BRIGHT SIDE OF NEWS About | Advertise | Contact

Users profile

Username:
Password:

Lost password

Please enter your email address

Send

New user

Proceed to registration page.

| Register

SUBSCRIBE Newsletter | RSS Feeds

HOME

APPLE

HARDWARE

CLOUD COMPUTING

ENTERPRISE

SOFTWARE

BUSINESS

ENTERTAINMENT

SECURITY

News Analysis Interviews Reviews Rumors

Tuesday, February 09, 2010

Email this to a friend.

Your friend's e-mail:
Your Name:
Your e-mail:
Message subject:

nVidia shows NV100/GF100 GPU running Unigine DX11 Benchmark

11/17/2009 by: Theo Valich - Get more from this author

Just as the swarm of AMD "Hemlock" Radeon HD 5970 reviews hit the Internet, nVidia decided to answer in a subtle way - on the company's Facebook page, "somebody" posted a picture showing the desktop version of NV100 GPU - "GF100" in action.

nVidia NV100-Class GPU is action - is this future GeForce 380?

nVidia NV100-Class GPU is action - is this the future GeForce 380?

The card in question looks quite ready for retail, as far as lack of wiring goes, and is running Unigine's DirectX 11 benchmark. The corner showing framerates is naturally, "cut by accident" but nevertheless, nVidia has desktop parts working, and in just a few weeks, members of the media and partners should start to see these cards live.

nVidia NV100-Class GPU - Fermi architecture in ASIC form. Picture source: nVidia presentation

nVidia NV100-Class GPU - Fermi architecture in ASIC form. Picture source: nVidia presentation

The board itself packs an 8-pin plus 6-pin power, meaning the GPU and memory can suck up to 300W, giving a very healthy overclocking headroom. In our discussions with members of nVidia staff, clock expectations and performance are running quite high, and if we compare the thermal characteristics of Tesla 20-Generation, you can see that "GeForce 380" has around 107W for overclocking purposes.

If you are wondering why are we calling the Fermi GPU "NV100", the reason is actually quite simple.

The name for the architecture is Fermi, while the previous architecture [GT200] was called Tesla. But the actual ASIC has several different names, depending on product where it is going to end up. As a generation of silicon, Fermi ASIC carries the name NV100 is the name for GPU itself, GF100 is name for the GeForce part, and we saw GT300GL for Quadro parts and GT300 for Tesla on several Tesla-based documents back in early 2009.

In any case, regardless of how many codenames the company use [AMD's own HD5970 was codenamed Evergreen as the architecture, Hemlock as the dual GPU part, R800 as the internal name for external companies and finally the board itself was codenamed as Zaphod Beeblebrox].

At the end of the day, it is important what this DirectX 11-class GPUs are doing and are they working properly or not.

Updated: Ostendo fails to deliver in more than one way by: Anshel Sag
nVidia GF100 Architecture: Alea iacta est by: Theo Valich
First DirectX 11 benchmark released by: Theo Valich
World's First DirectCompute, GPGPU benchmark released by: Theo Valich
How ASUS notebook SNAFU caused a war of words between AMD and nVidia by: Theo Valich
Codemasters goes DirectX 11: DiRT 2, GRID 2, and F1 2010 by: Theo Valich
nVidia silently deploys DX10.1 parts to OEMs: G210, GT220 are out! by: Theo Valich
nVidia GF100 Fermi silicon cost analysis by: Theo Valich

Tags:

NVDA, nVidia, NV100, GF100, GeForce 380, GTX380, DirectX 11, NVIDIA DX11, ATI vs NVIDIA, ati vs nvidia 2010, DX11, Fermi, Furbie, cGPU, GPGPU, Unigine, DX11 benchmark , benchmark

Share and enjoy :)

Comments:

Uhh, now we have benchies! by: Anonymous on 1/26/2010

And the NVIDIA benchies K F'in A !
I am really excited by this whole new modular architecture - man nvidia hit ANOTHER ONE out of the park !
Way to go nvidia !
You even played it smart and waited for tsmc to get their 40nm straightened around instead of blowing bazillions on 2 million red rooster chips while 8 million came out dead do-do's like ati the dipsquats did. DUHHH DUMMM ati !
Ya gonaa lose a fortune again ati - you bad business noobs!
Thanks for being the ginea pigs (err.. red roosters) with all the losses !

@not real by: Anonymous on 1/19/2010

sorry but you are wrong. the white reflection is the pci slot.

not real! by: Anonymous on 1/19/2010

dudes u can see its not real cuz of the reflection on the card! there is a blue pci-e thats reflected as white on the card! so shut up and dont argue about image file extensions!

Tesselation by: Anonymous on 12/28/2009

the fermi gpu's don't have a tessellation unit, it runs off the cuda programming sset in the gpu. Therefore tessellation can be faster, but detracts from the gpu's overall power.

Rumor by: Anonymous on 12/18/2009

Interesting rumor...

Fermi tessellation is 2 to 2.5 X faster than AMD's solution...

2 MBs by: Anonymous on 12/14/2009

Hey, have you all noticed the second motherboard on top? It has an Audigy2 sound card fitted ... tres bizzare!

by: Anonymous on 12/7/2009

I wish I could see the results of that bench

off topic. by: Anonymous on 12/5/2009

What is the picture on the monitor from. a game, a movie, a TV show/cartoon, or what? also what is the name of the game, TV show/cartoon, movie, etc? I'm sorry that this is off topic but when i don't know something i really want to know what it is for no reason other then to know.

by: Anonymous on 11/27/2009

1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7% 1.7%

derp by: Anonymous on 11/27/2009

wood screws

by: Anonymous on 11/25/2009

It doesn't have to be photo-shopped to be fake. There could be a system with an HD 5850 or HD 5870 running off camera. Then a picture a motherboard with a fake video card on it, that might even have a real working fan spinning, just for realistic effect. This is proof of nothing.

by: Greg442 on 11/22/2009

Not getting into the whole fan spinning thing, but this photo was definitely photoshopped. I work with photoshop quite a bit and you can easily see this photo was altered as the pixels don't match when you zoom in on the center of the fan.

by: Anonymous on 11/22/2009

Oh god, get over debating the pictures authenticity, LOL...

by: Anonymous on 11/21/2009

Are you people seriously debating whether the picture is real or not?

Images by: Sean Kalinich on 11/20/2009

The image was shot RAW, converted to JPG by Adobe Photoshop CS4 then I reopened the RAW file using Canon's ZoomBrowser EX. From there I took a screen shot of the image in that browser so that you could see the settings used by the camera [this is saved by default to PNG]. The original converted JPG is here;

http://www.flickr.com/photos/29512214@N04/4117954941/

RAW, jpg and png by: Gipsel on 11/20/2009

"on a side note, any JPG is compressed. There is always a loss of image information. Otherwise the file sizes would be the same. a 10.1 MP JPG is around 5-8MB while a RAW file is up to 15MB. There is information that is not present."

Of course is a jpeg compressed (there is even a rarely used lossless variant). But I understood your original post that you trimmed down the picture (still as RAW; side note: if a 10MP RAW file is only 15MB, it uses already [lossless] compression) and uploaded it as a ~1.8MB file to Flickr. But it looked to me you uploaded it actually as png, not as jpg. So if you haven't saved it as jpg in some intermediate step, the png file should introduce no compression artifacts as it features lossless compression. Or does Flickr some automatic conversions internally to save storage space?

PS:
It gets more and more off topic. We should really stop ;)

by: Kyocera on 11/20/2009

You two sound like a long time married couple.

:D :D

Just to end it by: Gipsel on 11/20/2009

Thanks for the description. But in fact my most explicit statement regarding the shot was:
"You definitely didn't use a continuous light source for the shot. What one sees "through" the black fan blade (blown up a bit in the linked picture) is the shot taken with the full exposure time but just the background light."

And that is entirely true. You have only used also artificial light for the background and triggered only the main directed light with the shutter (so most of the picture was taken just in the last moment of the open shutter). But as I said, only you could have known the exact lighting conditions.

With a flash it should look even less motion blurred (shorter light pulse), but that is something everyone can try at home with much simpler equipement than you used as I really doubt most people would make that effort or have your equipment to shoot a photo of a spinning fan ;)

artifacts by: Sean Kalinich on 11/20/2009

My comment on artifacts is not about the fan blades. I mentioned that because you said "By the way, some really crappy energy saving lamps are also flashing with double the line frequency. Something a studio light shouldn't be doing at all (can give quite some artifacts)." Where this was in your comment sounded like you were saying I was using them.

I looked at the image and saw where the tops of the copper tubes were washed out and thought you meant that as an artifact.

on a side note, any JPG is compressed. There is always a loss of image information. Otherwise the file sizes would be the same. a 10.1 MP JPG is around 5-8MB while a RAW file is up to 15MB. There is information that is not present.

Coming from an standard film back ground [35mm Canon EOS 1V HS and Mamiya Medium Format] I found that out the hard way when converting my film to digital format. I made the mistake of saving to JPG instead of Tif and could visibly see the loss of quality later when I looked at them. I spent a ton of money to correct that mistake.

We should stop at some point! by: Gipsel on 11/20/2009

"As for your question about the fan it is not relevant. You said it could not be done and I did it. Seems like you are nitpicking the image that was presented instead of admitting it is possible.

It does not matter how I did it, I did it and the photo was not altered in any way."

To be correct, I said how it can be done before you even had the idea to do the shot (in my very first comment here).

And if I remember right, your first try to explain that one sees the cooling fins through the fan blade was some rescaling or compression artifact until I pointed out you used a lossless compression.

I see no reason to backtrack any comment here, I just tried to put it into context as you apparently misunderstood some of them. A very clear example is the side note about the energy saving lamps, how about backtracking your claim that I wrote you used such lamps?

I guess we could agree that one should take each others words with a pinch of salt as we speak different languages and also the motivation of the winks and smileys may not come across as intended (trying to build bridges here ;).

But you are correct in one point, the whole discussion gets largely irrelevant (as I said already some time ago).

But since you want to know by: Sean Kalinich on 11/20/2009

The shot was obtained with four lights two facing the front one pointed at the back these are open lights that fire into reflectors but they provide quite a bit of light.

The fourth light is behind and over the left side of the camera. It has a defuse cloth filter on it. It is tied in with the shutter release cable I use. It turns the light on in time with the shutter. But leaves it on.

As 99% of digital cameras do not have anything resembling back fill flash of a method of delaying the flash I find this useful for in studio shots of people and product.

It is not using a pulse flash at all.

Four directed lights were used.

See all comments

Leave a comment:

Author:

Title:

Comment:

Enter the code shown above:

(Note: If you cannot read the numbers in the above
image, reload the page to generate a new one.)