Amd Bulldozer tested [youtube]

[sunny7day]

B0 stepping only.
Please watch in HD mode, thanks.

Bench at 3.2Ghz:
http://www.youtube.com/watch?v=Z12e7oKPJd8

DDR3-2400Mhz:
http://www.youtube.com/watch?v=aS-QScAfdR4

Bulldozer vs 2500k at 4G:
http://www.youtube.com/watch?v=d8QRKdyBzKQ

Some hardware info and wPrime 32M result.
http://www.youtube.com/watch?v=uCZmIzfGbiI

[fellix_bg]

The overall cache performance in AIDA64 is quite abnormal. It's probably due to the unoptimized benchmark code.

[newls1]

not even a stable oc!!!! if this is real, im not impressed at all!

[eXa]

Can you please do cinebench R11.5???

[sbattosai]

B0 stepping -> FAIL

[informal]

It's the same chip they are testing in chiphell forum... Look at the L1/L2/L3 benchmarks,write test cannot even finish properly lol.

[XRL8]

Yeah, its B0.BD is going to get load of bad press before launch.Congratz AMD once again.

edit:
And whats up with timings in cpuz ,2-16-2 ?!

Also.
Note the cpuz screens.They show the infamous 186W and some other stuff people doubted.Its 1.5.7.1 tho ,and not 5.6 liek cpuz creator had.

[slaveondope]

Whats with the ram timings? 2-16-2-22-0 AMD surely gimped the hell out of these chips to keep everyone off track to true performance.

[radaja]

why is AMD letting these horrendous chips out in the wild?
whats the point of gimping them so bad?

[XRL8]

Well, no one knows about real performance, thats whats for.
However, people that arent calm enthusiasts, will think that AMD failed big time.Doube edge sword.

[informal]

Average Joe has not even heard about Zambezi or FX or Scorpius.They have no clue.Why do you think they will know about gimped BD ES on some Chinese forum or even youtube.
They will know about BD when Zambezi/FX/Scorpius launches.Then we will have non-gimped chips with good retail prices.That will average Joe like.

[SKYMTL]

The number of Bulldozer threads is beginning to get out of hand.

Honestly though, I wouldn't even bother paying attention to these early results since like I said in another thread: they'll give you a completely unbalanced opinion about the chip and its capabilities.

[BrowncoatGR]

why is AMD letting these horrendous chips out in the wild?
whats the point of gimping them so bad?

Well their purpose was probably to be used as validation chips by M/B OEMs so performance wasn't really a factor.

[jimbo75]

All publicity is good publicity.

[Manicdan]

The number of Bulldozer threads is beginning to get out of hand.

+1
this could have easily been added to the BD screenshots thread in the AMD section

[LordEC911]

AMD loves to sandbag.

[repman]

It's the same chip they are testing in chiphell forum... Look at the L1/L2/L3 benchmarks,write test cannot even finish properly lol.

I think you were right when you said that the frequency isn't read correctly, one of the comments: "appearently there is something wrong with the multiplier, can't tell if it is really running at 3.2G"

And also, could it be that the cache is disabled completely?

[BeepBeep2]

Guys, quick heads up, ram timings are 6-9-6 you can see that in the AIDA CPUID window.

[QuadDamage]

This doesn't look too promising to say the least.

[Particle]

I don't understand how anyone here could be dense enough to treat an early ES that is performing at a level obviously below where it should as any real indicator of final silicon performance. If it performed great it would tell us something. Since it doesn't, it tells us nothing.

[Lanek]

This doesn't look too promising to say the least.

They don't have to be promising .... Imagine what result could have been shown with a A1 revision of GTX480 .... lol.. ES is ES ...

[BeepBeep2]

I don't understand how anyone here could be dense enough to treat an early ES that is performing at a level obviously below where it should as any real indicator of final silicon performance. If it performed great it would tell us something. Since it doesn't, it tells us nothing.

Too many people are complaining because most Intel ES's perform just like retail ones or even better than retail ones.

[Blaber]

Too many people are complaining because most Intel ES's perform just like retail ones or even better than retail ones.

[nex_73]

also you can request other benchmark testing.

wPrime
Run as admin
Advanced Settings -> Thread Count: 8 -> Save
Run 32M Benchmark

Run it first @ 3200MHz with C6 OFF
Then run it @ 1600MHZ (8x multiplier) with C6 OFF

My guess: around 10s, in both cases! (IF ES is TDP limited)
If HT Refclk limited to 100 - then the last test will be around 20s

Update: Can you also show us the system summary page from HWiNFO64?
Or HWiNFO32 if you're running 32bit OS...

[imamage]

why is AMD letting these horrendous chips out in the wild?
whats the point of gimping them so bad?

These are ES
They aren't "suppose" to show up in market