NVIDIA A100
The NVIDIA A100 is a Data Center graphics card based on the Ampere architecture. Released on 2020-05-14 with a launch price of $10,000.00, it features 6,912 CUDA cores and 40 GB of HBM2e memory.
General info about A100
| Architecture | Ampere (GA100) |
| Market Segment | Data Center |
| Release Date | 2020-05-14 |
| Launch Price | $10,000.00 |
| Manufacturing Process | 7nm |
Technical specs of A100
| CUDA Cores | 6,912 |
| Base Clock Speed | 1410 MHz |
| Boost Clock Speed | 1410 MHz |
| Transistor Count | 54.2B |
| VRAM Capacity | 40 GB HBM2e |
| Memory Bus Width | 5120 bits |
| Memory Bandwidth | 1555.0 GB/s |
| TDP | 250 W |
Key Technical Parameters
Key technical parameters of the NVIDIA A100 include its 6,912 CUDA cores, 1410 MHz base clock, and 1410 MHz boost clock, delivering high performance for data center applications.
Benchmark performance of A100
FP32 Performance
19.5
TFLOPS
FP16 Performance
312
TFLOPS
The NVIDIA A100 achieves 19.5 TFLOPS in FP32 performance, making it ideal for data center workloads.
Gaming performance of A100
Gaming Performance
The NVIDIA A100 is not designed for gaming. It is optimized for AI training, machine learning, and high-performance computing workloads.
Features and connectivity of A100
Tensor Cores
Gen 3
Transformer Engine
Disabled
Flash Attention
Not Supported
Connectivity
NVLink 3, PCIe Gen4
Physical dimensions of A100
| Length | 10.5in |
| Width | 4.4in |
| Height | 2-slot |
Rankings and efficiency of A100
| Ranking Position | #3 |
| Popularity Ranking | #3 |
| Cost Effectiveness | 1.0/5 |
| Power Efficiency | 1.3/5 |
Cloud Availability
Available Instances
1
Starting Price
$1.29/hr
Top GPUs for Training and Inference
| Category | Rank 1 | Rank 2 | Rank 3 |
|---|---|---|---|
| Best for Training | NVIDIA H200 | NVIDIA H100 | NVIDIA B200 |
| Best for Inference | NVIDIA A40 | NVIDIA A100 | NVIDIA A10 |
Compare GPU specifications and cloud instances to find the best GPU for your workload.