AMD Instinct MI325X vs NVIDIA GH200 Superchip — Comparación de GPU (Apr 2026)
AMD Instinct MI325X (256GB HBM3e, 1,307 TFLOPS FP16, CDNA 3) vs NVIDIA GH200 Superchip (96GB HBM3, 989 TFLOPS FP16, Hopper). Cloud pricing: AMD Instinct MI325X from $2.00/hr. Compare specs, VRAM, performance, and pricing across 1 cloud providers to find the best GPU for your AI workload.
|
AMD Instinct MI325X
256GB HBM3e · CDNA 3
|
NVIDIA GH200 Superchip
96GB HBM3 · Hopper
|
||
|---|---|---|---|
| Especificaciones | |||
| Fabricante | AMD | NVIDIA | |
| Arquitectura | CDNA 3 | Hopper | |
| VRAM | 256 GB HBM3e | 96 GB HBM3 | |
| Ancho de Banda | 6,000 GB/s | 4,000 GB/s | |
| FP16 (Tensor) | 1307.0 TFLOPS | 989.0 TFLOPS | |
| FP32 | 163.4 TFLOPS | 494.5 TFLOPS | |
| TDP | 1000W | 700W | |
| Año de Lanzamiento | 2024 | 2023 | |
| Segmento | Data center | Data center | |
| Mejor Para | AI training large model inference | Large-scale AI training HPC | |
| Precios en la Nube | |||
| Más Barato Bajo Demanda | $2.00/hr | — | |
| Más Barato Spot | — | — | |
| Proveedores | 2 | 0 | |
| Precios del Proveedor (Bajo Demanda) | |||
|
$2.00/hr | No aplica | |
AMD Instinct MI325X
NVIDIA GH200 Superchip