Talk:GeForce 30 series

Computing Low‑importance

	This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.ComputingWikipedia:WikiProject ComputingTemplate:WikiProject ComputingComputing articles
Low	This article has been rated as Low-importance on the project's importance scale.
	This article is supported by Computer hardware task force (assessed as Mid-importance).

Computer graphics Mid‑importance

	This article is within the scope of WikiProject Computer graphics, a collaborative effort to improve the coverage of computer graphics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.Computer graphicsWikipedia:WikiProject Computer graphicsTemplate:WikiProject Computer graphicscomputer graphics articles
Mid	This article has been rated as Mid-importance on the project's importance scale.

Wiki Education Foundation-supported course assignment

This article was the subject of a Wiki Education Foundation-supported course assignment, between 6 September 2020 and 7 December 2020. Further details are available on the course page. Student editor(s): Martinmadison.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 21:36, 17 January 2022 (UTC)[reply]

FLOPS

The GFLOPS in the table look wrong. Single and half precision definitely don't have the same flops! There should be a factor 4 discrepancy according to the NVIDIA Ampere architecture in depth blog post.^[1]

Maxiantor (talk) 12:32, 23 December 2020 (UTC)[reply]

References

^ https://developer.nvidia.com/blog/nvidia-ampere-architecture-in-depth/

Untitled section

RTX 3050 and RTX 3050 Ti have not been officially confirmed yet and RTX 3080 is a high end GPU and RTX 3090 is the only enthusiast grade GPU! — Preceding unsigned comment added by Kaptainvet (talk • contribs) 14:25, 24 January 2021 (UTC)[reply]

Driver disable

The driver thing isnt worded that well. It should say that the older (dev only IIRC?) driver did not yet have the hash rate limiter. In other words the hash rate limiter was circumvented instead of disabled Anon contributor 375 (talk) 09:02, 16 July 2021 (UTC)[reply]

3050 and 3090 ti reverts

I reverted some unsourced changes to the 3050 and 3090 ti, as I could not find any confirmed information to support the edits. Instead, I found conflicting information regarding the release date and pricing of the 3090 ti:.^[1] To my knowledge, the unsourced changes were based on rumours (see WP:RUMOUR point 5).

After my revert, I made some changes that should not be controversial: I calculated the TFLOPS and GB/s rates from the confirmed data.

An anonymous user reverted my changes, as well as my revert. Could I get some third person's opinion (or discussion with said anonymous user(s)) on the validity of my changes, as well as my revert.

Thank you, Logical user (talk) 12:26, 9 January 2022 (UTC)[reply]

I see the 3090 Ti and "40 Series" have been added. I will remove these on sight unless and until a reliable source not repeating rumors is provided. WP:CRYSTALBALL applies to this type of content and is not allowed here. —Locke Cole • t • c 05:15, 5 March 2022 (UTC)[reply]

References

^ [1]

RTX 3090 Ti already officially announced

https://www.youtube.com/watch?v=6KoscQVKpAQ&t=1933s

https://www.anandtech.com/show/17173/nvidia-briefly-teases-geforce-rtx-3090-ti-more-details-later-this-month

https://images.anandtech.com/doci/17173/CES_2022_RTX_3090_Ti.jpg

(Personal attack removed) — Preceding unsigned comment added by 42.190.137.74 (talk) 2022-03-06T10:39:01 (UTC)

They announced it at CES but nothing since. They promised to provide additional information by the end of January but they missed that. Now we're back to rumors. This is precisely why WP:CRYSTALBALL exists, to avoid companies making paper launches of products that they have no intention or capability of delivering on. #5 from WP:CRYSTALBALL applies here: Wikipedia is not a collection of product announcements and rumors. You might also want to familiarize yourself with WP:NPA before calling people names. Or you can be permablocked. I don't care either way. =) —Locke Cole • t • c 21:09, 6 March 2022 (UTC)[reply]

RTX 3090 Ti is real

https://www.nvidia.com/en-us/geforce/news/geforce-rtx-3090-ti-out-now/ https://www.nvidia.com/en-us/geforce/graphics-cards/30-series/rtx-3090-3090ti/

(Personal attack removed) — Preceding unsigned comment added by 42.190.164.122 (talk) 13:11, 29 March 2022 (UTC)[reply]

Using "unofficial" card names inside infobox (and/or tables)

I noticed this on other articles, I kinda rolled with it, but after recent edit, I have to ask.

What is the point of 3050 Ti, 3060 Ti GA103 and 3080 (12 GB) on cards list?

Subquestion#1: Are mobile/laptop GPUs considered in same class as desktop GPUs (Mid-range, High-end,...) or not part of a list?

1) No desktop 3050 Ti, 3050 Ti alone is confusing without mobile or laptop specification.See [nvidia 30 series laptop list] and [notebookcheck - device name on images]

2) 3060 Ti GA103 is not official name.GA103 is mentioned inside table (same with 3060 GA104), but both remain as 3060 and 3060 Ti.

3) 3080 (12 GB) also not official name (unlike with 40 series), it's just 3080 for both versions.See [nvidia 30 series desktop list] and [nvidia 3080 family]

Subquestion#2:Since 3080 "(10 GB)" and 3080 (12 GB) are on the list, why not include mobile 3080 (8 GB) and 3080 (16 GB)? If 3050 Ti is on the list, we can have up to 4x 3080s.(But...please don't) Rando717 (talk) 20:59, 28 September 2022 (UTC)[reply]

Forgot to leave a reply... I removed unofficial cards from infobox, also merged same models inside table (I must admit, it's little bit jam-packed...compact but crowded).

However recent silent refresh (3060 8GB, 3060 Ti GDDR6X, 3070 Ti GA102) is potential return of unofficial names.

All of them are already posted on Asia/China region from vendors like Zotac, Asus, Colorful.For now no updates on Nvidia site (Not a global refresh?).

Adding them inside table (under current models) is gonna be little tricky (especially both G6X transfer speeds), unless we decide to use only effective speeds of GDDR6X (for all models). Rando717 (talk) 05:56, 27 October 2022 (UTC)[reply]

Looks like Nvidia updated 3060 family product page and officially confirmed 2/3 with new driver post: [Geforce News]. Rando717 (talk) 18:53, 28 October 2022 (UTC)[reply]

[...], unless we decide to use only effective speeds of GDDR6X (for all models).
— User:Rando717

The whole List of Nvidia graphics processing units article denotes GDDR6X memory transfer speeds with T/s as in "bits transmitted per second" not as in "symbols transmitted per second". The "9.5 GT/s" numbers now proclaimed here, as opposed to the "19000 MT/s" cited in the other article(s), are highly confusing/misleading in the given context. The questionably worded (and more or less obscured) footnote "Cards with GDDR6X send two bits per transfer" doesn't help much with that anyway. Where did either of those come from? Which of the referenced sources specify "9.5 GT/s"? This can't both be right, the prevalent inconsistency must be eliminated in one way or the other. Also, any article about SDRAM defines "transfers per second" as "bits per pin per second" (effective data rate) as well, not as "cycles (or whatever) per pin per second". WPmurphy (talk) 03:39, 24 November 2023 (UTC)[reply]

[1] ttps://developer.nvidia.com/blog/nvidia-ampere-architecture-in-depth/

[2] [1]

[1]

[1]