User:ShuffyIosys/sandbox Source: en.wikipedia.org/wiki/User:ShuffyIosys/sandbox

About this page
This page is to show some examples of tables to propose for the List of Nvidia graphics processing units page. Some of these tables are not complete, they're meant to only provide a template/example.

Some notes about the thought process:

Any stat that can be derived is listed instead its breakdown or another form of it. For example, in GPUs since G80, NVIDIA grouped the shaders into clusers, so you only need to list how many of these clusters are with a note of how many shaders are in each cluster
For memory, prefer transfers/second over clock speed. Also wanted to avoid having the memory type as its own field. Maybe it could be its own model (but I don't have an example for this)
Specific generation features like "third gen ray tracing cores" or "supports transparency" are removed, because that's what the main page is for.
Not really sure what to do with some other secondary/tertiary specs like display connectors, outside of maybe a separate feature table.

Example Tables

Note: This is a proposed table for NVIDIA GPUs before GeForce 8

GeForce 7 (7xxx) series

All models support Direct3D 9.0c and OpenGL 2.1

Model	Launch		Codename	Bus Interface	Core Config ^[a]	Clock Speeds		Memory		SLI
Model	Date	MSRP (USD)	Codename	Bus Interface	Core Config ^[a]	Core Clock (MHz)	Memory (MT/s)	Size (MB)	Bus-Width (Bit)	SLI
GeForce 7300 GT	May 15, 2006		G73	AGP 8x PCIe 1.x x16	8:5:8:4	350	750 (DDR2) 1400 (GDDR3)	128 256	128	No
GeForce 7600GT	March 9, 2006		G72	PCIe 1.x x16	12:5:12:8	560	800 (DDR2) 1400 (GDDR3)	256	128	2-way
GeForce 7600GT	July 15, 2006		G72	AGP 8x	12:5:12:8	560	800 (DDR2) 1400 (GDDR3)	256	128
GeForce 7800 GTX	June 22, 2005		G70	PCIe 1.x x16	24:8:24:16	430	600	256	256
GeForce 7800 GTX	Nov 14, 2005		G70		24:8:24:16	550	850	512	256
GeForce 7950 GX2	June 5, 2006		2xG71		2x 24:8:24:16	500	600	2x 512	256

^ Pixel shaders: vertex shaders: texture mapping units: render output units

Note: This is a proposed table for NVIDIA GPUs from GeForce 900 and after

GeForce 900 series

All models support the following APIs: Direct3D 12_1, OpenGL 4.6, OpenCL 3.0 and Vulkan 1.3 and CUDA 5.2
With the exception of the GeForce GT 945A, all models support 3rd Generation NVENC

Model	Launch		Codename	Bus Interface	SMM ^[a]	Clock Speeds		Memory		Board Power (W)	SLI
Model	Date	MSRP (USD)	Codename	Bus Interface	SMM ^[a]	Core Clock (MHz)	Memory (MT/s)	Size (GB)	Bus-Width (Bit)	Board Power (W)	SLI
GeForce GT 945A	Feb, 2016	OEM	GM108	PCIe 3.0 x8	4	1072 (1176)	1800	1, 2	64	33	No
GeForce GTX 950	Aug 20, 2015	$159	GM206-250	PCIe 3.0 x16	6	1024 (1188)	6600	2	128	75, 90	2-way
GeForce GTX 960	Jan 22, 2015	$199	GM206-300		8	1127 (1178)	7000	2, 4	128	120	2-way
GeForce GTX 970	Sep 18, 2014	$329	GM204-200		13	1050 (1178)		3.5 + 0.5	224 + 32	145	4-way
GeForce GTX 980	Sep 18, 2014	$549	GM204-400		16	1126 (1216)		4	256	165	4-way

^ Stream Multiprocessor, Maxwell (SMM). Each SMM contains 128 FP units and 8 texture mapping units

GeForce 20 series

Supported APIs: Direct3D 12 Ultimate (12_2), OpenGL 4.6, OpenCL 3.0, Vulkan 1.3 and CUDA 7.5
NVENC 6th generation (B-frame, etc.)

Model	Launch		Codename	Bus Interface	SM ^[a]	Core Clock Speeds		Memory			Board Power (W)	SLI
Model	Date	MSRP (USD) ^[b]	Codename	Bus Interface	SM ^[a]	Base (MHz)	Boost (MHz)	Size (MB)	Speed (GT/s)	Bus-Width (Bit)	Board Power (W)	SLI
GeForce RTX 2060	Jan 15, 2019	$349	TU106	PCIe 3.0 x16	30	1365	1680	6	14	192	160	No
	Jan 10, 2020	$299	TU104		30	1365	1680	6			160
	Dec 7, 2020	$299	TU106		34	1470	1650	12			185
GeForce RTX 2060 Super	Jul 9, 2019	$399			34	1470	1650	12		256	175
GeForce RTX 2070	Oct 17, 2018	$499 $599			36	1410	1620	8			175
GeForce RTX 2070 Super	Jul 9, 2019	$499	TU104		40	1605	1770				215	2-Way NVLink
GeForce RTX 2080	Sep 20, 2018	$699 $799			46	1515	1710				215
GeForce RTX 2080 Super	Jul 23, 2019	$699			48	1650	1815				250
GeForce RTX 2080 Ti	Sep 27, 2018	$999 $1199	TU102		68	1305	1545	11	15.5	352	250
TITAN RTX	Dec 18, 2018	$2499	TU102		72	1305	1770	24	14	384	280

GeForce 30 series

Supported APIs: Direct3D 12 Ultimate (12_2), OpenGL 4.6, OpenCL 3.0, Vulkan 1.3 and CUDA 8.6
Supported display connections: HDMI 2.1, DisplayPort 1.4a
NVENC 7th generation

Model	Launch		GPU	Bus Interface	SM ^[c]	Core Clock Speeds		Memory			Board Power (W)
Model	Date	MSRP (USD)	GPU	Bus Interface	SM ^[c]	Base (MHz)	Boost (MHz)	Size (GB)	Speed (GT/S)	Bus-Width (bit)	Board Power (W)
GeForce RTX 3050	Dec 16, 2022	$249	GA107	PCIe 4.0 x8	20	1552	1777	8	14.0	128	115
GeForce RTX 3050	Jan 27, 2022	$249	GA106	PCIe 4.0 x8	20	1552	1777	8	14.0	128	130
GeForce RTX 3060	Oct 27, 2022	$329	GA106	PCIe 4.0 x16	28	1320	1777	12	15.0	192	170
	Feb 25, 2022		GA106
	Sep 1, 2021		GA104
GeForce RTX 3060 Ti	Dec 2, 2020	$399	GA104		38	1410	1665	8	14.0	256	200
GeForce RTX 3060 Ti	Oct 27, 2022	$399	GA104		38	1410	1665		19.0		200
GeForce RTX 3070	Oct 29, 2020	$499	GA104		46	1500	1725		14.0		220
GeForce RTX 3070 Ti	Jun 10, 2021	$599	GA104		48	1575	1770		19.0		290
GeForce RTX 3080	Sep 17, 2020	$699	GA102		68	1440	1710	10		320	320
GeForce RTX 3080	Jan 27, 2022	$799	GA102		70	1260	1710	12		384	350
GeForce RTX 3080 Ti	Jun 3, 2021	$1199	GA102		80	1365	1665	12
GeForce RTX 3090	Sep 24, 2020	$1499	GA102		82	1395	1695	24	19.5
GeForce RTX 3090 Ti	Mar 29, 2022	$1999	GA102		84	1560	1860	24	21.0		450

GeForce 40 series

Supported APIs: Direct3D 12 Ultimate (12_2), OpenGL 4.6, OpenCL 3.0, Vulkan 1.3 and CUDA 8.9<ref>
Supported display connections: HDMI 2.1, DisplayPort 1.4a
8th Generation NVENC

Model	Launch		GPU	Bus Interface	SM ^[c]	Core Clock Speeds		Memory			Board Power (W)
Model	Date	MSRP (USD)	GPU	Bus Interface	SM ^[c]	Base (MHz)	Boost (MHz)	Size (GB)	Speed (GT/S)	Bus-Width (bit)	Board Power (W)
GeForce RTX 4060	Jul 2023	$299	AD107	PCIe 4.0 x8	24	1830	2460	8	17.0	128	115
GeForce RTX 4060 Ti	Mar 24, 2023	$399	AD106		34	2310	2540	8	18.0		160
GeForce RTX 4060 Ti	Jul 2023	$499	AD106		34	2310	2540	16	18.0		160
GeForce RTX 4070	Mar 24, 2023	$599	AD104	PCIe 4.0 x16	46	1920	2475	12	21.0	192	200
GeForce RTX 4070 Ti	Jan 5, 2023	4799	AD104		60	2310	2610				285
GeForce RTX 4080 (12 GB)	Cancelled	$899	AD104		60	2310	2610				285
GeForce RTX 4080	Nov 16, 2022	$1199	AD103		76	2210	2505	16	22.4	256	320
GeForce RTX 4090	Oct 12, 2022	$1599	AD102		128	2230	2520	24	21.0	384	450

^ Stream Multiprocessor (SM). Each SM contains 64 INT units, 64 FP units, 8 tensor cores, 4 texture mapping units, and 1 ray tracing core
^ If present, the second price reflects the Founder's Edition price
^ ^a ^b Stream Multiprocessor (SM). Each SM contains 32 FP/INT units, 32 FP units, 4 tensor cores, 4 texture mapping units, and 1 ray tracing core


Model	Launch Date & Price (USD)	GPU & Fab Process	Transistors & Die Size	Core		Processing Power						Memory					TBP	Bus Interface
Model	Launch Date & Price (USD)	GPU & Fab Process	Transistors & Die Size	Config	Clock Speed (MHz)	Pixel Fillrate (GP/s)	Texture Fillrate (GT/s)	FP16/32 (TFLOPS)	FP64 (TFLOPS)	Tensor (TFLOPS)	Ray Tracing	Type	Size (GB)	Speed (MT/s)	Bandwidth (GB/s)	Width (bits)	TBP	Bus Interface
GeForce RTX 4060	Jul 2023 $299	AD107-400 TSMC 4N	18.9B 158.7mm²	3072:96:48 96:24 (24)(3)	1830 2460	118.1	236.2	11.2 15.1	0.176 0.236	60 121	35	GDDR6	8	17000	272	128	115	PCIe 4.0 x8
GeForce RTX 4060 Ti	Mar 24, 2023 $399	AD106-350 TSMC 4N	22.9B 187.8mm²	4352:136:48 136:34 (34)(3)	2310 2540	121.9	345.4	20.1 22.1	0.314 0.345	88 177	51		8	18000	288		160
GeForce RTX 4060 Ti	Jul 2023 $499	AD106-351 TSMC 4N	22.9B 187.8mm²	4352:136:48 136:34 (34)(3)	2310 2540	121.9	345.4	20.1 22.1	0.314 0.345	88 177	51		16	18000	288		160
GeForce RTX 4070	Mar 24, 2023 $599	AD104-350 TSMC 4N	35.8B 294.5mm²	5888:184:64 184:46 (46)(4)	1920 2475	158.4	455.4	22.6 29.1	0.353 0.455	117 233	67	GDDR6X	12	21000	504	192	220	PCIe 4.0 x16
GeForce RTX 4070 Ti	Jan 5, 2023 $799	AD104-400 TSMC 4N		7680:240:80 240:60 (60)(5)	2340 2610	208.8	626.4	35.5 40.1	0.554 0.627	142 284	92.7						285
GeForce RTX 4080 (12 GB)	Cancelled $899	AD104-400 TSMC 4N		7680:240:80 240:60 (60)(5)	2340 2610	208.8	626.4	35.5 40.1	0.554 0.627		92.7						285
GeForce RTX 4080	Nov 16, 2022 $1199	AD103-400 TSMC 4N	45.9B 378.6mm²	9728:304:112 304:76 (76)(7)	2210 2505	280.6	761.5	43.0 48.8	0.672 0.761	172 344	112.7		16	23000	717	256	320
GeForce RTX 4090	Oct 12, 2022 $1599	AD102-300 TSMC 4N	76.3B 608.5mm²	16384:512:176 512:128 (128)(11)	2235 2520	443.5	1290.2	73.1 82.6	1.142 1.291	292 585	191		24	21000	1008	384	450

[FixedCoreConfig-1] Pixel shaders: vertex shaders: texture mapping units: render output units

[SMMConfig-2] Stream Multiprocessor, Maxwell (SMM). Each SMM contains 128 FP units and 8 texture mapping units

[TUSMConfig-3] Stream Multiprocessor (SM). Each SM contains 64 INT units, 64 FP units, 8 tensor cores, 4 texture mapping units, and 1 ray tracing core

[FENote-4] If present, the second price reflects the Founder's Edition price

[SMConfig-5] Stream Multiprocessor (SM). Each SM contains 32 FP/INT units, 32 FP units, 4 tensor cores, 4 texture mapping units, and 1 ray tracing core

[a]

[a]

[a]

[b]

[c]