FERMI Benchmarked?

[StAndrew]

Found here: http://forums.nvidia.com/index.php?showforum=172

Original source: http://www.guru3d.com/news/geforce-g...80-benchmarks/

Do not ask us how or where these benchmarks source/originate from as we do not know. For all we care they are fake. However, a user in our forums posted screenshots of benchmarks from what seems to be a NVIDIA Slide Deck presentation. I can confirm that the layout and style is exactly what NVIDIA uses for it's presentations.

The slides reveal internal NVIDIA benchmarks of Far Cry 2, Resident Evil 5 and STALKER Clear Sky. Now this remains speculation, but if the numbers are right, then the future is looking bright and sunny for NVIDIA -- well if they get their pricing and TDP right that is.

Anyway, have a look and props go out to forum user "Memorian" for posting these in this GTX 300 series discussion thread.

To reiterate, take with a grain of salt...

But if true...

[ajaidev]

Holy crap...!!! GTX360 </> 5970

Dam 5870 is not even in the picture wtf??

Seems fake tough

If GTX 380 performs sooo well how will GTX390/5 perform 280 fps in RE5 :O

These are very unlikely just like the first fermi demos that appeared!!

[NaMcO]

Wouldn't surprise me at all... But rather wait and see actual card benches than commenting on those.

[Helmore]

I'll keep these benchmarks numbers in the "seems plausible" category of rumors. One thing of note though, the GTX380 is actually more than twice as fast as the GTX285. Quite an achievement if that turns out to be true.

[Shadov]

Seems fake, take a look at the same benchmarks with HD 5870 and HD 5970:





And now compare to:

[N19h7m4r3]

Very interesting.

But I don't believe them really.

If this was the case, why would NV not just send a card or 2 out to review sites.

That would cause all doubt to be erased and get more people to wait for Fermi.

[duckbillclinton]

Very interesting.

But I don't believe them really.

If this was the case, why would NV not just send a card or 2 out to review sites.

That would cause all doubt to be erased and get more people to wait for Fermi.

NDA maybe, but yes, they could send 1 or 2 out then with the "leaked" scores...

[zalbard]

PR crap probably.
ATI had some mind blowing performance graphics, too, iirc.

[NaMcO]

Those PCLAB.PL graphics are a joke, my 295 performs way better than what they claim at those settings. I think right now anyone can throw out fermi benchmarks because nobody knows how they perform but on the other hand many people know how 5870 and 295 perform and can smell fake benches by a mile.

[eric66]

love this kinda anonymous benchs lol

[Helloworld_98]

would like to see AF benchmarks, AA is becoming a bit of old hat now that higher and higher resolutions are being used.

[ajaidev]

This guy at a Chinese forum made an excellent point he edited the slides and put these up lol







I hope fermi gets out soon... or else people are gonna mock it like above.

[eric66]

damn hemlock rocks in sec test must be some error there :P

[orangekiwii]

Seems fake, take a look at the same benchmarks with HD 5870 and HD 5970:





And now compare to:

You obviously CAN'T compare yourself. The Stalker clear sky settings are different - 8AA vs 0AA

and the Far Cry 2 ones are basically identical - settings and fps wise... seems plausible to me

[Genius_pt]

Wow, if true not bad really.

Finally we get some good explanations (amazing performance) for the amount of time we've waited for this cards .

[eric66]

there are ppl who believe this crap thats even amazing :P clocks aren't fnalized but somehow we already have gtx 360 and 380

[orangekiwii]

why SHOULDN'T we believe it

g80 was such a massive jump over g70, why can't fermi be a similar if not even larger jump over the gtx280?


just because it hasn't happened doesn't mean it can't


yes its unlikely, but stop being so pessimistic towards everything

[eric66]

slides look too cheap imho don't know if those were real nvidia would be probably the one posting it all over the net

[Luka_Aveiro]

Results seem plausible, more than doubled performance compared to GTX285.

[JohnZS]

Holy smokes Batman!!! I hope this IS true

[N19h7m4r3]

I just want the GTX 380 to use a single 6 or 8pin power connector and be at MOST 9.5"

[bladerash]

Well.. I remember somewhere on the internet... here: http://www.semiaccurate.com/2009/12/...-silicon-oven/
some info about insider saying the targeted clock at 750Mhz would make fermi about 40% faster than cypress.
But their A2 silicons was only able to reach 500Mhz.. which means 50% more clockspeed to go... They just got A3 and ya think it will clock at 750Mhz?

These numbers showing increase in performance of way beyond 50-70% over the cypress is something I am prolly not gonna put my faith on.

[JohnZS]

But bladerash there is a clue in the name of that website... SEMI-accurate lets hope they hit 600Mhz and wanted to go the whole at least 700Mhz hog with A3!

[Chumbucket843]

here are some fakes i found. note how the vram is incorrect with the bus widths. take benchies like these with a grain of salt.

[ajaidev]

@Chumbucket843 there were specs there also lol fake specs


Specifications

GTX 380
CUDA cores 512
core clock 730MHz
shader clock 1600MHz
Memory 1100MHz 2GB
TDP(watts) 225w

GTX 360
CUDA cores 480
core clock 650MHz
shader clock 1400MHz
Memory 1000MHz 1.8GB
TDP(watts) 195w

http://www.tomshardware.com/forum/27...-gtx380-gtx360