Talk:GeForce/Archive 1

This is a summary page
I don't think this page needs detailed lists of part specifications added to it, 6800 or otherwise. Its a portal to the main GeForce pages, and nothing more. If you think Wiki is lacking, and at this point I would beg to differ there, having personally done a lot of the cleanup and created the navigation bar, please add the detail to the linked pages. Timharwoodx 07:00, 23 Apr 2005 (UTC)


 * i think this page needs gpu power consumption stats. that would be awesome:)


 * here are some
 * http://www.xbitlabs.com/articles/video/display/gpu-consumption2006.html
 * http://www.xbitlabs.com/articles/video/display/ati-powercons.html


 * Summary page or not I wouldn't want to have to open 10 tabs to check card vs. card when I could easily look at a table and be done with it. I second the power consumption stats as well.

Neutrality dispute
Seems like much of this article was deposited by nVIDIA's copy writers. In particular, the GeForce Go section uses uncited phrases like "flexible and powerful." ---Ransom (--208.25.0.2 19:49, 29 March 2006 (UTC))


 * Fixed? I changed this:
 * "Since the GeForce 2, NVIDIA has produced number of counterpart designs for notebook computers. Branded GeForce Go, the notebook graphics processing units (GPU) from NVIDIA are flexible and powerful.  GeForce Go provides top-to-bottom solutions for many mobile platforms including thin and light, desktop replacement, and business notebooks, as well as mobile workstations. NVIDIA’s notebook GPUs deliver long battery life with good system performance."
 * to:
 * "Since the GeForce 2, NVIDIA has produced number of counterpart designs for notebook computers, such as the GeForce Go, a notebook graphics processing unit (GPU). GeForce Go provides solutions for many mobile platforms including thin and light, desktop replacement, and business notebooks, as well as mobile workstations."


 * If anyone doesn't like it, feel free to revert, but I took out such claims as "delivers long battery life with good system performance", and the already mentioned "flexible and powerful". 207.42.160.58 22:09, 2 April 2006 (UTC)


 * It's hard to look at this page as being unbiased. The whole article is dismally low on citations and rife with advertising language. There needs to be some serious cleanup here.NorsemanII 20:33, 5 April 2006 (UTC)


 * Did some work a while ago (like removing claims of the cards' "prowess"), but looking back it still needs a lot of work. --72.140.244.130 02:41, 30 May 2006 (UTC)


 * Other than quips like, "once again established NVIDIA's performance leadership over its competitors," the article is fine. Just needs some TLC to make it infomrative instead of pitchy.

Quadro
Nothing on the Nvidia Quadro workstation lines (both desktop and laptop)? --Nantonos 17:32, 30 August 2006 (UTC)
 * Quadro != Geforce. It has its own article.
 * I was imprecise, apologies. Nothing on comparisons of Quadro with GeForce? There is a lot of shared silicon between the two lines (eg for mobile chisets, Quadro 2500M ~= GeForce 7900GTX). Some comparative material would be good.
 * I'm under the impression that for games both are about the same, but for professional applications the Quadro is just shockingly faster. --Swaaye 18:07, 2 September 2006 (UTC)

It depends, the Quadro is a professional series card, but they have low end and high end. Some are very basic corporate desktop cards, others are high end cards these tend to be called DCC (digital content creation) cards. It's difficult to track them w.r.t. the mainstream cards due to the way they are named and marketed. Quadros started out as very similar hardware with features enabled like anit-aliased lines and two sided lighting which would affect 'professional' application performance & benchmarks, but they have grown into an entire differentiated product line, you can no longer make simple comparrisons between GeForce and Quadro, the DCC stuff is the high end where NVIDIA seems to have adopted the FX product label so the high end is the Quadro FX 5600, the core is probably G80 based but it may well be a dedicated silicon design, they tend to have significantly enhanced memory performance w.r.t. mainstream cards and drivers certified for use with specific applications.

The advert tag
See the neutrality dispute above, I'll add some specifics here.


 * "is an evolutionary improvement" - Can anyone tell me what that means?
 * "once again established NVIDIA's performance leadership over its competitors" - This isn't advertising?
 * “perform extremely well on older DirectX 7 and 8 titles.” - According to what standard?
 * “the high-end GeForce 256” - Which means what exactly? Details, not advertising babble! -NorsemanII 00:28, 25 October 2006 (UTC)
 * "is an evolutionary improvement" - Can anyone tell me what that means?
 * sure. This means it's not a major change but offers refinements. It is similar to its predecessor in many ways.


 * "once again established NVIDIA's performance leadership over its competitors" - This isn't advertising?
 * Well, GeForce2 GTS outperformed its competitors for quite a while. Voodoo5 was no match and Radeon could only win in a few circumstances with 32-bit color. Same with Kyro.


 * “perform extremely well on older DirectX 7 and 8 titles.” - According to what standard?
 * This is tricky. NV3x was built to perform very well on older titles. It has an inefficient hardware design to do this. Unfortunately it was not really a useful design choice because it hindered DX9 performance while boosting speed in old games that didn't need help anyway.


 * “the high-end GeForce 256” - Which means what exactly? Details, not advertising babble!
 * There are two GeForce 256 boards. The value board with SDR and the high-end model with DDR. This isn't properly conveyed in this sentence though so it will be removed.


 * The above dispute is 6 months old and the page has been overhauled since then. Those "advertisement" paragraphs were multi-paragraph glorifications before. It seems you are sensitive to overly descriptive adjectives. Some of the writing is inaccurate. I will fix your concerns. This is not worthy of a giant whiny tag though. --Swaaye 03:17, 25 October 2006 (UTC)

NV15 is NOT the earliest GeForce2
The article states "GeForce2 - Launched in April 2000, the first GeForce2 (NV15) ...". I do have the following video card according to lspci: VGA compatible controller: nVidia Corporation NV11 [GeForce2 MX/MX 400] (rev a1) / Class 0300: 10de:0110 (rev a1). Someone got it wrong - the article or lspci? j.engelh 21:49, 2 January 2007 (UTC)

The loop (reboot) error in Windows XP
Please take into account it's very hard to give resources to something that mainly lives in forums. But it does affect many users. -Lwc4life 14:22, 30 May 2007 (UTC)
 * I believe that it's mainly POV. We don't need such kinds of statements here. rohith 19:04, 12 June 2007 (UTC)
 * Stating facts does not violate POV policies..... 71.105.218.44 06:39, 23 June 2007 (UTC)
 * This is completely irrelevant to the article. Allowing some user with a driver issue to contaminate a very generic article with a very narrow issue is ridculous, especially when it takes up so much space. There are/have been innumerable bugs with various generations of graphics cards, this is not the place for this stuff, the issue is not whether it is factual it is whether it belongs in this article and whether it should take up more space than the meat. This seems better served with an external link or perhaps a reference to an article about a loop reboot NVIDIA driver issue. —Preceding unsigned comment added by 199.106.103.254 (talk • contribs) 19:16, 26 July 2007
 * And now people with the bug will first come here looking for help only to realize they won't get it in a place where any anonymous user can remove whole sections not only without a consensus but without a reason. -Lwc4life (talk) 18:04, 4 January 2008 (UTC)

'Naming Scheme' Table
Just a quick one... The Shader Resolution in the Budget row reads "≤50%" - ie. "less than or equal to 50%". Should this not instead read "≥50%" - "more than or equal to 50%"? 86.6.44.148 23:46, 19 August 2007 (UTC)

New GeForce G92??
I think the new GeForce card (possibly the G92 everyones been on about) is the following...

http://www.fudzilla.com/index.php?option=com_content&task=view&id=3430&Itemid=1

Might need a little more research but I don't have time at the moment. Danno81 08:22, 15 October 2007 (UTC)

Potential 9 Series
I am not sure about this but he N92 which is officially the 8800 GT came out today, it is better than a GTS and close if not better than a GTX. It is not better than the Ultra having 16 less shaders and having a smaller memory clock —Preceding unsigned comment added by 12.16.74.28 (talk) 00:09, 30 October 2007 (UTC)

It's not more powerful than the GTX (research some online benchmarks) and the Ultra has exactly the same specs in regards to shaders etc as the GTX because they are exactly hte same card, the only difference is the Ultra is simply clocked higher. The 8800GT is a high midrange card, rather than a competition for the high end GTX/Ultra cards. Yes it does beat the GTS though in most cases. Danno81 09:54, 30 October 2007 (UTC)

9 Series Release Date
I think Q4 2007 release date should be removed as this is not confirmed by Nvidia anymore —Preceding unsigned comment added by Achandab (talk • contribs) 22:56, 30 October 2007 (UTC)

Suffix order
The table outling card performance lists the suffixes for midrange cards as (weakest to most powerful) "LE, GS, GTS, GT, XT". However, the 8600 GTS is more powerful than the 8600 GT. Is the table just plain wrong, or did the order listed hold true for older cards? Bombot 15:13, 16 November 2007 (UTC)
 * What do those even stand for? I'm not seeing GTS here. --DocumentN (talk) 00:29, 21 February 2008 (UTC)

Dubious statement on GeForce 9
Where's the source? I can quote from two sources to dispute this claim. and. Apparently, according to these two websites, D9E is to be a dual GPU version of the GeForce 8. speaks rohith. 19:20, 3 December 2007 (UTC)

Look for newer sources. That was published in October before anything was known. It is mostly speculation at this point. —Preceding unsigned comment added by 76.175.229.87 (talk) 02:57, 6 December 2007 (UTC) the nvidea 8800 is good —Preceding unsigned comment added by 88.107.40.107 (talk) 17:12, 7 December 2007 (UTC)

Reason for undoing revision
User 76.175.229.87, you have quoted two sources (DigiTimes and X-bit Labs) and say that GeForce 9 is pretty much confirmed, according to "confirmed sources". However, the article on X-bit Labs uses the DigiTimes article as its reference. As it is, I have stated above that there is at least one source that disputes this. Keep in mind that NVIDIA has neither confirmed, nor denied the existence of a GeForce 9, but we do know from an official about the close to 1 TFLOPS statement. speaks rohith. 11:27, 9 December 2007 (UTC)

capitalized t in The
The was the first word in the sentence and uncapitalized. I capitalized it.

64.252.83.100 (talk) 01:44, 22 December 2007 (UTC)

Notability of removal of video mirroring feature?
The article says that the decision to remove the video mirroring feature was "controversial", but the only quoted source is a discussion board thread where someone asks if it is available and several people are unhappy than it is not. This should not be considered a 'notable' source - you can find people complaining on discussion boards about anything. 81.156.187.223 (talk) 23:58, 7 February 2008 (UTC)


 * Duly tagged with vc, so speaks rohith. 22:17, 9 August 2008 (UTC)

Removal of Vidoe mirroring - This section is incorrect
Removal of the video mirroring feature is not inherent in the geforce 8 series graphics cards, it's a feature of windows vista. Vista lacks the video overlay, which is required for "video mirroring" or full screen video overlay as it is better known Vista lacks this as part of its HDCP compliance. XP on the other hand will full screen video overlay quite happily even on 8 series graphics cards as i do frequently —Preceding unsigned comment added by 131.111.121.12 (talk) 12:57, 19 August 2008 (UTC)

Fair use rationale for Image:GeForce newlogo.png
Image:GeForce newlogo.png is being used on this article. I notice the image page specifies that the image is being used under fair use but there is no explanation or rationale as to why its use in this Wikipedia article constitutes fair use. In addition to the boilerplate fair use template, you must also write out on the image description page a specific explanation or rationale for why using this image in each article is consistent with fair use.

Please go to the image description page and edit it to include a fair use rationale. Using one of the templates at Fair use rationale guideline is an easy way to insure that your image is in compliance with Wikipedia policy, but remember that you must complete the template. Do not simply insert a blank template on an image page.

If there is other fair use media, consider checking that you have specified the fair use rationale on the other images used on this page. Note that any fair use images lacking such an explanation can be deleted one week after being tagged, as described on criteria for speedy deletion. If you have any questions please ask them at the Media copyright questions page. Thank you.

BetacommandBot (talk) 23:46, 22 December 2007 (UTC)

GT100 series article?
I'm unsure of whether or not this series should have its own article.

There are very few solid details to be had, but all of the GPUs appear to be rebadged 9 series parts. Charlie from The Inquirer alleges that the rebadging is an attempt to sell defective 9 series GPUs:

http://www.theinquirer.net/inquirer/news/214/1050214/evidence-that-nvidia-renamed-9xxx-gpus-tips-up

That allegation is the most noteworthy thing about the series, hence the question of whether or not it should have its own article. 68.154.235.30 (talk) 06:37, 7 February 2009 (UTC)

Should be added
As shown on the official page on the nVidia site: http://www.nvidia.com/object/geforce_family.html

I also think "GeForce GTX 200" should be renamed "GeForce 200 Series", to be in line with the nVidia website. Lonaowna (talk) 10:43, 7 March 2009 (UTC)

Product Naming Scheme
This section is original research. 58.165.254.91 (talk) 00:29, 28 April 2009 (UTC)
 * Yes, but I don't know where you'd find any up to date, detailed sources. Given the neutral and descriptive nature of the content, I think it's acceptable. And for what it's worth, it's entirely accurate so far as I can tell. —Preceding unsigned comment added by 66.157.206.160 (talk) 18:51, 28 April 2009 (UTC)
 * It's only accurate for a subset of Geforce parts. Perhaps it would be improved by outlining which cards, or ranges thereof, from each generation that fall within each category?  59.167.39.47 (talk) 11:51, 5 May 2009 (UTC)

Hardware T&L in software APIs
The GeForce 256 (NV10) was the first PC graphics chip with hardware transform, lighting, and shading although 3D games utilizing this feature did not appear until later

Now that applies only to Direct3D-based games. All OpenGL-based software uses this feature automatically (think Quake III). Anyone to rewrite this? - Anonymous 06:25, 8 February 2010 (UTC) —Preceding unsigned comment added by 87.207.151.253 (talk)

List of Nvidia video card manufacturers
List of Nvidia video card manufacturers now points just to NVidia. Which is sad, because I came here looking for a list of NVida video card manufacturers. —Preceding unsigned comment added by 203.206.162.148 (talk) 03:34, 6 April 2010 (UTC)
 * You're just as good as any of us to create the article. You'll be more than welcome to. :) --uKER (talk) 04:49, 6 April 2010 (UTC)

Added "In Development" section
Just added this section after poking around to see what is coming up in the pipeline. I hestitated before adding the first bullet about the GTX 585 -- I know this article is about the GeForce product lines in general, and not specific/individual cards. However, I think this warrants mention here because (1) it is an important development in the GeForce product line, but cannot be fit into the GeForce 400 Series bullet section, because it has a different name, and (2) because it has a different name (suggesting a 500 series), it could create confusion and the article can address that confusion.

As for the Kepler and Maxwell series, these are (like the GTX 585) officially announced and I believe are clearly worth mentioning. Mjgilbert (talk) 18:09, 3 November 2010 (UTC)

As of 2009, there have been eleven iterations of the design.
As of 2011, there have been...? --Rogington (talk) 00:12, 29 March 2011 (UTC)

GTX Titan
The GTX Titan is part of the 600 GeForce series. Go to http://www.geforce.com/drivers, if you see the GTX Titan appears in the 600 series section. --Ravotm (talk) 21:31, 21 February 2013 (UTC)

Nouveau driver
Recently, someone added information about the open-source Nouveau driver for Nvidia graphics cards. Then it got removed without a good explanation.

I think there should be a small section about alternative drivers such as Nouveau, with a link to Nouveau (software). Thoughts? --Lonaowna (talk) 18:05, 24 July 2014 (UTC)


 * Hello there! To me, having  section as part of the article is just fine, though I'd say that its  subsection is a bit overdetailed and would look better slightly compacted. &mdash; Dsimic (talk | contribs) 19:10, 24 July 2014 (UTC)


 * I completely agree. I have made some modifications to the section and trimmed down the alternative driver sections. I think the "proprietary" section needs some trimming down as well. I hope you agree with my modifications. --Lonaowna (talk) 19:59, 24 July 2014 (UTC)


 * The whole section now looks good to me, and I'd say that now there's isn't too much detail in any of its two subsections.  I've just cleaned it up a bit, restored the deleted explicit anchor, and restored some of the deleted content together with replacing the associated reference, please .  The rest of the deleted content belongs to the nouveau article anyway, so we should be set –  if you agree. &mdash; Dsimic (talk | contribs) 08:01, 25 July 2014 (UTC)


 * Great, I think it looks good too. Apparently doesn't want to join the discussion. If he keeps deleting I guess we have to keep reverting. Thanks for your effort! --Lonaowna (talk) 08:15, 25 July 2014 (UTC)


 * Oh, thank you for starting it in the first place! :) Yeah, for some reason 175.138.232.130 doesn't want to discuss, but we'll see how's that going to unwind. &mdash; Dsimic (talk | contribs) 20:57, 25 July 2014 (UTC)

I have started a discussion about at Administrators' noticeboard/Incidents, since he keeps removing the section. Lonaowna (talk) 09:24, 27 July 2014 (UTC)


 * Thank you! As we know, having a different opinion is perfectly fine, but discussing the associated changes (and providing references) is the key –  what was completely lacking from 175.138.232.130. &mdash; Dsimic (talk | contribs) 03:08, 28 July 2014 (UTC)

Some time ago Nvidia has started to support Nouveau development by giving them knowledge and code, so that might be added to the article. For examples search "nvidia contributes code to nouveau". 62.61.46.117 (talk) 11:27, 9 September 2014 (UTC)


 * Good suggestion! I have added a small sentence to the article. Lonaowna (talk) 14:19, 9 September 2014 (UTC)


 * ... and I've the section a bit. :) &mdash; Dsimic (talk | contribs) 05:04, 17 September 2014 (UTC)

128 CUDA cores in each SM
There is a statement in the article: "In Pascal, an SM (streaming multiprocessor) consists of 64 CUDA cores, ... ". This statement is not correct. There are 128 CUDA cores in each SM in GeForce GTX 1080.

Please, see in GeForce GTX 1080 Whitepaper:

http://international.download.nvidia.com/geforce-com/international/pdfs/GeForce_GTX_1080_Whitepaper_FINAL.pdf

see "Figure 5: GP104 SM Diagram", and see on page 8: "... Each SM contains 128 CUDA cores, ... ".

Only GP100 has 64 CUDA cores in each SM, but it is not GeForce. It is Tesla. — Preceding unsigned comment added by 91.131.134.230 (talk) 13:33, 3 September 2016 (UTC)

Free and open-source driver
The "Nouveau" driver supports neither openCL or CUDA so it's ineffective for stream computing. 208.127.217.116 (talk) 22:11, 9 December 2016 (UTC)

It's spelled "GeFORCE"
it's GeFORCE not GeForce.

— Preceding unsigned comment added by 88.218.27.9 (talk • contribs) 09:01, 31 August 2006 (UTC)


 * Have you looked at Nvidia's website lately?


 * — Preceding unsigned comment added by Pkaulf (talk • contribs) 14:07, 31 August 2006 (UTC)

the logo in the article reads GeFORCE. i saw the latest logo and saw that now it's GeForce (latest geforce GPU logo).

— Preceding unsigned comment added by 88.218.27.9 (talk • contribs) 12:41, 1 September 2006 (UTC)

Volta redirects here
Volta WAS mentioned as the successor to Maxwell. If you 'rely' on the references provided, that claim was last made in 2013 and was CONTRADICTED (but not retracted) in 2014. Volta was to have stacked DRAM as a key improvement. In 2014 Pascal was announced as the successor to Maxwell AND it has "stacked memory" close to the gpu. So, what is needed is either an edit to remove the redirect or an authoratative reference confirming that as of say 2016 (last year) Pascal will be followed by Volta.40.142.190.18 (talk) 18:51, 15 January 2017 (UTC)

License
On 2 July 2017, Blakegripling ph removed the section on the GEForce License, saying, "Is the licence section even necessary? Almost all proprietary stuff has clauses against reverse-engineering." However the License section had much more than reverse engineering. It covered permission to track users, letting advertisers track users, and ignoring "Do Not Track" signals. This is a sensitive subject, discussed at https://www.ghacks.net/2016/11/07/nvidia-telemetry-tracking/ and elsewhere. Most Wikipedia readers would not expect to be tracked by their own GPU, so it is an important aspect of GeForce's business model, and I propose to reinstate the license section, after allowing a period for others to discuss it here. Numbersinstitute (talk) 01:28, 3 August 2017 (UTC)
 * I updated the article, since there was no further discussion for a week. Numbersinstitute (talk) 16:20, 10 August 2017 (UTC)
 * Apologies if I was a bit brash on removing said statements, but that being said, would it be better to further discuss about the controversy or criticism surrounding it than merely state or quote how Big Brother Nvidia is supposedly spying on us? Blake Gripling (talk) 00:02, 19 September 2017 (UTC)
 * Good idea to add more on the controversy. Important to keep what we have factual, not accusatory. This is a legitimate company choosing a particular business model. I saw the controversy discussed on some blogs, and didn't put them in, since I wasn't sure they'd be called reliable sources. I certainly won't object if someone does put them in. Not sure that putting licensing last is the best place. Perhaps the long list of "Generations" could be put last instead? Few readers will care about the earliest generations, though I certainly think these should stay in the article. They matter to people and museums with older chips or caring about history. Numbersinstitute (talk) 21:27, 24 September 2017 (UTC)

Regarding Volta's placement in the GeForce article
Hi all. This is my first talk post so please bear with my faults.

Getting to the point, I believe Volta should be removed from the GeForce article in some way or another, if it can't be removed entirely. The reason is that Volta wasn't intended for the general consumers (i.e gamers), even if benchmarks showed that it performed better or at least neck-n-neck with the GTX 1080 ti. The clear lack of plans for Nvidia releasing Volta consumer cards, instead opting for Turing to be the real GeForce successor over Pascal, the popular consumer consensus regarding the Titan V and Volta architecture in general and its restricting price tag (to 'prosumers') means Volta shouldn't be placed here as if its a GeForce / consumer successor to Pascal, as its placement in this article implies.

If anyone got any thoughts, ideas and rebuttals for what I say, please do speak up.

P.S I read a little bit of the previous talk section and was surprised that Volta was announced as a successor to Maxwell as opposed to Pascal. TheFledglingLearner (talk) 02:41, 14 September 2018 (UTC)