backtop


Print 77 comment(s) - last by skroh.. on Oct 10 at 5:38 PM

More G80 features abound

As if we mere mortals needed more reasons to be excited about G80, here are a couple more tidbits: 128-bit high dynamic-range and antialiasing with 16X sampling.

The high dynamic-range (HDR) engine found in GeForce 7950 and Radeon series graphics cards is technically a 64-bit rendering.  This new HDR approach comes from a file format developed by Industrial Light and Magic (the LucasFilm guys).  In a nutshell, we will have 128-bit floating point HDR as soon as applications adopt code to use it. OpenEXR's features include:
  • Higher dynamic range and color precision than existing 8- and 10-bit image file formats.
  • Support for 16-bit floating-point, 32-bit floating-point, and 32-bit integer pixels. The 16-bit floating-point format, called "half", is compatible with the half data type in NVIDIA's Cg graphics language and is supported natively on their new GeForce FX and Quadro FX 3D graphics solutions.
  • Multiple lossless image compression algorithms. Some of the included codecs can achieve 2:1 lossless compression ratios on images with film grain.
  • Extensibility. New compression codecs and image types can easily be added by extending the C++ classes included in the OpenEXR software distribution. New image attributes (strings, vectors, integers, etc.) can be added to OpenEXR image headers without affecting backward compatibility with existing OpenEXR applications.
NVIDIA already has 16X AA available for SLI applications.  The GeForce 8800 will be the first card to feature 16X AA on a single GPU.  Previous generations of GeForce cards have only been able to support 8X antialiasing in single-card configurations.

This new 16X AA and 128-bit HDR will be part of another new engine, similar in spirit to PureVideo and the Quantum Effects engines also featured on G80.


Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Problem with article
By Chillin1248 on 10/6/2006 2:03:28 AM , Rating: 2
I quote this from a B3D member:

quote:
The DailyTech article completely misinterpreted the whole "NVIDIA and OpenEXR share fp16/32 pixel formats" issue (which has been true since NV30 and the launch of EXR). It may have been referenced again in the marketing materials on which these posts were based, presumably in the context of the new fp32 framebuffer blending (required for DX10...) However, this is just a matter of using a common binary representation for 16-bit and 32-bit floating-point image values. EXR, the file format, is a much more complex standard than the uncompressed fp16/32 pixel formats it specifies (and shares with NVIDIA).

DailyTech then went to openexr.org and read about the OpenEXR *software* library -- a C++ library for reading and writing EXR files, including RLE, wavelet, etc. compressed disk-storage formats.

This is not at all related to G80.


Can we clear this up?


-------
Chillin




RE: Problem with article
By Chillin1248 on 10/6/2006 2:05:28 AM , Rating: 2
Another thing pointed out by another B3D member:

quote:
It should also be noted that DailyTech lists some rather unlikely numbers in terms of memory clocks and texture rate. They list the GTX as having 86GB/s of bandwidth, and the GTS only 64GB/s. However, they also claim both have 900MHz GDDR3, which is highly unlikely. It should be relatively clear that this was a simple typo, and that the GTS actually sports 800MHz GDDR3. As for texture rates, they give the number of 38.4GPixels/s for the GTX, but few logical ways exist to attain such a number with the specified clockrates. As such, it seems much more likely that this part of the specification relates to the GeForce 8800GTS and its 1200MHz stream processors.


Looking foward for a response from DailyTech on what's going on.

-------
Chillin


By KristopherKubicki (blog) on 10/6/2006 7:01:47 AM , Rating: 2
Unfortunately, until we have a way to test some of these metrics I can't really go into much more detail other than what was provided for us.


RE: Problem with article
By KristopherKubicki (blog) on 10/6/2006 6:49:59 AM , Rating: 2
quote:
Can we clear this up?


You'll have to bear with me a little here because some of this information is paraphrased from marketing materials, and we obviously would rather not have people tracking down the exact materials we are referencing.

The 128-bit implementation of HDR on the G80 is based on several technologies from OpenEXR, the binary storage method I'm sure just being one of them. This part was quoted near verbatim from the included material we received, though as he B3D reader pointed out you'll have to look at it in context and not as a "feature" in G80.


RE: Problem with article
By Chillin1248 on 10/6/2006 10:35:15 AM , Rating: 2
Thank you very much for responding, looking foward to further updates.

Your truly,
-------
Chillin


RE: Problem with article
By skroh on 10/10/2006 5:38:58 PM , Rating: 2
Let me clear it up for you even more.

You're quoting the forums at Beyond3D. Those forums are noted for a membership that is rabidly, passionately, unabashedly pro-ATi/anti-nVidia.

That you are finding posts there that are trying to "debunk" the apparently fantastic specs of the G80 with nitpicks about OpenEXR and nakedly subjective speculation such as "doubtful" and "highly unlikely" should come as no surprise. This has been the tone of discourse on that board for a long time now.

Bringing their pro-ATi bigotry over here and using it to imply criticism of the competence and journalistic integrity of the DailyTech staff is just spreading the disease.


"So if you want to save the planet, feel free to drive your Hummer. Just avoid the drive thru line at McDonalds." -- Michael Asher

Related Articles













botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki