GeekRadar / The comparison desk

COMPARE.

GPU

NVIDIA

GRID A100B

Graphics processor

Video memory: 48 GB^*
Memory bus: 6,144 bit^*
Rated memory bandwidth: 1,870 GB/s^*

No matched benchmark results are available for this entry yet.

Processing & platform

Architecture: Ampere^*
Process node: 7 nm^*
GPU cores: 6,912^*
GPU base clock: 900 MHz^*
Reported GPU frequency: 900 MHz^*
Graphics interface: PCIe 4.0 x16^*
Reported chip die area: 826 mm²^*
Chip/package TDP: 400 W^*
Transistors: 54,200 million^*
GPU boost clock: 1,005 MHz^*
GPU L2 cache: 48 MB^*
Graphics processor: GA100^*
Texture mapping units: 432^*
Raster output units: 192^*
Streaming multiprocessors: 108^*
Tensor cores: 432^*
Chip foundry: TSMC^*
Card or module format: IGP^*
Declared GPU TDP: 400 W^*
Suggested system PSU: 800 W^*
External power connectors: None^*
OpenCL: 3.0^*
CUDA compute capability: 8.0^*

Memory & storage

Memory options: 48^*
Video memory: 48 GB^*
Video memory type: Reported value 1: HBM2e^*
Reported value 2: HBM2e^*
Reported value 3: GDDR5X^*
Reported value 4: HBM2e^*
Reported value 5: HBM2e^*
Reported value 6: HBM2e^*
Memory bus: 6,144 bit^*
Rated memory bandwidth: 1,870 GB/s^*

Power specifications

Thermal design power: 400 W^*

System configuration

Detailed specifications

Specifications 1 · 8 fields

Product Name: GRID A100B^*
GPU Chip: GA100^*
Released: May 14th, 2020^*
Bus: PCIe 4.0 x16^*
Memory: 48 GB, HBM2e, 6144 bit^*
GPU clock: 900^* MHz
Memory clock: 1,215^* MHz
Shaders TMUs ROPs: 6912 / 432 / 192^*

Specifications 2 · 13 fields

Foundry: TSMC^*
Vendor: NVIDIA^*
FP16 GFLOPS: 55,570^* GFLOPS
FP32 GFLOPS: 13,890^* GFLOPS
FP64 GFLOPS: 6,947^* GFLOPS
Product: NVIDIA GRID A100B^*
Type: GPU^*
Release Date: 2020-05-14^*
Process Size (nm): 7^* nm
TDP (W): 400^* W
Die Size (mm^2): 826^* mm²
Transistors (million): 54,200^*
Freq (MHz): 900^* MHz

Specifications 3 · 29 fields

proc transistors: 54,200 million^*
proc die size: 826^* mm²
cores: 6,912^*
tensor cores: 432^*
sm count: 108^*
base clock: 900^*
boost clock: 1,005^*
memory clock: 1,215^*
ram: 48^*
name: NVIDIA GRID A100B^*
ram type: HBM2e^*
bus width: 6,144^*
bus interface: PCIe 4.0 x16^*
ram bandwidth: 1.87 TB/s^*
fp16 performance: 55.57^*
fp32 performance: 13.89^*
fp64 performance: 6.947^*
tdp: 400^* W
suggested psu: 800^* W
slot width: IGP^*
api cuda: 8^*
is retail board: No^*
gpi value: 19.92^*
vendor: nvd^*
manufacturer: NVIDIA^*
arch name: Ampere^*
card release date: 2020-05-14^*
proc foundry: TSMC^*
proc process size: 7^* nm

Specifications 4 · 35 fields

CUDA compute capability

8.0^*

OpenCL

3.0^*

Architecture

Ampere^*

Graphics interface

PCIe 4.0 x16^*

Memory bus

6,144^* bit

Reported shader processors

6,912^*

Reported theoretical FP16 peak

55.57^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Reported theoretical FP32 peak

13.89^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Reported theoretical FP64 peak

6.947^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Declared L1 cache configuration

192 KB (per SM)^*

Declared GPU L2 cache

48 MB^*

Reported display outputs

No outputs^*

Theoretical pixel fill rate

193.0 GPixel/s^*

External power connectors

None^*

Reported chip die area

826 mm²^*

Chip foundry

TSMC^*

Process node declaration

7 nm^*

Transistors

54,200 million^*

Graphics processor

GA100^*

Reported video memory

48^* GB

Rated memory bandwidth

1.87 TB/s^*

Video memory type

HBM2e^*

Raster output units

192^*

Reported shading units

6,912^*

Card or module format

IGP^*

Streaming multiprocessors

108^*

Suggested system PSU

800 W^*

2:4 tensor sparsity

Yes^*

Declared GPU TDP

400 W^*

Tensor cores

432^*

Theoretical texture fill rate

434.2 GTexel/s^*

Texture mapping units

432^*

GPU cores

6,912^*

Video memory

48^* GB

Video memory type

HBM2e^*

Specifications 5 · 8 fields

Product Name: GRID A100B^*
GPU Chip: GA100^*
Released: May 14th, 2020^*
Bus: PCIe 4.0 x16^*
Memory: 48 GB, HBM2e, 6144 bit^*
GPU clock: 900^* MHz
Memory clock: 1,215^* MHz
Shaders / TMUs / ROPs: 6912 / 432 / 192^*

Specifications 6 · 47 fields

Brand: NVIDIA^*
Top BUS WIDTH: 6,144^* bit
Graphics Processor GPU Name: GA100^*
Graphics Processor Architecture: Ampere^*
Render Config SM Count: 108^*
Graphics Features CUDA: 8^*
Render Config Tensor Cores: 432^*
Graphics Processor Process Size: 7^* nm
Graphics Processor Transistors: 54,200 million^*
Graphics Processor Die Size: 826^* mm²
Graphics Card Release Date: May 14th, 2020^*
Graphics Card Generation: GRID (Ax)^*
Graphics Card Production: Active^*
Name: GRID A100B^*
Graphics Card Bus Interface: PCIe 4.0 x16^*
Clock Speeds Memory Clock: 1215 MHz 2.4 Gbps effective^*
Memory Memory Size: 48^* GB
Memory Memory Type: HBM2e^*
Memory Memory Bus: 6,144^* bit
Memory Bandwidth: 1.87 TB/s^*
Render Config TMUs: 432^*
Top GRAPHICS PROCESSOR: GA100^*
Render Config ROPs: 192^*
Theoretical Performance Pixel Rate: 193.0 GPixel/s
Theoretical Performance Texture Rate: 434.2 GTexel/s
Board Design Slot Width: IGP^*
Board Design TDP: 400^* W
Board Design Suggested PSU: 800^* W
Board Design Outputs: No outputs^*
Graphics Features OpenCL: 3^*
source file: nvidia2020.json^*
Graphics Processor Foundry: TSMC^*
Graphics Processor Density: 65.6M / mm²^*
Top TMUS: 432^*
Graphics Processor Chip Package: BGA-2743^*
Top ROPS: 192^*
Clock Speeds Base Clock: 900^* MHz
Clock Speeds Boost Clock: 1,005^* MHz
Top CORES: 6,912^*
Render Config Shading Units: 6,912^*
Render Config L2 Cache: 48^* MB
Top MEMORY SIZE: 48^* GB
Theoretical Performance FP32 (float): 13.89^* TFLOPS
Theoretical Performance FP64 (double): 6.947 TFLOPS (1:2)
Render Config L1 Cache: 192 KB (per SM)^*
Theoretical Performance FP16 (half): 55.57 TFLOPS (4:1)
Top MEMORY TYPE: HBM2e^*

Specifications 7 · 27 fields

manufacturer: NVIDIA^*
manufacturer: NVIDIA^*
rop: 192^*
rop: 96^*
igp: No^*
igp: No^*
bus: PCIe 3.0 x16^*
bus: PCIe 4.0 x16^*
memType: HBM2e^*
memType: GDDR5X^*
gpuChip: GA100^*
gpuChip: GP102^*
productName: GRID A100B^*
productName: GRID A100B^*
releaseYear: 2,020^*
releaseYear: 2,020^*
memSize: 48^*
memSize: 48^*
memBusWidth: 6,144^*
gpuClock: 900^*
gpuClock: 1,303^*
memClock: 1,808^*
memClock: 1,215^*
unifiedShader: 6,912^*
unifiedShader: 3,840^*
tmu: 240^*
tmu: 432^*

* Provisional data