Report an Error

NVIDIA A100 SXM4

Graphics Processor
GA100
Cores
6912
TMUs
432
ROPs
160
Memory Size
40 GB
Memory Type
HBM2e
Bus Width
5120 bit
The A100 SXM4 is a professional graphics card by NVIDIA, launched in May 2020. Built on the 7 nm process, and based on the GA100 graphics processor, the card supports DirectX 12 Ultimate. The GA100 graphics processor is a large chip with a die area of 826 mm² and 54,200 million transistors. It features 6912 shading units, 432 texture mapping units, and 160 ROPs. Also included are 432 tensor cores which help improve the speed of machine learning applications. NVIDIA has paired 40 GB HBM2e memory with the A100 SXM4, which are connected using a 5120-bit memory interface. The GPU is operating at a frequency of 1410 MHz, memory is running at 1215 MHz.
Its power draw is rated at 400 W maximum. This device has no display connectivity, as it is not designed to have monitors connected to it. A100 SXM4 is connected to the rest of the system using a PCI-Express 4.0 x16 interface.

Graphics Processor

GPU Name
GA100
Architecture
Ampere
Foundry
TSMC
Process Size
7 nm
Transistors
54,200 million
Die Size
826 mm²

Graphics Card

Release Date
May 14th, 2020
Generation
Tesla
(Axx)
Production
Active
Bus Interface
PCIe 4.0 x16

Relative Performance

1%
GeForce 210
1%
GeForce 9400 GT
1%
Radeon HD 4550
2%
Radeon HD 5450
2%
Radeon HD 6450
2%
GeForce GT 520
3%
GeForce GT 220
4%
GeForce GT 430
4%
Radeon HD 5570
4%
Radeon HD 4670
4%
GeForce GT 440
4%
GeForce GT 240
5%
GeForce 9600 GT
5%
Radeon HD 5670
6%
GeForce GT 640
6%
GeForce 9800 GT
6%
Radeon HD 6670
6%
Radeon HD 4830
7%
Radeon HD 4770
7%
Radeon HD 4850
7%
GeForce GTS 250
8%
GeForce GTS 450
8%
Radeon HD 5750
9%
Radeon HD 7750
9%
Radeon HD 5770
9%
GeForce GTX 260
9%
Radeon HD 4870
9%
Radeon Vega 8
10%
GeForce GTX 550 Ti
10%
GeForce GTX 650
10%
Radeon HD 5830
11%
Radeon HD 6790
11%
Radeon HD 4890
11%
GeForce GTX 460
11%
Radeon RX Vega 11
12%
GeForce GT 1030
12%
GeForce GTX 275
12%
Radeon HD 7770 GHz Edition
12%
GeForce GTX 280
12%
GeForce GTX 465
12%
Radeon HD 6850
13%
GeForce GTX 285
14%
Radeon HD 5850
14%
GeForce GTX 650 Ti
14%
Radeon HD 7790
14%
Radeon RX 550
15%
GeForce GTX 470
15%
Radeon HD 6870
16%
Radeon HD 5870
17%
GeForce GTX 560 Ti
17%
Radeon HD 4870 X2
17%
GeForce GTX 750 Ti
17%
Radeon HD 6950
18%
GeForce GTX 295
18%
GeForce GTX 650 Ti Boost
18%
Radeon HD 7850
18%
GeForce GTX 480
19%
Radeon HD 6970
20%
Radeon R7 265
20%
Radeon RX 460
20%
Radeon R7 370
20%
GeForce GTX 660
20%
GeForce GTX 570
20%
Radeon RX 560
22%
GeForce GTX 950
22%
Radeon HD 7870 GHz Edition
22%
Radeon R9 270X
23%
GeForce GTX 660 Ti
23%
GeForce GTX 580
23%
GeForce GTX 1050
23%
Radeon HD 7950
23%
Radeon HD 5970
24%
GeForce GTX 760
26%
GeForce GTX 670
26%
GeForce GTX 960
27%
Radeon R9 380
27%
Radeon R9 285
28%
Radeon HD 7970
28%
GeForce GTX 680
29%
GeForce GTX 1050 Ti
29%
GeForce GTX 770
30%
Radeon HD 6990
31%
Radeon R9 280X
31%
GeForce GTX 590
31%
Radeon HD 7970 GHz Edition
34%
GeForce GTX 780
36%
GeForce GTX 1650
38%
Radeon RX 470
38%
Radeon R9 290
41%
Radeon RX 570
41%
Radeon R9 390
41%
Radeon R9 290X
42%
GeForce GTX 970
42%
GeForce GTX TITAN
43%
GeForce GTX 780 Ti
44%
Radeon R9 390X
44%
Radeon HD 7990
44%
Radeon RX 480
44%
GeForce GTX 690
46%
GeForce GTX 1060 6 GB
46%
Radeon RX 5500 OEM
47%
Radeon RX 580
47%
Radeon RX 5500 XT
48%
GeForce GTX 980
48%
GeForce GTX 1650 SUPER
48%
Radeon R9 FURY
50%
Radeon RX 590
53%
Radeon R9 295X2
53%
GeForce GTX 1660
54%
Radeon R9 FURY X
55%
GeForce GTX 980 Ti
56%
GeForce GTX TITAN X
60%
GeForce GTX 1660 SUPER
61%
GeForce GTX 1660 Ti
61%
GeForce GTX 1070
65%
Radeon RX Vega 56
69%
GeForce GTX 1070 Ti
71%
Radeon RX Vega 64
72%
GeForce RTX 2060
72%
GeForce GTX 1080
75%
Radeon RX 5700
81%
GeForce RTX 2060 SUPER
84%
GeForce RTX 2070
86%
Radeon RX 5700 XT
87%
Radeon VII
89%
TITAN X Pascal
92%
GeForce GTX 1080 Ti
93%
GeForce RTX 2070 SUPER
99%
GeForce RTX 2080
100%
A100 SXM4
105%
GeForce RTX 2080 SUPER
116%
GeForce RTX 2080 Ti
153%
GeForce RTX 3080
168%
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

GPU Clock
1410 MHz
Memory Clock
1215 MHz
2.4 Gbps effective

Memory

Memory Size
40 GB
Memory Type
HBM2e
Memory Bus
5120 bit
Bandwidth
1,555 GB/s

Render Config

Shading Units
6912
TMUs
432
ROPs
160
SM Count
108
Tensor Cores
432
L1 Cache
192 KB (per SM)
L2 Cache
40 MB

Theoretical Performance

Pixel Rate
225.6 GPixel/s
Texture Rate
609.1 GTexel/s
FP16 (half) performance
77.97 TFLOPS (4:1)
FP32 (float) performance
19.49 TFLOPS
FP64 (double) performance
9.746 TFLOPS (1:2)

Board Design

Slot Width
IGP
TDP
400 W
Suggested PSU
800 W
Outputs
No outputs
Power Connectors
None

Graphics Features

DirectX
12 Ultimate (12_2)
OpenGL
4.6
OpenCL
2.0
Vulkan
1.2
CUDA
8.0
Shader Model
6.5