[Lightman]

Originally Posted by fellix

AMD Bulldozer FX 4110: First Look
AMD Bulldozer FX 4110: Windows Experience Index (WEI Test)
AMD Bulldozer FX 4110: Temperatures FULL and IDDLE.

Nice!
My Phenom X6 @3.4GHz gets WEI score of 7.6
Looks like Bulldozer 4 cores are quicker in this test than my 6 cores at the same speed

[DarthShader]

WEI score is stored in XML files and can be easily changed, there's no CPU-Z screen and there's no individual core temp values (maybe mobo issue?)

[Lightman]

Originally Posted by DarthShader

WEI score is stored in XML files and can be easily changed, there's no CPU-Z screen and there's no individual core temp values (maybe mobo issue?)

Correct, but my score is not fake

If that Bulldozer score is fake we will only find out at launch I think

[Blazkowicz]

Originally Posted by DarthShader

WEI score is stored in XML files and can be easily changed, there's no CPU-Z screen and there's no individual core temp values (maybe mobo issue?)

ROFL
so it's the same as cheating the high scores in minesweeper?

[Mize]

Since when did WEI mean anything?

[swaaye]

It's like 3dmark. Important to some forumfolk.

[Psycho]

Regarding the 3dmark vantage cpu score..
http://www.futuremark.com/pressroom/..._v100_Rev2.pdf
cpu score: 2500/477.9*5775.19 + 2500/12*281.74 = 88907 (and not 81917)
3dmark score: 1 / ( 0.75/19718 + 0.25/81917) = 24337 (and not 26676)

[denev2004]

Originally Posted by fellix

Source

One core only? It's with defect?

[DarthShader]

Those are very old. AFAIK it was a B1 stepping which had problems.

[Blazkowicz]

in the pics, it's just CPU-Z showing it as six single core CPUs.

[eastmen]

Originally Posted by fudzilla
The list includes two eight-core parts, which were already leaked some time ago. The FX-8130P is a 3.8GHz part and it can overclock to 4.2GHz on turbo. It has a TDP of 125W and should sell for $320. The FX-8110 also packs eight cores and it’s clocked at 3.6GHz sans turbo, or 4.0GHz with turbo. The TDP stands at 95W and the price is $290.

http://www.fudzilla.com/processors/i...ulldozer-specs

I could use a 3.8ghz 8 core cpu for video editing , should blow away my q9650 at 3.8ghz hopefully

[Lightman]

Originally Posted by eastmen

http://www.fudzilla.com/processors/i...ulldozer-specs

I could use a 3.8ghz 8 core cpu for video editing , should blow away my q9650 at 3.8ghz hopefully

It's 3.2GHz + 1.2GHz TC

In other words 3.2 / 4.4GHz yet supposedly retail will be different

[eastmen]

Originally Posted by Lightman

It's 3.2GHz + 1.2GHz TC

In other words 3.2 / 4.4GHz yet supposedly retail will be different

thats not what the article i posted says .

[Lightman]

Originally Posted by eastmen

thats not what the article i posted says .

I know, I wasn't referring to that article, only correcting it.

If retail samples are better than early ES (usually the case), then we might see well over 5GHz on air with single BD core overclocked .

[eastmen]

Originally Posted by Lightman

I know, I wasn't referring to that article, only correcting it.

If retail samples are better than early ES (usually the case), then we might see well over 5GHz on air with single BD core overclocked .

i dunno , i just hope for really good video editing and encoding.

[hoho]

Originally Posted by eastmen

i dunno , i just hope for really good video editing and encoding.

Once that video editing-encoding software gets AVX support BD's performance vs Intel will be nearly cut in half as it has only one AVX unit shared between two int cores in one block while each Intel core has it's own AVX unit.

Of course as long as the software still uses the regular 128bit SIMD stuff then BD will be rather awesome while Intel has half of it's SIMD unit just sitting idle.

[entity279]

Assuming they have the same number of cores

[hoho]

Originally Posted by entity279

Assuming they have the same number of cores

You meant integer cores there, not FPU ones. BD kind of makes it hard to exactly define what a core actually means

[entity279]

Yes, I meant marketing cores.

The point was that most of Intel's offering is still made of quads so your point with half the FP performance is not standing.

[eastmen]

Kyle talked about a hardocp event for bulldozer in july and from whats getting siad over there it seems like we wont see bulldozer till then.

Really pathatic amd !

[mczak]

Originally Posted by entity279

Yes, I meant marketing cores.

The point was that most of Intel's overing is still quads so you're point with half the FP performance is not standing.

I think his point was that intel really handles 256bit AVX natively whereas AMD needs to split it up into 2 instructions.
That said, the AMD SSE unit is definitely more beefy otherwise (though I think latency is quite a bit worse). There's only one sse unit handling muls for instance in Sandy Bridge, so doing nothing but muls is twice as fast on BD (and if you're doing 256bit muls still as fast). It only gets better if you'd do FMAD (though not much since with as many muls as adds you can do the adds "for free" on SB). Not sure how though they'd behave with real code where the most frequent instructions tend to be the shuffle and pack/unpack ones .

[hoho]

Originally Posted by mczak

I think his point was that intel really handles 256bit AVX natively whereas AMD needs to split it up into 2 instructions.

My point was that one BD module with 1x256bit SIMD and 2x int cores has roughly equivalent SIMD throughput to one Intel core. Or in other words, BD has half the 256-bit SIMD throughput per int core vs Intel.

Originally Posted by mczak

That said, the AMD SSE unit is definitely more beefy otherwise (though I think latency is quite a bit worse)

Most likely true on both cases. Only question will be how much software support will be needed to actually benefit from it. If most devs just opt to go with lowest common dominator and not write special code for BD it won't good to say the least.

[rpg.314]

Originally Posted by mczak

I think his point was that intel really handles 256bit AVX natively whereas AMD needs to split it up into 2 instructions.

It doesn't matter in the slightest.

[3dilettante]

There is a relatively minor performance degradation in 256-bit mode, not an improvement.
It matters enough that gcc will emit two 128-bit ops instead of one 256-bit for BD, or at least for this first-gen BD.
FMA4 was the horse AMD bet on, though there is no word on whether Intel will re-adopt it.

[rpg.314]

I thought the splitting referred to 2 uops inside the decoder. I didn't know that compiler itself was tuned to crack them into 2 instructions.