Report an Error

NVIDIA GeForce RTX 3080 20 GB

Graphics Processor
GA102
Cores
8704
TMUs
272
ROPs
96
Memory Size
20 GB
Memory Type
GDDR6X
Bus Width
320 bit
GPU Chip
GPU
I/O
I/O
Top
Top
Bottom
Bottom
Back
Back
This product is not released yet.
Data on this page may change in the future.
The GeForce RTX 3080 20 GB will be an enthusiast-class graphics card by NVIDIA, that is expected to launch in 2020. Built on the 8 nm process, and based on the GA102 graphics processor, in its GA102-200-KD-A1 variant, the card supports DirectX 12 Ultimate. This ensures that all modern games will run on GeForce RTX 3080 20 GB. Additionally, the DirectX 12 Ultimate capability guarantees support for hardware-raytracing, variable-rate shading and more, in upcoming video games. The GA102 graphics processor is a large chip with a die area of 628 mm² and 28,300 million transistors. Unlike the fully unlocked GeForce RTX 3090, which uses the same GPU but has all 10496 shaders enabled, NVIDIA has disabled some shading units on the GeForce RTX 3080 20 GB to reach the product's target shader count. It features 8704 shading units, 272 texture mapping units, and 96 ROPs. Also included are 272 tensor cores which help improve the speed of machine learning applications. The card also has 68 raytracing acceleration cores. NVIDIA has paired 20 GB GDDR6X memory with the GeForce RTX 3080 20 GB, which are connected using a 320-bit memory interface. The GPU is operating at a frequency of 1440 MHz, which can be boosted up to 1710 MHz, memory is running at 1188 MHz (19 Gbps effective).
Being a dual-slot card, the NVIDIA GeForce RTX 3080 20 GB draws power from 1x 12-pin power connector, with power draw rated at 320 W maximum. Display outputs include: 1x HDMI, 3x DisplayPort. GeForce RTX 3080 20 GB is connected to the rest of the system using a PCI-Express 4.0 x16 interface. The card measures 285 mm in length, 112 mm in width, and features a dual-slot cooling solution. Its price at launch will be 799 US Dollars.

Graphics Processor

GPU Name
GA102
GPU Variant
GA102-200-KD-A1
Architecture
Ampere
Foundry
Samsung
Process Size
8 nm
Transistors
28,300 million
Die Size
628 mm²

Graphics Card

Release Date
2020
Availability
2020
Generation
GeForce 30
Production
Unreleased
Launch Price
799 USD
Bus Interface
PCIe 4.0 x16

Relative Performance

1%
GeForce 210
1%
GeForce 9400 GT
1%
Radeon HD 4550
1%
Radeon HD 5450
1%
Radeon HD 6450
2%
GeForce GT 520
2%
GeForce GT 220
2%
GeForce GT 430
3%
Radeon HD 5570
3%
Radeon HD 4670
3%
GeForce GT 440
3%
GeForce GT 240
3%
GeForce 9600 GT
4%
Radeon HD 5670
4%
GeForce GT 640
4%
GeForce 9800 GT
4%
Radeon HD 6670
4%
Radeon HD 4830
5%
Radeon HD 4770
5%
Radeon HD 4850
5%
GeForce GTS 250
5%
GeForce GTS 450
5%
Radeon HD 5750
6%
Radeon HD 7750
6%
Radeon HD 5770
6%
GeForce GTX 260
6%
Radeon HD 4870
6%
Radeon Vega 8
6%
GeForce GTX 550 Ti
6%
GeForce GTX 650
7%
Radeon HD 5830
7%
Radeon HD 6790
7%
Radeon HD 4890
7%
GeForce GTX 460
7%
Radeon RX Vega 11
8%
GeForce GT 1030
8%
GeForce GTX 275
8%
Radeon HD 7770 GHz Edition
8%
GeForce GTX 280
8%
GeForce GTX 465
8%
Radeon HD 6850
8%
GeForce GTX 285
9%
Radeon HD 5850
9%
GeForce GTX 650 Ti
9%
Radeon HD 7790
9%
Radeon RX 550
10%
GeForce GTX 470
10%
Radeon HD 6870
11%
Radeon HD 5870
11%
GeForce GTX 560 Ti
11%
Radeon HD 4870 X2
11%
GeForce GTX 750 Ti
11%
Radeon HD 6950
12%
GeForce GTX 295
12%
GeForce GTX 650 Ti Boost
12%
Radeon HD 7850
12%
GeForce GTX 480
13%
Radeon HD 6970
13%
Radeon R7 265
13%
Radeon RX 460
13%
Radeon R7 370
13%
GeForce GTX 660
13%
GeForce GTX 570
13%
Radeon RX 560
14%
GeForce GTX 950
14%
Radeon HD 7870 GHz Edition
15%
Radeon R9 270X
15%
GeForce GTX 660 Ti
15%
GeForce GTX 580
15%
GeForce GTX 1050
15%
Radeon HD 7950
15%
Radeon HD 5970
16%
GeForce GTX 760
17%
GeForce GTX 670
17%
GeForce GTX 960
18%
Radeon R9 380
18%
Radeon R9 285
18%
Radeon HD 7970
18%
GeForce GTX 680
19%
GeForce GTX 1050 Ti
19%
GeForce GTX 770
20%
Radeon HD 6990
20%
Radeon R9 280X
20%
GeForce GTX 590
20%
Radeon HD 7970 GHz Edition
23%
GeForce GTX 780
23%
GeForce GTX 1650
25%
Radeon RX 470
25%
Radeon R9 290
27%
Radeon RX 570
27%
Radeon R9 390
27%
Radeon R9 290X
27%
GeForce GTX 970
28%
GeForce GTX TITAN
28%
GeForce GTX 780 Ti
29%
Radeon R9 390X
29%
Radeon HD 7990
29%
Radeon RX 480
29%
GeForce GTX 690
30%
GeForce GTX 1060 6 GB
30%
Radeon RX 5500 OEM
31%
Radeon RX 580
31%
Radeon RX 5500 XT
31%
GeForce GTX 980
31%
GeForce GTX 1650 SUPER
32%
Radeon R9 FURY
33%
Radeon RX 590
35%
Radeon R9 295X2
35%
GeForce GTX 1660
35%
Radeon R9 FURY X
36%
GeForce GTX 980 Ti
37%
GeForce GTX TITAN X
39%
GeForce GTX 1660 SUPER
40%
GeForce GTX 1660 Ti
40%
GeForce GTX 1070
43%
Radeon RX Vega 56
45%
GeForce GTX 1070 Ti
47%
Radeon RX Vega 64
47%
GeForce RTX 2060
47%
GeForce GTX 1080
49%
Radeon RX 5700
53%
GeForce RTX 2060 SUPER
55%
GeForce RTX 2070
56%
Radeon RX 5700 XT
57%
Radeon VII
58%
TITAN X Pascal
60%
GeForce GTX 1080 Ti
61%
GeForce RTX 2070 SUPER
65%
GeForce RTX 2080
69%
GeForce RTX 2080 SUPER
76%
GeForce RTX 2080 Ti
100%
GeForce RTX 3080
100%
GeForce RTX 3080 20 GB
110%
GeForce RTX 3090
Based on TPU review data: "Performance Summary" at 1920x1080, 4K for 2080 Ti and faster.
Performance estimated based on architecture, shader count and clocks.

Clock Speeds

Base Clock
1440 MHz
Boost Clock
1710 MHz
Memory Clock
1188 MHz
19 Gbps effective

Memory

Memory Size
20 GB
Memory Type
GDDR6X
Memory Bus
320 bit
Bandwidth
760.3 GB/s

Render Config

Shading Units
8704
TMUs
272
ROPs
96
SM Count
68
Tensor Cores
272
RT Cores
68
L1 Cache
128 KB (per SM)
L2 Cache
5 MB

Theoretical Performance

Pixel Rate
164.2 GPixel/s
Texture Rate
465.1 GTexel/s
FP16 (half) performance
29.77 TFLOPS (1:1)
FP32 (float) performance
29.77 TFLOPS
FP64 (double) performance
465.1 GFLOPS (1:64)

Board Design

Slot Width
Dual-slot
Length
285 mm
11.2 inches
Width
112 mm
4.4 inches
TDP
320 W
Suggested PSU
700 W
Outputs
1x HDMI
3x DisplayPort
Power Connectors
1x 12-pin
Board Number
PG132 SKU 30

Graphics Features

DirectX
12 Ultimate (12_2)
OpenGL
4.6
OpenCL
2.0
Vulkan
1.2
CUDA
8.5
Shader Model
6.5