PNY NVIDIA A100 40GB HBM2 Passive Ampere GPU
SKU: GPURETAILNA10040GBPCIE
For world-leading performance in AI, data analytics and HPC tasks look no further than the latest NVIDIA Tesla A100 GPU. Based upon the groundbreaking Ampere architecture, the A100 is ideally suited to accelerating data centre platforms. With new technologies such as Multi-Instance GPU (or MIG), you can partition a single GPU into seven isolated GPU instances. Furthermore, the NVIDIA Tesla A100 can also scale up to thousands of GPUs, has NVLINK support for 600GB/s interconnectivity and is PCIe Gen4 compatible, making it now the go-to choice for data scientists.
Out Of Stock (End of Life)
System Configuration
Out Of Stock (End of Life)
Details & Specifications
Details & Specifications
:
| Website Product Image URL | Click Here |
|---|
Cooling:
| Passive Cooling |
1 |
|---|
Dimensions:
| Item Width (CM) |
26.77 |
|---|---|
| Item Depth (CM) |
11.115 |
| Item Weight (Kg) |
1.196 |
GPU:
| GPU Slot Size | |
|---|---|
| GPU Architecture | |
| # CUDA Cores |
5120 |
| NVLINK Support |
1 |
| Int 8 Performance (TFLOPS) |
312 |
| FP32 Performance (TFLOPS) |
19.5 |
| FP64 Performance (TFLOPS) |
9.7 |
| Memory Interface (bit) |
5120 |
Memory:
| Memory Type |
HBM2 |
|---|---|
| Memory Capacity (GB) |
40 |
| Memory Bandwidth (GB/s) |
1600 |
PCIe:
| PCIe Version |
PCIe 4.0 |
|---|
Power:
| Wattage |
250 |
|---|
Warranty:
| Mfr Warranty Length (M) |
36 |
|---|---|
| Mfr Warranty Type |
Return to Base |
* Specifications for older discontinued variants can be view by clicking here