[OBR] amd's 8350 tested

[The Alias]

note : this is a eng sample http://www.obr-hardware.com/2012/09/preview-amd-fx-8350-piledriver-last.html
discuss

 

[Smoblikat]

If thats true then this MIGHT be a viable competitor for SB, my 2500K at 4.3 gets around 6.8 in cinebench, and the 8350 @ 4.2 gets 6.8, though it is double the cores. So I assume single thread performance still sucks major dick. Also, look at the voltage required for that low OC. I understand its just an ENG sample, but it should give an accurate example of general performance. Unless they make a revision that changes the architecture Id say that this will be another BD.

EDIT - It looks like they did the comparison with the PD at 600mhz higher than BD, if thats true then performance might have gone down . Not unheard of from AMD.

 

[pelov]

EDIT - It looks like they did the comparison with the PD at 600mhz higher than BD, if thats true then performance might have gone down . Not unheard of from AMD.

The CPUID picture where it says "Zambezi" should be a clear giveaway. These are the same pictures that were posted before and they were fake. The benchmark figures were all over the place, from way better to 10-20% worse.

 

[Riek]

If thats true then this MIGHT be a viable competitor for SB, my 2500K at 4.3 gets around 6.8 in cinebench, and the 8350 @ 4.2 gets 6.8, though it is double the cores. So I assume single thread performance still sucks major dick. Also, look at the voltage required for that low OC. I understand its just an ENG sample, but it should give an accurate example of general performance. Unless they make a revision that changes the architecture Id say that this will be another BD.

EDIT - It looks like they did the comparison with the PD at 600mhz higher than BD, if thats true then performance might have gone down . Not unheard of from AMD.

the 8350 runs at 4Ghz not 4.2Ghz (thats it boost).
if cpuid is wrong about the id... its probably wrong about everything else including voltage....


The results are somewhat in line with my expectations for the real chip though. Around 15% higher performance than a 8150 in fpu intensive applications or in applications where the difference is mostly due to the added clockspeed.

a ~4% gain in fpu intensive applications at the same clockspeed is pretty good considering there were no major changes for this.

but still its from obr so...

 

[tulx]

How can anyone this site seriously when they write: "This means performance of FX-8350 is on par with Core i7-2600K ... somewhere better, somewhere worst. But in average is Core i7-2600K good competitor for FX-8350. Pathetic result ...

Because, FX-8350 is the same shit like FX-8150 price of that "crap" can be only around 200 USD."

or "But it has grown only little, wet dream of children and AMD fans has not been realized. IPC is increased up to a few percent. "

It doesn't even matter how good or bad the product is that you're describing - if that's the best wording you can come up with, you're simply stupid...

 

[Vesku]

Please put [OBR] or [Rumor - OBR] in title, thank you.

 

[NTMBK]

How can anyone this site seriously when they write: "This means performance of FX-8350 is on par with Core i7-2600K ... somewhere better, somewhere worst. But in average is Core i7-2600K good competitor for FX-8350. Pathetic result ...

Because, FX-8350 is the same shit like FX-8150 price of that "crap" can be only around 200 USD."

or "But it has grown only little, wet dream of children and AMD fans has not been realized. IPC is increased up to a few percent. "

It doesn't even matter how good or bad the product is that you're describing - if that's the best wording you can come up with, you're simply stupid...

Agreed. Even the petty levels that Semiaccurate descends to don't compare with how utterly ridiculous OBR comes across. (Not to mention that the fact that the site is named after a gun is rather ridiculous, in my eyes- although I am British.)

 

[Arkadrel]

double thread, so same reply as other thread:

ORB.... no thanks.

Rather see if anything showed up on Chiphell or the like, not really too trusting of ORB.
He's been wrong to often, or claimed things that where wrong, or lied/made things up.
That happends when your getting page hits off of rumors, and the more juicy the better.

These are the same pictures that were posted before and they were fake.

Orb likes page hits, enough to make up things on the spot.
If ever a title deserved something like [Rumor site ORB] in title, nows the time.
The "amd 8350 tested" sounds too legit, for something thats just fake(s) or made up news for page hits by ORB.

 

[pelov]

Rather see if anything showed up on Chiphell or the like, not really too trusting of ORB.

It did. That too was fake.

We already have a good idea just how the chips will perform, though. Single digit IPC bump overall and small increase in clock speeds that vary depending on the turbo. Judging by the clocks, the multi-threaded performance should be a good increase while the single-threaded performance would only increase by whatever their IPC increase was.

 

[Homeles]

As far as performance goes, Piledriver just puts AMD in the same place they were in before Ivy Bridge launched. The power consumption front should look phenomenally better, though.

 

[guskline]

Never put much stock in an engineering sample. My Bulldozer 8150 runs on less voltage at 4.2Ghz. Heck I have it at 1.425V at 4.5Ghz and it just passed 6 hrs of prime without a blip. I think this might be a bogus article. Beside, the source was very negative about the Bulldozer. Fnally my 8150 at 4.5 Cinebenches 7.32.At 4.2 it does 6.92. HMMM???

 

[zebrax2]

Last i remember that was the site that released those fake bulldozer? benches then later told everyone that it was fake.

Anyway I'm not really expecting that much out of PD

 

[LOL_Wut_Axel]

10% faster than Bulldozer if this is accurate. Big deal.

If this is accurate it's still slower than Sandy Bridge and about on par with Nehalem, which came out four years ago. And that's only in multi-threaded, mind you. In single-threaded it's still a dog.

"You can't polish a turd."

 

[bunnyfubbles]

Problem with this is that the base clock is ~11% higher, and thus the IPC appears to be realistically only ~4% improvement. I suppose not terrible considering same 32nm fab, but really nothing to write home about and certainly not enough to get AMD back into the game.

 

[The Alias]

oh I didn't know obr was not to be trusted lol sorry guys

 

[MLSCrow]

Problem with this is that the base clock is ~11% higher, and thus the IPC appears to be realistically only ~4% improvement. I suppose not terrible considering same 32nm fab, but really nothing to write home about and certainly not enough to get AMD back into the game.

Why do you think AMD is advertising Steamroller and not Piledriver? Because, like you said, it's nothing to write home about. Steamroller is AMD's true saving grace. Piledriver is simply the refresh of Crapdozer. It may finally be a teeny upgrade from Thuban, but Steamroller is what most AMD fans should truly be excited about, IMO.

 

[Borealis7]

my pet octopus makes better predictions than OBR.

 

[StrangerGuy]

Objectively Bad Reputation.

 

[Idontcare]

oh I didn't know obr was not to be trusted lol sorry guys

my pet octopus makes better predictions than OBR.

The problem with OBR is not just that he literally makes stuff up for the pagehits (an accusation that could be leveled against any of Mike Magee's understudy's - Theo (BSN*), Faud (Fudzilla), Charlie (S/A), etc); rather, the problem with OBR is that he intentionally makes stuff up with the goal of misleading and infuriating the fanboys.

:hmm: hmmm, on second thought, maybe OBR and the others aren't so different after all...

 

[Olikan]

Update:

disclaimer: this is not tested by obr itself, but by chinese owner of this web. Cpu is final es sample, but bios has only early support of or-c0 chips. There are lot of problems with stability a freezes during testing. Final performance can be different, but i dont believe that. It was the same with es bulldozer or-b0 chips. Es has the same performance as retail cpu. This preview is for flanker, biggest amd fan on the planet. Btw, thanks for fx-8150 results buddy

 

[Arkadrel]

@Olikan

a footnote that claims the chip was tested by some chinese dude that owns the website, doesnt change the fact that ORB (and his site) are AMD bashers, that love to make up rumors and falsify test results/pictures ect.

Nothing anywhere close to ORB or comeing from that site of his, reguardless of who he claims obtained the test results are to be trusted. The record they have is just to spotted.

 

[piesquared]

Nm

 

[Hatisherrif]

 

[sandorski]

There's so much wrong about that pic. I loled though.

 

[Nemesis 1]

Last i remember that was the site that released those fake bulldozer? benches then later told everyone that it was fake.

Anyway I'm not really expecting that much out of PD

Yep and it was tallked about alot here. Than BD was released and it showed orbs bugus benchmarks were right on the money . It was the AMD guy that told the stories of fantasy here at AT forums .