AMD Zambezi news, info, fans !

[flyck]

How do you know that it's the CPU and not the GPU that consumes the power budget? Besides, the silicon and design could be limited by the compatibility with the GPU.

First indication:

AMD A8-3850
Radeon HD 6550D
100W
4
2.9GHz
400
600MHz


AMD A8-3800
Radeon HD 6550D
65W
4
2.4/2.7GHz
400
600MHz


TDP significant lower while the only constant is gpu resources and clockspeeds.

second indications:

Reviews A8 3850. They include power figures. it is trading power draw blows with similar clocked 45nm x4 parts.
http://www.anandtech.com/show/4476/amd-a83850-review/9
POwer draw figure cpu only is right where it is for gaming.

A Phenom II X6 at 32nm would be almost half the size of a Thuban if caches scales as good as cores when shrinking the process. BD must significantly outperform Thuban to justify this change in architecture.

Like i mentionned, that is the ideal situation, but the same is true for BD. Ideally on a good process BD would be far better than what they can do now with the current issues. Your hypotetical BD replacement with thuban cores in llano form would simply not do what on this process node what you would expect them to do.

[Dimitriman]

snip

Is the CPU PIB bundled with the Corsair H80 or similar cooling?

[Leeghoofd]

No cooling in my kit... Gotmail today that apparently Corsair is sending some 2 X4 GB dimms to the current reviewers (2000c9) However dunno if I still have the kit here when the rams arrive lol :p AMD in fact gave me one day at start, I begged on the phone to keep it over the weekend... time for Bad Company 2...

[Dimitriman]

No cooling in my kit... Gotmail today that apparently Corsair is sending some 2 X4 GB dimms to the current reviewers (2000c9) However dunno if I still have the kit here when the rams arrive lol :p AMD in fact gave me one day at start, I begged on the phone to keep it over the weekend... time for Bad Company 2...

Thanks. there are a lot of rumours that it comes with a self contained water cooler, hope those are true it would be a decent bundle for the "Black Edition" version.

[BeepBeep2]

i mean all Thubans can do 4.2-4.3Ghz 24/7 on 45nm,

My Thuban, a 1100T, does just over 4.1 on water.

Not sure where you pull your facts from.

[radaja]

Snip

are you saying AMD at first wanted you to review it and send it back the next day?

[I.nfraR.ed]

My Thuban, a 1100T, does just over 4.1 on water.

Not sure where you pull your facts from.

Completely agree:

View Full-Sized

[Leeghoofd]

are you saying AMD at first wanted you to review it and send it back the next day?

Not many samples in Benelux region mate, same for motherboards, GPUs etc... this time we got the cpu first, usually we are 5th in line...

[Canis-X]

How are you feeling about the chip so far....good ....bad....undecided?

[charged3800z24]

Completely agree:

[]Edit chart out
[]

It hasn't been updated it appears, I can do 4.2 easy on mine Jeck the last post in that thread. That still doesn't mean all can do 4.2, but believe most can.

[Dimitriman]

Snip

Tx for the comments, careful not to breach any NDA :P

But I´m left wondering when is AMD finally going to learn their lesson and work on high IPC. You'd think that Phenom II getting spanked in IPC was bad but having the new processor doing even less IPC is a slap in the face.

Sure the future is multithreading, but 8x high IPC should be better than 8x low IPC + high clocks.

[dess]

@xdan:
I have a feeling you're laying the foundations of your opinion solely on floating-point performance, more exactly existing, unoptimized scalar FP code (Cinebench). Didn't you noticed FP performance in not indicative of the INT one - the latter being much better? Also, don't you think even FP performance can become better with newer code, utilizing f.ex. FMA?

Regarding K10.x@32nm: you can put two more cores in, right, but you can't lift up the clocks enough (because it's very uarch dependant) to compete with SB core-for-core... It's a dead-end. With BD you can, even if it needs high clocks and so perhaps more power. Power efficiency can be bettered with further ramping and/or newer steppings and revisions.

[informal]

I'm with you Dimitriman .. I can't understand the logic they used to get where they are now. They knew ,back in 2008,with what they are dealing with from the other side. They could even make projections with what will intel come up come 2010/2011. With all this foreknowledge they produce chip that on paper has 33% more cores,but since each core is weaker,especially in SIMD workloads (which are crucial aspect of modern day CPU performance),it can't outperform the old design by more than 20%. Looking at highly MT scores,Thuban is just behind it in handbrake(video transcode) and cinebench(3d rendering)-and these 2 are supposed to be strongholds of Bulldozer design. As for ST performance,they talked about super prefetch and other things that supposed to offset the deeper pipeline and they talked about whole flexfp executing 256bit for single thread code etc. None of these things help Bulldozer outperform a 2007 core(K10) that got a very mild facelift in 2008/09 (in Deneb iteration). It's pretty disappointing.

Now there is this clock speed headroom which is a definite plus of the design,but users will run in the wall around 4.8Ghz on air. This is 17% faster than average Thuban OC so in best case you are just matching Thuban's "pure speed"(aka IPC) in apps like games (which won't support 8 threads anytime soon so core count advantage won't kick in) and poorly threaded workloads which are just the majority of desktop application types. Where you can actually use 8 cores/threads ,then this 4.8Ghz OC vs Thuban's 6C running at 4-4.1Ghz will probably be around 33% faster,but this is just in a few applications .

[xVeinx]

Wasn't what I was hoping to hear...

[pumero]

Thanks. there are a lot of rumours that it comes with a self contained water cooler, hope those are true it would be a decent bundle for the "Black Edition" version.

Way too expensive for AMD or only with a special bundle that's more expensive than the "standard" FX boxed version without it.
Only Intel could afford it to include such a water cooling solution with their Extreme Editions without additional costs for the end user.


Regarding extremely short review times, I heard something similar from a friend.
AMD told him they might receive the CPUs "soon" and then send one to him "soon" but he can only have it for 3-4 days.
It's probably one week before launch and they work with "soon" and only give reviewers a minimum amount of time to create articles.

An in depth article, which would take more than a few days to make, released a couple of days or weeks later will only reach those who are really interested in such stuff but the broad mass won't read it, as it is either to technical for them or they will say "I've read a number of reviews already, why read another one".

In my opinion that's ridiculous and amateurish and one could get the impression that this is done on purpose, to make it impossible to make a real in depth review (which outlines the weaknesses of the processor) right in time for the products launch, when most of these articles will be read.

However, it's in line with the (imo) failed marketing strategy for BD, saying it's a 8-core cpu. They create high expectations which they can't meet and thus the overall reception by the generic user will be "what, 8 cores and not better than 4 core sandy bridge? what a piece of crap ..."

[BeepBeep2]

It hasn't been updated it appears, I can do 4.2 easy on mine Jeck the last post in that thread. That still doesn't mean all can do 4.2, but believe most can.

Well, then my Thuban is a peice of . It failed prime last night at 4130 1.51v.

Oh wait, I chew* told me that was just an excuse.

[PerryR]

Because an architecture of cpu's waited for 3-4 years, fails to beat the mainstream of Intel.

How long do you think Intel worked on SB? Were Intel supposed be twiddling their thumbs during BD's development? Also, I thought SB 2600k was considered high-end?

They are no threat to even 2010 Intel hexa cores and now it's soon 2012.

You mean 2011?

Because AMD remains again in the back.

Back of what? Are you saying Bulldozer should have annihilated SB?

Because they had the performance of SB from january or even earlier, they delay 3-4 months and they couldn't do anything to improve much more the performance to at least equal SB 2600K.

So it would have made a difference if they hadn't delayed for 4 months? I'm inclined to believe that it would have still been a "fail" in some peoples eyes.

[informal]

This was posted by Dresdenboy some time ago:
http://semiaccurate.com/forums/showp...&postcount=824

Ah, it's N1truX and he posted in my first language.. IIRC he works for some magazine. He had to obscure his informations because of a NDA. He saw BD and made photos (might result in some article) and he heard something interesting about BD's performance directly from John Taylor and other ppl. Further he wrote that if leaked benches are close to the truth (although they seem to be a bit too low to him) any results (esp. latest Sandra leak) might be explained somehow. But the problem is not with Bulldozer (core/whole processor?) but with something else. The solution won't be available until launch which will also mean bad first reviews (at least sub par of what a fixed system would be capable of).

If there is any truth to this then it might be OS scheduling problem or a platform(bios?) problem. But don't get your hopes high.

[PerryR]

Did Lee remove his comment? NDA violation? Anyway, it's a bit strange that this processor performs like this.

[xdan]

Socket 1366 with i7 980( 583$) and i7 980X and 990X(999$) are high end not 2600K.
i7 980X was launch on Q1 2010 http://ark.intel.com/products/47932/...GTs-Intel-QPI), so after 1 year and a half AMD doesn't even touch that performance in multithread.
BD with "8core" should have been between 2600K an i7 980 in multithread always and closer to the six core Intel than to 2600K to be a very good cpu.

[devguy]

Whether or not AMD has some sort of hidden ace that they won't release until launch, I take exception to those who believe they can label a product a failure before it has been officially released and properly reviewed by reputable reviewers. Maybe it will be disappointing, maybe it won't. I'm not saying pre-launch speculation shouldn't be allowed (I think it should be encouraged), but pre-launch labeling is unfair. Just a few more days, and then feel free to give it a label.

[xVeinx]

This was posted by Dresdenboy some time ago:
http://semiaccurate.com/forums/showp...&postcount=824



If there is any truth to this then it might be OS scheduling problem or a platform(bios?) problem. But don't get your hopes high.

I get the BIOS update, but at the same time, that should be available to reviewers now. If people are going to publish results that aren't based on the right BIOS/kernel patch/whatever, that seems like a bit of a problem to me...

[informal]

I wouldn't bother launching it if this was the case. They (reviews) will generate a lot of negative buzz even if the mythological "fix" eventually arrives.

[PerryR]

I wouldn't bother launching it if this was the case. They (reviews) will generate a lot of negative buzz even if the mythological "fix" eventually arrives.

I was saying the same thing when this point was brought up initially. It doesn't make sense to release a flawed product to bad press.

I still say something doesn't add up, unless the FX-6100 ( I assume 4 cores 6 threads) I'm eyeing is only slightly better than a 1090t; but anyone expecting a SB killer is going to be "disappointed."

[Spectrobozo]

single threaded performance on CB10 looks to be around a 3.2GHz Conroe or something right?