer micro-architecture as well as DirectX 11-class graphics engine" ?

AMD Financial Analyst Day 2010, revealed the 2nd generation is to be scheduled for 2012. AMD currently refers to this as Enhanced Bulldozer.?


Trinity is based on our [enhanced] x86 bulldozer core?

[XRL8]

@Jazkat, seems youre right.

Ive read somewhere article aboud Llano and there was a part where they stated AMD is going to use in each generation cpu core from past strictly CPU generation cpus, like Llano does with stars currently.
Thats interesting that theyre accelerating that.

[jazkat]

@Jazkat, seems youre right.

Ive read somewhere article aboud Llano and there was a part where they stated AMD is going to use in each generation cpu core from past strictly CPU generation cpus, like Llano does with stars currently.
Thats interesting that theyre accelerating that.

yes this is why i had that thought, i mean if they have a working enhanced BD core already maybe this is the real reason for the zambezi delay.

im sure amd want all the performance they can get when bringing zambezi to market?

[2senile]

yes this is why i had that thought, i mean if they have a working enhanced BD core already maybe this is the real reason for the zambezi delay.

im sure amd want all the performance they can get when bringing zambezi to market?

I had a similar thought when I heard about the Trinity laptop demo . While I know nothing about the industry I would imagine that they would have had to start to "do the switch" months ago. ?????

[Solus Corvus]

No matter when BD is released, I would expect Trinity to use an enhanced version of those cores since it comes out later.

[JF-AMD]

A. I thought I already addressed ES on this forum, why are we still beating that dead horse?
B. I have said (in other places) that the delay is scheduling. Client moved from Q2-Q3 because of some re-scheduling, server is staying in Q3. If you want to read anything into that, you can see that there is not a huge amount of time difference between them, and the fact that server stays in Q3 should be an indicator that all of the rumors and speculation are massively overplayed.

[FlanK3r]

JF: I think, its because higher priority for AMD is now Llano (notebooks), so one or two months sheduled of BD )in desktop) is nothing...I believe BD will be good . 32nm capacity is maybe limited. When will be new FAB completed? Thx.