AMD Zambezi news, info, fans !

[Mechanical Man]

I just hope the NDA lifts on the 19th of this month. I think that's one point everyone in this thread can agree on.

We can agree that you want that.

We cant agree that all of us want that, I want it to lift now

[informal]

Launch should be imminent. We should know soon what's the deal with Zambezi. I'm still not expecting much,maybe slightly better than 2500K or on that level.

[metalazzo]

I read somewhere that end of nda is not before end of September.

[informal]

Well that's 2 weeks or so,pretty close in my book .

[Formula350]

Good find! Have we known about these tools and/or these pages? Would be nice to have access to all of them.

I think some people have known, but I for one have not and it is ALL new to me! I agree though, I was hoping the other pages worked too lol I'd be content with even only having gotten PSCheck haha I like the fact you have SO much control from Windows (hell, 1/2 that would be nice even from the BIOS!) since you are able to adjust power states, make new power states, and run tests. I have a feeling that running the "Random" test would be a very good way to test stability, either on it's own or combined with one of their system-load producing tools. If it can switch PStates under heavy load, I think that's a sign of stability

Is this the most frustrating CPU release of all time or what? So many people are getting twisted and out of shape waiting for it. I used to come check out this thread every few days, but now I seem to be checking it lots of times daily. I think a lot of us need to chill out and just go do something else until it's released. Having said that, I will probably be back in an hour .

While I've not really kept my nose in the wind with other CPU releases, I do suspect this is one of the more dramatic releases, at the very least only as far as the users are concerned heh And yea the rate of new posts have substantially increased! Where a month ago or so I'd only check back once or twice a day to read maybe 6 to 9 new posts, now just over night there are TWO PAGES!! I suppose we're not called Enthusiasts for nothing haha

[El Gappo]

While I've not really kept my nose in the wind with other CPU releases, I do suspect this is one of the more dramatic releases, at the very least only as far as the users are concerned heh And yea the rate of new posts have substantially increased! Where a month ago or so I'd only check back once or twice a day to read maybe 6 to 9 new posts, now just over night there are TWO PAGES!! I suppose we're not called Enthusiasts for nothing haha

Should see the thread on OCN, posted yesterday, now over 550 posts. Would probably be 1000+ but half of them were dragon ball Z over 8000 comments

[Manicdan]

I think some people have known, but I for one have not and it is ALL new to me! I agree though, I was hoping the other pages worked too lol I'd be content with even only having gotten PSCheck haha I like the fact you have SO much control from Windows (hell, 1/2 that would be nice even from the BIOS!) since you are able to adjust power states, make new power states, and run tests. I have a feeling that running the "Random" test would be a very good way to test stability, either on it's own or combined with one of their system-load producing tools. If it can switch PStates under heavy load, I think that's a sign of stability

holy crap i think i need to play around and maybe write up a review. i love messing with pstates and turbo settings.

[massman]

Should see the thread on OCN, posted yesterday, now over 550 posts. Would probably be 1000+ but half of them were dragon ball Z over 8000 comments

I hope you informed them it's supposed to be over 9000 ...

[informal]

I can tell you one thing folks. This level we are seeing from vrzone,coolaler was not the projected one 9 months ago. Something changed in the meantime and we now get "this". What changed is anyone's guess. The 2 benchmarks from vrzone ,3dmar06 cpu test and c11.5 are clear signs of this "change". 9 months ago,Thuban couldn't touch zambezi in both. Today Zambezi can't touch Thuban. Situation looks grim.

[FlawleZ]

I can tell you one thing folks. This level we are seeing from vrzone,coolaler was not the projected one 9 months ago. Something changed in the meantime and we now get "this". What changed is anyone's guess. The 2 benchmarks from vrzone ,3dmar06 cpu test and c11.5 are clear signs of this "change". 9 months ago,Thuban couldn't touch zambezi in both. Today Zambezi can't touch Thuban. Situation looks grim.

What's "this?"

[duron]

I hope you informed them it's supposed to be over 9000 ...

[informal]

What's "this?"

You have seen "it". The vrzone and now coolaler posted same results. "This" means slower than Thuban or barely matching Thuban. And I use 8150 for comparison,not 8120. Granted there are some cases where 8150 with this level of performance will still probably spank 1100T,but nothing like the projections from December last year showed. Where is the double-sized SMT capable FlexFP performance? Where is the HPC improvement that is quoted for server parts gone? Remember we are left to think that even in legacy code server parts will be ~50% faster with same clocks as previous MC cores. This equates to around 10% better per core performance in SIMD workloads. All Zambezi leaks thus far show no such thing . On the contrary they show very poor performance of this "powerful double-sized FPU".

[BobyTT]

Sooo when we can buy those buldozers?

[Andi64]

You have seen "it". The vrzone and now coolaler posted same results. "This" means slower than Thuban or barely matching Thuban. And I use 8150 for comparison,not 8120. Granted there are some cases where 8150 with this level of performance will still probably spank 1100T,but nothing like the projections from December last year showed. Where is the double-sized SMT capable FlexFP performance? Where is the HPC improvement that is quoted for server parts gone? Remember we are left to think that even in legacy code server parts will be ~50% faster with same clocks as previous MC cores. This equates to around 10% better per core performance in SIMD workloads. All Zambezi leaks thus far show no such thing . On the contrary they show very poor performance of this "powerful double-sized FPU".

I'm lost. Maybe because the thread is over 100 pages, lol. Which coolaler benchmarks are you talking about? Can you link them? Are you talking about those benchs with 760G (or some crappy chipset like that) or maybe I missed something?

The CB11.5 tests with that chipset were in the 5.5 points range o something like that, really low. But "he who shall not be named" got 6.9 points in CB11.5 with 8150 stock. Anyway, I don't like that score either but we shouldn't jump to conclusions.

[FlawleZ]

You have seen "it". The vrzone and now coolaler posted same results. "This" means slower than Thuban or barely matching Thuban. And I use 8150 for comparison,not 8120. Granted there are some cases where 8150 with this level of performance will still probably spank 1100T,but nothing like the projections from December last year showed. Where is the double-sized SMT capable FlexFP performance? Where is the HPC improvement that is quoted for server parts gone? Remember we are left to think that even in legacy code server parts will be ~50% faster with same clocks as previous MC cores. This equates to around 10% better per core performance in SIMD workloads. All Zambezi leaks thus far show no such thing . On the contrary they show very poor performance of this "powerful double-sized FPU".

I don't understand why you're still trying to draw a rational conclusion from yet again "just another leaked benchmark." The results are all over the place in terms of performance within the last 6 months. I thought we had decided that it makes the most sense to take these things with a giant bowl of salt until we have official results once the NDA is lifted?

[FlawleZ]

I'm lost. Maybe because the thread is over 100 pages, lol. Which coolaler benchmarks are you talking about? Can you link them? Are you talking about those benchs with 760G (or some crappy chipset like that) or maybe I missed something?

The CB11.5 tests with that chipset were in the 5.5 points range o something like that, really low. But "he who shall not be named" got 6.9 points in CB11.5 with 8150 stock. Anyway, I don't like that score either but we shouldn't jump to conclusions.

I do believe he's referring to these:

http://translate.google.com.my/trans...p%3Ft%3D273646

[informal]

I don't understand why you're still trying to draw a rational conclusion from yet again "just another leaked benchmark." The results are all over the place in terms of performance within the last 6 months. I thought we had decided that it makes the most sense to take these things with a giant bowl of salt until we have official results once the NDA is lifted?

Well you probably read it,but this level of performance is in reviewers guide(I hope you know what that is). 9 months ago, in some docs 8C Zambezi at 3.5-4Ghz was shown as 70% faster than 1100T. There is your discrepancy. Now 1100T is actually looking better from price/perf. POV.

[FlawleZ]

Well you probably read it,but this level of performance is in reviewers guide(I hope you know what that is). 9 months ago, in some docs 8C Zambezi at 3.5-4Ghz was shown as 70% faster than 1100T. There is your discrepancy. Now 1100T is actually looking better from price/perf. POV.

Can you explain why CB would report 8120 and CPU-Z would report 8130? I don't think these benchmarks are legitimate.

[undone]

Well you probably read it,but this level of performance is in reviewers guide(I hope you know what that is).

What's 'reviewers guide'? If you're telling us that coolaler is something special, why they even have no NDA in hand.(NDA is not lifted)

[TESKATLIPOKA]

FlawleZ IT's funny how the lattest CPU-Z is still showing wrong info like TDP 174W but the core voltage is just 1.164V.

[informal]

What's 'reviewers guide'? If you're telling us that coolaler is something special, why they even have no NDA in hand.(NDA is not lifted)

Have you seen Radeon 6870/69xx launch slides? The performance comparisons(versus competing products from competitors and their own previous gen products),uarchitecture details,pricing strategy ,all made by manufacturer and sent to press as briefing material and guide for their own review .

And no,coolaler is nothing "special". Nor is vrzone for that matter. I'm just telling you that the level 8120 is showing in the leaks is close to what it will be showing when it launches. And this level is pretty dismal ,at least compared to all marketing talk we had and previous projections(by AMD themselves) that showed MUCH higher performance.
Bulldozer is supposed to be "multithreaded beast" . In benchmarks that utilize all the threads(or cores if you like),like cinebench for example,there is no sign of the beast. It's tamed. The "double-sized"(compared to K10) and "very-powerful" 256bit FPU ,even in single thread workload such as Cinebench10's single core test mode,is failing against Thuban's one core. It has even higher Turbo on its disposal in that case...

[Andi64]

Can you explain why CB would report 8120 and CPU-Z would report 8130? I don't think these benchmarks are legitimate.

CB is reading the "Specification" string wich is the same in CPU-z, not the "Name". I think "Name" in CPU-z is gathered from a Database, and "Specification" is read from the CPU itself.

[undone]

Have you seen Radeon 6870/69xx launch slides?

I've forgot them.

BTW it seems, that you have something in common with those guys who satirized AMD running a platform with slides instead of a CPU.

[Mechanical Man]

Have you seen Radeon 6870/69xx launch slides? The performance comparisons(versus competing products from competitors and their own previous gen products),uarchitecture details,pricing strategy ,all made by manufacturer and sent to press as briefing material and guide for their own review .

And no,coolaler is nothing "special". Nor is vrzone for that matter. I'm just telling you that the level 8120 is showing in the leaks is close to what it will be showing when it launches. And this level is pretty dismal ,at least compared to all marketing talk we had and previous projections(by AMD themselves) that showed MUCH higher performance.
Bulldozer is supposed to be "multithreaded beast" . In benchmarks that utilize all the threads(or cores if you like),like cinebench for example,there is no sign of the beast. It's tamed. The "double-sized"(compared to K10) and "very-powerful" 256bit FPU ,even in single thread workload such as Cinebench10's single core test mode,is failing against Thuban's one core. It has even higher Turbo on its disposal in that case...

I dont buy AMD giving any material to reviewers until its ~week to launch. That is based on what Sampsa said about it.

Then my guess would be that they got material 12th and launch indeed is 19th.

[repman]

I can tell you one thing folks. This level we are seeing from vrzone,coolaler was not the projected one 9 months ago. Something changed in the meantime and we now get "this". What changed is anyone's guess. The 2 benchmarks from vrzone ,3dmar06 cpu test and c11.5 are clear signs of this "change". 9 months ago,Thuban couldn't touch zambezi in both. Today Zambezi can't touch Thuban. Situation looks grim.

I found a few posts by JF which you might find interesting, I think he was commenting the sandra benchmarks:

I have no idea if they are even real, but I have good reason to believe that they are not representative. That is all I can say.

First, how does anyone know if the results are real? How do we know if they even had an actual processor? And if they did have a processor, how do we know what level of silicon they had?

Secondly, there are some reasons why even if they did have a chip, I can feel pretty confident that those numbers don't represent actual shipping silicon. Can't get into that right now, but given a little time you'll see why.

The real problem is I can't say the results are fake because nobody on this board can say that. Only the person that posted them can say that. But, just stop for a second and think what if they are real. Do you know if the person who ran them optimized the BIOS settings? Do you know if they specifically had the latest BIOS, the latest drivers, etc.?

There is better probability that there is something wrong than not.

I said that the benchmarks are "not representative", not inaccurate.

We just hit production on server processors. So are those benchmarks on final silicon? Highly unlikely. Are they using a final BIOS? Final drivers? OS optimizations? There are a ton of things that all need to be in place in order to get the best performance.

If I posted a benchmark showing the my product against an intel product but refused to give you the detail on the config, would you trust me? Probably not. But you would trust me if I gave you my score and compared to the best intel score that they ran, right?

There have been so many fake benchmarks, including those that were published before the first silicon hit our test labs.

Up until we hit production, none of the benchmarks could be considered representative. And, the only thing that we have publicly said is production on server (I cannot confirm if zambezi is in production or not - that is not my job.)

Once they are in production, even then, there are lots of variables that still have to be considered. So there are plenty that are not representative out there. Showing both lower and higher performance. Both sides of the coin seem to want to push an agenda. I recommend not trusting anything until launch.

If AMD does anything prior to launch, I would gladly confirm it and most likely would have a blog post as well.

These are from [H] and were posted in august, but there are quite a few hints there.

I think everyone should relax, there are so many fake benchmarks flying around no one can tell which one (if any) is true.

We do however know that IPC increases and clock is increased shouldn't that result in better performance?