Talk:GeForce 600 series/Archive 1

Naming
The naming is far from final and there is some suggestion for this being a 700 series. As such this should be renamed NVIDIA Kepler or some such thing, because the only confirmed information is that these will be NVIDIA's kepler GPUs. Rlinfinity (talk) 13:32, 4 January 2012 (UTC)

I'm deleting any and all mentions of "EngineFLUX" from this page unless it is sourced. That name caused enough of a commotion the last time it appeared, which AFAIK was on the GeForce 9 Series page over 4 years ago (may have also been on other pages). 71.64.14.2 (talk) 19:15, 15 January 2012 (UTC)

The naming is as good as final, all the parts in the 600 series are rebadged Fermi for the oem market. — Preceding unsigned comment added by 124.168.11.227 (talk) 04:21, 6 February 2012 (UTC)

I see the kepler believers are citing information from the fake specs lenzfire posted. — Preceding unsigned comment added by 220.235.101.12 (talk) 07:22, 7 February 2012 (UTC)

99.142.36.30 is an idiot, EngineFlux is a term describing change, not a code name for any in development part. the follow up for Kepler is Maxwell as in James Clerk Maxwell 124.149.98.175 (talk) 03:46, 8 February 2012 (UTC)
 * Do not attack other editors (WP:NPA). Until someone provides a reliable source from Nvidia, don't add what the architecture allegedly will or will not be based upon. Edit warring will get you both blocked.  Falcon8765  (T ALK ) 04:01, 8 February 2012 (UTC)
 * Blocking isn't worth crap on a dynamic ip. and there he goes again putting his EngineFlux crap in again. reverted.124.149.98.175 (talk) 04:30, 8 February 2012 (UTC)
 * Yeah, threatening to edit war as you are does not lead to your version being the one included. That's not how things work here. If you actually want to provide reliable sources for your edits, by all means.  Falcon8765  (T ALK ) 04:39, 8 February 2012 (UTC)
 * Being a complete failure at google is why theres so much bogus information on wikipedia, most of it published by the assigned editors.
 * FYI, it was the editors on here that can't discern the difference between page hit fishing and the nvidia geforce product articles that are the edit warrers124.149.98.175 (talk) 04:40, 8 February 2012 (UTC)
 * If it's so easy to Google a reliable source as Wikipedia defines it for your information, why don't you provide one? Also, repeatedly reverting another editor you disagree with is edit warring, so both of you are guilty. If you are actually interested in contributing constructively by conforming with Wikipedia's policies and guidelines, you might actually get the article improved.  Falcon8765  (T ALK ) 04:45, 8 February 2012 (UTC)
 * Posting hex dumps and inf snippets of the driver violates copyright. Posting a Geforce 600/Kepler article was as good as violation of the wikipedia verifiability clause in the first place since no parts and Device strings are obtainable.124.149.98.175 (talk) 04:51, 8 February 2012 (UTC)
 * Hence my removal of any architecture listing. Your own analysis of hex dumps and inf snippets of the driver are original research and thus wouldn't be usable as a source and third-party analysis would be conjecture.  Falcon8765  (T ALK ) 04:57, 8 February 2012 (UTC)

This generation is certainly called "Kepler", and I don't know where the "EngineFLUX" thing came from, but the next generation is called "Maxwell". See here, and a more "citable" source here. --uKER (talk)
 * Thank you for the source, feel free to add and cite the information onto the article.  Falcon8765  (T ALK ) 19:28, 8 February 2012 (UTC)
 * Series =! Architecture. Kepler is not confirmed on the 600 series. put it back to the previous page. 124.168.151.151 (talk) 08:14, 10 February 2012 (UTC)
 * Do you have a source that contradicts the one posted above? Falcon8765  (T ALK ) 18:15, 10 February 2012 (UTC)
 * This is an encyclopedic site, rumourmill and a random website called semiaccurate are not a valid source. — Preceding unsigned comment added by 124.169.28.178 (talk) 13:39, 12 February 2012 (UTC)
 * 124.168 is correct, the source you gave in no way means that even one chip in the 600 series will be Kepler. In fact, it seems like a sizable proportion of the 600 numbers will be rebadged Fermis. So, Falcon8765's comment "Do you have a source that contradicts the one posted above?" is not valid because that source doesn't indicate that Kepler = 600 series to begin with. I won't change this article to what we do know about the 600 series (a number of low-end mobile parts), because then it will inevitably be reverted by those who don't actually know anything about the subject at hand. — Preceding unsigned comment added by 140.254.179.213 (talk) 17:29, 24 February 2012 (UTC)

The EngineFLUX vandal is back again, please permanently ban these IPs, 114.249.138.85, 114.249.142.245, 114.246.160.244, 114.249.136.239 and the whole 114 IP range. — Preceding unsigned comment added by 60.54.26.6 (talk) 10:38, 21 April 2012 (UTC)
 * The EngineFlux vandal keeps trying to edit my post to make as if I wrote things which is not what I wrote. — Preceding unsigned comment added by 60.51.57.123 (talk) 07:49, 25 April 2012 (UTC)
 * Just looked up the IP block, the vandal is in China. Seriously, just ban all of China from vandalising Wikipedia.

Yea right.Nvidia has already comfirm the 600 series its Kepler.It doest'n get any righter than that.And FYI when did the press comfirm it,if they ask nvidia about it then they can comfirm it.Otherwise it BS.. — Preceding unsigned comment added by Cretman121 (talk • contribs) 06:29, 27 April 2012 (UTC)

Typical China vandals, it's obvious the country doesn't teach them any manners or common sense. — Preceding unsigned comment added by 60.53.115.149 (talk) 08:24, 27 April 2012 (UTC)

DX11.1 ?
i'm totally confused about the support of DX11.1, because i did not see any official nvidia page showing DX11.1 support. There is always written DX11. Can anyone name sure sources? — Preceding unsigned comment added by 129.69.232.162 (talk) 07:57, 23 March 2012 (UTC)

Yes, that is true that nVidia don't report about full support DirectX 11.1. All I know that it almost support, but not supported some not gaming function
 * Target-Independent Rasterization (2D rendering only)
 * 16xMSAA Rasterization (2D rendering only)
 * Orthogonal Line Rendering Mode
 * UAV in non-pixel-shader stages

So 60.54.29.138 please show link that confirm full support of DirectX 11.1, in my editing I show that it support only DirectX 11, and describe why not full support of DirectX 11.1


 * The GPUs only support Direct3D 11.0 and therefore DirectX 11.0 ONLY. Kapitaenk (talk) 21:23, 20 February 2013 (UTC) — Preceding unsigned comment added by 91.64.206.196 (talk)

Graphics :) -- Scientific computing :(
Plan on waiting if you plan to use it for scientific computing. — Preceding unsigned comment added by 208.127.16.186 (talk) 22:00, 27 March 2012 (UTC)

OpenCL 1.2
I'm confused about the mentioning of OpenCL 1.2 - can anyone confirm that these cards really support OpenCL 1.2 as of now? Of course they will do so (most probably along with older cards), as soon as Nvidia publishes drivers supporting it, but as of now I'm pretty sure there are no official OpenCL 1.2 drivers from Nvidia, but still only OpenCL 1.1, so I'd say the OpenCL version should be changed to 1.1 untill those drivers exist... — Preceding unsigned comment added by 92.229.83.171 (talk) 19:57, 10 April 2012 (UTC)

See for example http://developer.nvidia.com/opencl - OpenCL 1.2 would be mentioned there if there was any support so far, wouldn't it? — Preceding unsigned comment added by 92.229.83.171 (talk) 20:02, 10 April 2012 (UTC)

This is true, see for example http://www.planet3dnow.de/vbulletin/showthread.php?t=405690 or http://forums.developer.nvidia.com/devforum/discussion/7401/when-will-there-be-opencl-1-2-drivers/p1 - you can also download the NVIDIA GPU Computin SDK and check for yourself. It seems though, that NVIDIA tries to keep information regarding this to a minimum (for example they removed the mentioning of any version number from http://developer.nvidia.com/opencl). — Preceding unsigned comment added by 212.255.35.165 (talk) 23:14, 25 June 2012 (UTC)

Red Screen of Death
It's been reported some of the GTX 680 wasn't running stably, causing Red Screen of Death. Could this be another 8 Series overheating problem? --Novadust (talk) 09:51, 13 April 2012 (UTC)
 * If there was reliable coverage of it (no, forum posts don't count), it would be suitable for inclusion. Until then... --uKER (talk) 17:28, 29 April 2012 (UTC)

Vandalism by China Unicom Beijing province network 114 IP user(EngineFlux vandal)
It's time to permanently ban this IP block, China Unicom Beijing province network 114 IP range. There is no EngineFlux, even Nvidia's CEO called the codename as Kepler in this video, http://www.youtube.com/watch?v=iq202LKkeHI — Preceding unsigned comment added by 60.53.115.149 (talk) 08:31, 27 April 2012 (UTC)


 * You are the vandal. The architecture is broadly known to be called Kepler, as it says so all over the very same video you just posted. Also confirmation is the chips' codenames being GK (K for Kepler), just like the previous gen was GF (F for Fermi) and the previous one GT (T for Tesla). As JSH says, all architectures in the current roadmap are called after scientists. You shouldn't expect an architecture to be called EngineFlux any time soon. --uKER (talk) 02:39, 28 April 2012 (UTC)

OpenGL 4.3 support information is invalid.
https://developer.nvidia.com/opengl-driver

For example, GTX 660 is claimed here to support OpenGL 4.3, but NVIDIA tells it does not. Having a GTX 660 (GK106) myself, with the latest drivers available installed (downloaded a hour ago, ver 306.97 with date 2012.10.10) i can confirm that OpenGL 4.3 is NOT supported.

Asking someone with more experience at editing the wiki (compared to my "zero") to do the corrections based on the link given. — Preceding unsigned comment added by 62.65.50.36 (talk) 19:58, 1 November 2012 (UTC)

GeForce GT 630 and GT 620
The article states that the GT 630 is a rebranded GT 540 but the latter is not mentioned in the article for the GeForce 500 Series. 181.48.44.66 (talk) 16:59, 16 January 2013 (UTC)

The references given in this topic are most decidedly false. Footnotes 4 and 5 give no information about either the GT620 or the GT630. Moreover, if the GT 620 is a re-branded GT530, why does one contain a GF118 and the other a GF119? I'd be very happy to find a reliable source for that information.

source
somebody wanted to use this: http://nvidia.custhelp.com/app/answers/detail/a_id/3196

the place to discuss it is here. thanks. Choyoołʼįįhí:Seb az86556 > haneʼ 09:05, 2 March 2013 (UTC)

GTX 660 Ti is considered midrange, as the GT64x series?
I don't think they should be together. The GT64x series cards are Fermi refreshes while the 660Ti has the same chip the 670 have (the EXACT same chip). I think the 660Ti should be moved to High Ends or the 64x series moved to the low end. I don't have a Wikipedia account yet, so I don't know if this is proper here. Sorry. — Preceding unsigned comment added by 87.219.89.112 (talk) 18:46, 26 August 2012 (UTC)
 * The memory bus is very different, so no. 3&#124;9&#124;3&#124;0&#124;K (talk) 19:26, 26 September 2012 (UTC)

The GTX660Ti exists between the GTX670 and the GTX660. In terms of performance it's in the middle of the two. However, it was released one month before the GTX660 and has 40% more shaders/cuda-cores. The release price was the same as a GTX470 (a high end GeForce GTX 400 series GPU). To me the GTX660Ti is a high end GeForce GTX 600 card. If you want to split hairs call it a mid-to-high end card. — Preceding unsigned comment added by Skgiven (talk • contribs) 22:29, 13 June 2013 (UTC)

GTX Titan
The GTX Titan is part of the 600 Series. Go to http://www.geforce.com/drivers, if you see the the Titan appears on the 600 series section. --Ravotm (talk) 21:28, 21 February 2013 (UTC)
 * the titan is actually a series in itself, its nor 600 or 700 as far as I know, I would like to propose to make a new article labeled GTX Titan, using the info on the Geforce 700 Page since its basiacly gear towards the titan anyways, I don't want an article move because they a lot of redirecting will happen and it will make 700 series page useless. Matthew Smith (talk) 16:28, 22 February 2013 (UTC)

GTX Titan double precision peak is 1.5 GFLOPs, not 1.3. It should be 1/3 of SP peak. — Preceding unsigned comment added by Marat Dukhan (talk • contribs) 00:30, 24 March 2013 (UTC)

No, it's 1.3, since that number is for the worst-case scenario in which the Titan downclocks to ~700 MHz. — Preceding unsigned comment added by 128.146.3.128 (talk) 14:10, 26 April 2013 (UTC) Agreed Skgiven (talk) 23:03, 13 June 2013 (UTC)

The GTX Titan is basically a Tesla. It's CC3.5, not CC3.0 like the GF GTX 600 GPU range. The GTX 780 is also a trimmed down Tesla, again CC3.5. Skgiven (talk)

Edit request on 1 April 2013
It would be good if someone updated this page with the new "GTX 650 Ti BOOST" model. Here is the data for the various fields in the chart:

Name: GeForce GTX 650 Ti BOOST Launch: March 26, 2013 Code name: GK106 Fab (nm): 28 Transistors (million): 2540 Die Size (mm²): 221 Die Count: 1 Bus interface; PCIe 3.0 x16 Memory (MiB): 1024, 2048 SM count: 4 Config core: 768, 64, 24 Core (MHz): 980 Average Boost (MHz): 1033 Max Boost: 1084 Shader (MHz): 980 Memory (MHz): 6000 Pixel (GP/s): 23.5 Texture (GT/s): 62.7 Bandwidth (GB/s): 144.2 DRAM type: GDDR5 Bus width (bit): 192 DirectX: 11 OpenGL: 4.3 OpenCL: 1.2 GLOPS (FMA): 1505.3 TDP (watts): 134 GFLOP/W: 11.23 Release Price (USD): 169

Source: http://www.hardwareluxx.com/index.php/reviews/hardware/vgacards/21922-test-evga-geforce-gtx-650-ti-boost-superclocked.html?start=1

66.189.181.92 (talk) 15:10, 1 April 2013 (UTC)
 * A couple questions: First, where in the table should it go? Right after GeForce GTX 650 Ti? Or does it go farther down the list? And second, does the infobox at the top of the article need to be updated to include the 650 Boost in one of the range categories? — Bility (talk) 17:44, 2 April 2013 (UTC)
 * Red information icon with gradient background.svg Not done for now: not response to question. Mdann52 (talk) 12:38, 12 April 2013 (UTC)

There are 1GB GT630 Rev. 2 editions, http://www.techpowerup.com/gpudb/b2023/zotac-gt-630-zone-edition-1-gb.html http://www.techpowerup.com/gpudb/b2058/zotac-gt-630-guild-wars-1-gb.html Skgiven (talk) 22:58, 13 June 2013 (UTC)


 * Red information icon with gradient background.svg Not done: please make your request in a "change X to Y" format. Mdann52 (talk) 12:49, 26 June 2013 (UTC)


 * I found this series GTX 650 ( note -there is no GT 650 MX series ) on phoronix .. perhaps someone can find out the best information there .?
 * http://www.phoronix.com/scan.php?page=article&item=nvidia_geforce_gtx650&num=4
 * Linuxchatter (talk) 18:53, 1 December 2013 (UTC)

Regarding Direct3D 12.0 support
Please see http://techreport.com/news/26210/directx-12-will-also-add-new-features-for-next-gen-gpus before you state that the GeForce 400, 500, 600, and 700 series fully support Direct3D 12 or DirectX 12. They do not support the full Direct3D 12.0 standard because that standard will support features that were not invented when those GPU series were built. Nvidia's blog was technically true but quite misleading. While those GPUs will run Direct3D 12, they will run that API at a reduced feature level according to the article I cited. Jesse Viviano (talk) 03:46, 10 May 2014 (UTC)


 * For some reason, a few IP addresses keep a constant on this, together with deleting pictures of graphics cards.  By the way, what's wrong with those pictures?  Let's discuss first, please, in order to see whether we agree on the provided references etc. &mdash; Dsimic (talk | contribs) 07:20, 20 May 2014 (UTC)

OpenGL support
Regarding, profanity isn't a substitute for reliable sources. Please, 210.187.216.58, be a team player and provide an updated reference for OpenGL 4.5 support – that will make everybody happy. Thank you. &mdash; Dsimic (talk | contribs) 13:25, 19 December 2014 (UTC)


 * I'm not sure how good is the reference provided in, but what I'm quite sure is that investing more of my time into this simply isn't worth it. 4.4 or 4.5, and 11.0 or 12.0 –  in the end, who cares about those version numbers? &mdash; Dsimic (talk | contribs) 14:57, 19 December 2014 (UTC)


 * Please keep in mind that the referenced version 347.09 of Nvidia drivers is still a beta. &mdash; Dsimic (talk | contribs) 15:02, 19 December 2014 (UTC)

http://www.nvidia.com/download/driverResults.aspx/80913/en-us OpenGL 4.5 support is now in mainline drivers, not developer drivers. Stop editing if you're a clueless person that doesn't know how to search for anything or keep up with Nvidia drivers news. — Preceding unsigned comment added by ‎210.187.216.58 (talk) 18:08, 19 December 2014‎ (UTC)


 * What's wrong with you? The page you've linked above also clearly confirms that the version 347.09 of Nvidia drivers is a beta, and even uses red color to emphasize that. &mdash; Dsimic (talk | contribs) 18:16, 19 December 2014 (UTC)


 * The article is now semi-protected . &mdash; Dsimic (talk | contribs) 21:57, 19 December 2014 (UTC)

http://www.nvidia.com/download/driverResults.aspx/80913/en-us

"347.09 WHQL"

Driver is now WHQL certified,  — Preceding unsigned comment added by 124.82.32.39 (talk) 17:35, 23 December 2014 (UTC)


 * Thank you for pointing that out. With the version 347.09 of Nvidia drivers promoted from beta to production-ready status, the article  to reflect the change in OpenGL 4.5 support. &mdash; Dsimic (talk | contribs) 15:10, 24 December 2014 (UTC)