Real confirmation, NDA over @ 1201 for Vishera.

[Skurge]

damnit, it's already past midnight here

 

[DeeDot78]

Preliminary specs, says it will. Stock for stock.

Wont happen, I just hope an 8350 at 4.5 can compete with phenom 2 this time.

 

[VirtualLarry]

I hope that this is an improvement in both performance and performance/watt compared to a Thuban X6 1045T @ 3.51Ghz (at 1.325v, IIRC, which was basically stock voltage).

Then I might consider an upgrade, even if it is $200 for the part.

Looking forward to getting me some AVX and FMA for DC.

 

[DeeDot78]

Look at that voltage. If its stable then, great.

 

[Skurge]

Are those benches I'm seeing over at XS legit? cause I see the 8350 regularly beating my 3570K.

I'm so hoping that's the case.

 

[Abwx]

Are those benches I'm seeing over at XS legit? cause I see the 8350 regularly beating my 3570K.

I'm so hoping that's the case.

The link of the relevant french site is in another thread at AT ,
but it seems that the review was pulled out at this moment....

 

[Olikan]

same clocks, if google crome translator is not trolling me
http://extrahardware.cnews.cz/recenze/fx-8350-vishera-vrati-osm-jader-piledriveru-na-4-ghz-amd-do-hry/strana/0/1

 

[sequoia464]

Are those benches I'm seeing over at XS legit? cause I see the 8350 regularly beating my 3570K.

I'm so hoping that's the case.

Linux benches look good - not that they do anything for me...

http://forums.anandtech.com/showpost.php?p=34131448&postcount=7

 

[FalseChristian]

I think the demise of AMD is greatly exaggerated. The new FX's look incredible. I think at 5GHz
it is comparable to a i7 3570K at stock.

 

[Rhezuss]

Damn damn damn...really hope it's time for me to upgrade

 

[ShintaiDK]

http://extrahardware.cnews.cz/recen...r-piledriveru-na-4-ghz-amd-do-hry/strana/0/11

No surprise, no magic. And its still using ~2½x more power than a 3570K.

 

[ShintaiDK]

Look at that voltage. If its stable then, great.

 

[pelov]

http://extrahardware.cnews.cz/recen...r-piledriveru-na-4-ghz-amd-do-hry/strana/0/11

No surprise, no magic. And its still using ~2½x more power than a 3570K.

Although still behind in many benchmarks, that's quite a big jump from Bulldozer and far more than I was expecting.

The power consumption is still ugly and it's a year too late.

The multi-threaded gains are impressive.

 

[DeeDot78]

That looks to be more realistic.

 

[leper84]

http://extrahardware.cnews.cz/recen...r-piledriveru-na-4-ghz-amd-do-hry/strana/0/11

No surprise, no magic. And its still using ~2½x more power than a 3570K.

So I take it from reading your posts you may not be a fan of AMD, huh?

 

[inf64]

The Linux numbers in that extrahardware review are kinda strange. For instance Michael from Phoronix gets much better (~2x IIRC) C-ray results with his 8350 versus what those guy get. I can't see what was the exact number for C-ray in Phoronox review since they pulled it (NDA prevails?).

 

[NostaSeronx]

The Linux numbers in that extrahardware review are kinda strange. For instance Michael from Phoronix gets much better (~2x IIRC) C-ray results with his 8350 versus what those guy get. I can't see what was the exact number for C-ray in Phoronox review since they pulled it (NDA prevails?).

Older compiler. PHO used GCC 4.7 but I am not sure if it was the latest one GCC 4.7.2.

Take it from a standpoint without optimization to architecture i7 3570K and FX-8350 perform the same. While once you optimize for both the FX destroys the i5/i7. The only issue is GCC is heavily biased towards Intel because the people who do the intel core optimization read Agner Fog's microarchitecture/instruction latency. While the AMD side is just ideas how the processor should work in theory.

GCC:
Intel sends data to GCC -> GCC then checks with Agner Fog data -> creates optimizations.
AMD sends data to GCC -> GCC implements optimization.

We all know about AMD and there awful software optimization guides. Rory Read definitely needs to get guys to proof read and get the cut and pastes from 00h and k7 out of there.

 

[Fox5]

Although still behind in many benchmarks, that's quite a big jump from Bulldozer and far more than I was expecting.

The power consumption is still ugly and it's a year too late.

The multi-threaded gains are impressive.

The gap's been closed to what it was with Phenom 2 versus Nehalem. Not a good situation for AMD to be in, but good enough that they're not completely out of the market. Shame it took an extra year to get this out though, had they just made a Phenom 3 they would have been this competitive last year.

Also, AMD may perform well in Linux with the latest GCC and special optimizations...but that's not a typical use case even in Linux. Most people just grab the packages for their distro, which will be compiled for x64 SSE2.

 

[exar333]

Makes me worried. When they avoid performance and start comparing 'check-box features' its all over.

 

[LOL_Wut_Axel]

So basically it's 5-10% faster than the 3570K in multi-threaded and in everything else including single-threaded and power consumption it loses massively.

It's still behind Intel in most metrics, but it's okay if you want it for video editing or transcoding on the cheap, I guess.

 

[TakeNoPrisoners]

Single threaded performance is a big let down; seems like Intel is still the way to go for gaming.

 

[Rvenger]

Available at newegg already


http://www.newegg.com/Product/Produc...ishera&x=0&y=0

$220 for 8350

 

[Schmide]

I see techpowerup
xbitlabs
toms

 

[alexruiz]

I see techpowerup

The "witch hunter" also got his online:
FX8350 at TR

 

[Magic Carpet]

Pricing will be vital for this. I think, I just might reconsider a few coming builds...