GeekRadar / The comparison desk

COMPARE.

GPU

NVIDIA

NVIDIA A100 PCIe 80 GB

Graphics processor

Video memory: 80 GB^*
Memory bus: 5,120 bit^*
Rated memory bandwidth: 1,940 GB/s^*

No matched benchmark results are available for this entry yet.

Processing & platform

Architecture: Ampere^*
Process node: 7 nm^*
GPU cores: 6,912^*
Graphics interface: PCIe 4.0 x16^*
Reported chip die area: 826 mm²^*
Transistors: 54,200 million^*
GPU L2 cache: 80 MB^*
Graphics processor: GA100^*
Texture mapping units: 432^*
Raster output units: 160^*
Streaming multiprocessors: 108^*
Tensor cores: 432^*
Chip foundry: TSMC^*
Card or module format: Dual-slot^*
Declared GPU TDP: 300 W^*
Suggested system PSU: 700 W^*
External power connectors: 8-pin EPS^*
OpenCL: 3.0^*
CUDA compute capability: 8.0^*

Memory & storage

Memory options: 80^*
Video memory: 80 GB^*
Video memory type: HBM2e^*
Memory bus: 5,120 bit^*
Rated memory bandwidth: 1,940 GB/s^*

Power specifications

Thermal design power: 300 W^*

System configuration

Detailed specifications

Specifications 1 · 20 fields

GPU architecture — NVIDIA A100 80GB PCIe — 80 GB HBM2e: Ampere^*
Theoretical peak: BF16 tensor dense per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 312^* TFLOPS
Theoretical peak: FP16 tensor dense per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 312^* TFLOPS
Theoretical peak: FP32 non-tensor dense per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 19.5^* TFLOPS
Native hardware support: FP4 tensor — NVIDIA A100 80GB PCIe — 80 GB HBM2e: No^*
Theoretical peak: FP64 non-tensor dense per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 9.7^* TFLOPS
Theoretical peak: FP64 tensor dense per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 19.5^* TFLOPS
Native hardware support: FP6 tensor — NVIDIA A100 80GB PCIe — 80 GB HBM2e: No^*
Native hardware support: FP8 tensor — NVIDIA A100 80GB PCIe — 80 GB HBM2e: No^*
Theoretical peak: INT8 tensor dense per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 624^* TOPS
Theoretical peak: TF32 tensor dense per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 156^* TFLOPS
Physical configuration — NVIDIA A100 80GB PCIe — 80 GB HBM2e: PCIe, dual-slot air-cooled or single-slot liquid-cooled^*
GPU interconnect bandwidth (NVLink 3) — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 600^* GB/s
PCIe interface declaration — NVIDIA A100 80GB PCIe — 80 GB HBM2e: PCIe Gen4^*
Maximum board power — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 300^* W
Peak memory bandwidth per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 1,935^* GB/s
Memory interface width per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 5,120^* bit
Memory capacity per GPU — NVIDIA A100 80GB PCIe — 80 GB HBM2e: 80^* GB
Memory technology — NVIDIA A100 80GB PCIe — 80 GB HBM2e: HBM2e^*
GPU model — NVIDIA A100 80GB PCIe — 80 GB HBM2e: NVIDIA A100 80GB PCIe^*

Specifications 2 · 32 fields

proc transistors: 54,200 million^*
proc die size: 826^* mm²
cores: 6,912^*
tensor cores: 432^*
sm count: 108^*
base clock: 1,065^*
boost clock: 1,410^*
memory clock: 1,512^*
ram: 80^*
name: NVIDIA A100 PCIe 80 GB^*
ram type: HBM2e^*
bus width: 5,120^*
bus interface: PCIe 4.0 x16^*
ram bandwidth: 1.94 TB/s^*
fp16 performance: 77.97^*
fp32 performance: 19.49^*
fp64 performance: 9.746^*
bf16 performance: 311.84^*
tf32 performance: 155.92^*
tdp: 300^* W
suggested psu: 700^* W
slot width: Dual-slot^*
power connectors: 8-pin EPS^*
api cuda: 8^*
is retail board: No^*
gpi value: 25.63^*
vendor: nvd^*
manufacturer: NVIDIA^*
arch name: Ampere^*
card release date: 2021-06-28^*
proc foundry: TSMC^*
proc process size: 7^* nm

Specifications 3 · 37 fields

CUDA compute capability

8.0^*

OpenCL

3.0^*

Architecture

Ampere^*

Reported theoretical BF16 peak

311.84^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Graphics interface

PCIe 4.0 x16^*

Memory bus

5,120^* bit

Reported shader processors

6,912^*

Reported theoretical FP16 peak

77.97^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Reported theoretical FP32 peak

19.49^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Reported theoretical FP64 peak

9.746^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Declared L1 cache configuration

192 KB (per SM)^*

Declared GPU L2 cache

80 MB^*

Reported display outputs

No outputs^*

Theoretical pixel fill rate

225.6 GPixel/s^*

External power connectors

8-pin EPS^*

Reported chip die area

826 mm²^*

Chip foundry

TSMC^*

Process node declaration

7 nm^*

Transistors

54,200 million^*

Graphics processor

GA100^*

Reported video memory

80^* GB

Rated memory bandwidth

1.94 TB/s^*

Video memory type

HBM2e^*

Raster output units

160^*

Reported shading units

6,912^*

Card or module format

Dual-slot^*

Streaming multiprocessors

108^*

Suggested system PSU

700 W^*

2:4 tensor sparsity

Yes^*

Declared GPU TDP

300 W^*

Tensor cores

432^*

Theoretical texture fill rate

609.1 GTexel/s^*

Reported theoretical TF32 peak

155.92^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Texture mapping units

432^*

GPU cores

6,912^*

Video memory

80^* GB

Video memory type

HBM2e^*

* Provisional data