AMD Bulldozer Eng. Sample leaked, benched

[twilyth]

article - http://wccftech.com/2011/06/10/amd-...benchmarked-tested-asus-sabertooth-990fx-am3/

 

[hellrazor]

That board looks like a beast from hell!

I'm not familiar with any of those benching programs, so I don't have anything else to say.

 

[Melvis]

Ill believe it when i see it, im calling fake on this sorry.

I cant see an 8 core CPU getting a worse score then the current X6 of today, doesn't add up.

 

[CrackerJack]

Ill believe it when i see it, im calling fake on this sorry.

I cant see an 8 core CPU getting a worse score then the current X6 of today, doesn't add up.

same here, something just isn't adding up... but who knows

 

[twilyth]

I have a few Bulldozer alerts set on Google and this just happened to hit my mailbox this morning. I'm just posting it for what it's worth, if anything. If you think I should change to title to make it less dramatic, suggest something and if I can still do it I will.

 

[Wile E]

If it is a real BD, it could just be a glitch with the program. Might be like the old dual cores and need a special driver to work properly in some multi-threaded apps.

I sure hope that's not the real performance, anyway. If it is, BD is a total and utter failure.

 

[twilyth]

If it is a real BD, it could just be a glitch with the program. Might be like the old dual cores and need a special driver to work properly in some multi-threaded apps.

I sure hope that's not the real performance, anyway. If it is, BD is a total and utter failure.

Please be more specific since I don't know much about benchmarks. thanks.

 

[CrackerJack]

Please be more specific since I don't know much about benchmarks. thanks.

Simple version: Those results were shit

 

[twilyth]

Simple version: Those results were shit

That's not helping me or anyone else understand.

 

[Wile E]

That's not helping me or anyone else understand.

Which part of my comment is unclear/needs explained?

 

[twilyth]

If it is a real BD, it could just be a glitch with the program. Might be like the old dual cores and need a special driver to work properly in some multi-threaded apps.

I sure hope that's not the real performance, anyway. If it is, BD is a total and utter failure.

Which part of my comment is unclear/needs explained?

The response you quoted was to CJ.

Why do you think the results if true mean BD is a failure. I mean I did specifically ask about benchmarks in relationship to your comment.

 

[Wile E]

The response you quoted was to CJ.

Irrelevant. I just quoted the last post on the general topic. You knew what I meant, and responded accordingly. That's all that matters.

Why do you think the results if true mean BD is a failure. I mean I did specifically ask about benchmarks in relationship to your comment.

You could've also been referring to my comment on possible glitches and the drivers some of the older AMD chips needed to perform properly in some programs. In case you missed it, I did make more than one point in my original comment.

If these benches are true, BD is a failure because it's IPC per core is terrible. Intel's last generation 4 cores with HT overpower it, let alone SandyBridge, or the upcoming skt2011 cpus.

 

[HalfAHertz]

My guess is that it flopped because of the "slow" memory they used. From what I've heard the BD memory controller is really something and to get the full performance you need to use 1866MHz sticks. Otherwise the memory controller quickly gets saturated because of all 8 cores accessing it and the performance drops exponentially.

 

[Wile E]

My guess is that it flopped because of the "slow" memory they used. From what I've heard the BD memory controller is really something and to get the full performance you need to use 1866MHz sticks. Otherwise the memory controller quickly gets saturated because of all 8 cores accessing it and the performance drops exponentially.

I can see that being a bottleneck, but I don't see it being one to this degree.

 

[twilyth]

Irrelevant. I just quoted the last post on the general topic. You knew what I meant, and responded accordingly. That's all that matters.



You could've also been referring to my comment on possible glitches and the drivers some of the older AMD chips needed to perform properly in some programs. In case you missed it, I did make more than one point in my original comment.

If these benches are true, BD is a failure because it's IPC per core is terrible. Intel's last generation 4 cores with HT overpower it, let alone SandyBridge, or the upcoming skt2011 cpus.

I was pointing something out I though you didn't realize. Not sure why that is a problem or "irrelevant", but whatever.

Where are you seeing IPC figures?

 

[InnocentCriminal]

I can't take anything from wccftech with any seriousness.

I'll hold out for some legitimate results.

 

[Steve-007UK]

I was reading the other day about how AMD were not happy with the performance of BD on the early sample chips, After reading that link in the OP my guess is that the CPU shown is one of the first (engineering samples) that AMD were not happy with.

My understanding is AMD are working on new stepping (B2 or something) that should be in the retail CPU's.

If what i have read is true then im also guessing that them benchmarks are not to be taking seriously as true scores, We just have to wait and see.

It also mentions the new stepping in the link.

 

[twilyth]

I was reading the other day about how AMD was not happy with the performance of BD on the early sample chips, After reading that link in the OP my guess is that the CPU shown is one of the first (eng samples) that AMD was not happy with.

My understanding is AMD are working on new stepping (B2 or something) that should be in the retail CPU's.

If what i have read is true then im also guessing that them benchmarks are not to be taking as true scores, We just have to wait and see.

It also mentions the new stepping in the link.

I read the same thing and I think you got that spot on. That's my understanding for why they pushed the launch from June to July-September. It does make sense.

Aren't these chips the first to use high-k gates? If so, manufacturing issues were bound to crop up.

 

[Velvet Wafer]

Temps of the chip were reported at 13.4C (Idle) and 16.2C (Load) but the user mentioned that it was a Glitch in the AMD Overdrive software as the temps reported in BIOS were hovering around 35-40C (Cool n Quiet Enabled).

So i was indeed right. The damn,worthless AOC is just bugged!

EDIT:
just saw it, there are vantage leaks out there too (also on this UK page).... these look better surely:
http://wccftech.com/2011/05/30/amd-bulldozer-fx8110-scores-81917-3dmark-vantage-cpu-test-details/
281 steps in CPU test 2 in Vantage is murderous in my opinion...i7 does not even 100,severely OCed

 

[hat]

I really hope those aren't true results...

 

[Red_Machine]

Also, Cool n Quiet was enabled, which lowers performance.

 

[repman244]

http://www.youtube.com/watch?v=d8QRKdyBzKQ

And everyone should be aware that these chips are B0 ES chips, it seems they all have severe restrictions hence the low scores. I mean come on, AMD would never release a chip that is slower than Thuban and call it FX.

 

[Velvet Wafer]

http://www.youtube.com/watch?v=d8QRKdyBzKQ

And everyone should be aware that these chips are B0 ES chips, it seems they all have severe restrictions hence the low scores. I mean come on, AMD would never release a chip that is slower than Thuban and call it FX.

i really liked that quote... seems to throw at least some information regarding the issue, even it can be false

B0 Stepping => Capped clock speed, broken turbo, memory bandwidth restricted...
The retail release is B2 Stepping. (B1 Stepping has speed scaling issue that needs to be corrected.)

 

[Fatal]

3.2 over clock? 1333 memory I would have tried to blow that thing up the scores are poor but I will wait for a legit review. I am starting to believe AMD is screwed it will be a miracle for them to touch Intel Sandy Bridge chip's performance.

 

[Nesters]

Why do you think the results if true mean BD is a failure

Scoring less than PII X6 and also costing more would be an epic failure. If BD wasn't quite an improvement, price cuts would have been smaller for existing processors.