NVIDIA A10 vs AMD Instinct MI325X
How the NVIDIA A10 and AMD Instinct MI325X compare across memory, compute, power and form factor, with the stronger figure highlighted on each row. Vendor-specific figures (CUDA vs stream processors, Tensor vs matrix cores) are listed on their own rows.
| Specification |  AMPERE NVIDIA A10 View Product → | CDNA 3 AMD Instinct MI325X NVIDIA alternative: NVIDIA H200 SXM → |
|---|---|---|
| Pricing | ||
| Price | – | Not sold by NVNexus |
| Overview | ||
| Architecture | Ampere | AMD CDNA 3 |
| Memory | ||
| Memory | 24 GB GDDR6 | 256 GB HBM3E |
| Memory Bandwidth | 600 GB/s | 6 TB/s |
| Memory Interface | – | 8192-bit |
| Compute | ||
| Stream Processors (AMD) | – | 19,456 (304 CUs) |
| Matrix Cores / AI Accelerators (AMD) | – | 1,216 |
| FP32 | 31.2 TFLOPS | 163.4 TFLOPS (matrix) |
| FP64 Tensor | – | 163.4 TFLOPS (matrix) |
| TF32 Tensor | 62.5 TFLOPS (125 TFLOPS with sparsity) | 653.7 TFLOPS / 1,307 (sparse) |
| FP16/BF16 Tensor | – | 1,307 TFLOPS / 2,615 (sparse) |
| FP8 Tensor | – | 2,615 TFLOPS / 5,230 (sparse) |
| INT8 Tensor | – | 2,615 TOPS / 5,230 (sparse) |
| Platform | ||
| Interface | PCIe Gen4 x16 | PCIe Gen5 x16 |
| NVLink / Infinity Fabric | – | Infinity Fabric, 896 GB/s aggregate |
| TDP | 150W | 1000W |
| Form Factor | Single-wide, full-height, full-length PCIe, passive | OAM module (8-GPU platform) |
| Cooling | – | Passive (chassis airflow) |
| Transistors | – | 153 billion |
Not sure which fits?
Talk to an Expert → Talk it through with a specialist
Tell us your workload and we will recommend the right GPU, configuration, and sourcing options.
More Comparisons
- NVIDIA A10 vs NVIDIA A100 80GB →
- NVIDIA A10 vs NVIDIA A16 →
- NVIDIA A10 vs NVIDIA A2 →
- NVIDIA A10 vs NVIDIA B200 →
- NVIDIA A10 vs AMD Instinct MI210 →
- NVIDIA A10 vs AMD Instinct MI300X →
- AMD Instinct MI325X vs AMD Instinct MI300X →
- AMD Instinct MI325X vs AMD Instinct MI210 →
- AMD Instinct MI325X vs AMD Instinct MI350X →
- AMD Instinct MI325X vs AMD Instinct MI355X →