Loading the catalog…
MI325X (AMD)
Blurred values are subscriber-only. Sign in to start a free trial — 2 full pages a day — or subscribe for unlimited access.
| Process Node | 5nm FinFET |
|---|---|
| Host Interface | 16-lane PCIe® Gen 5 host interface |
| Tdp W | 1000 |
| Peak Engine Clock Mhz | 2100 |
| Gpu Compute Units | 304 |
| Architecture | CDNA 3 |
| Gpu Memory Gb | 256 |
| Fp16 Bf16 Tflops Dense | 1307.4 |
| Memory Bandwidth Tb S | 6 |
| Fp8 Tflops Dense | 2614.9 |
| Fp32 Tflops | 163.4 |
| Int8 Tops Dense | 2614.9 |
| Memory Bus Width Bits | 8192 |
| Accelerator Form Factor | OAM |
| Memory Type | HBM3E |
| Interconnect | seven AMD Infinity Fabric links for full connectivity between eight GPUs in a ring |
| Fp16 Bf16 Tflops Sparse | 2614.9 |
| Amd Infinity Cache Last Level | |
| Document Type | |
| Fp64 Tflops | |
| Fp8 Tflops Sparse | |
| Int8 Tops Sparse |
| Matrix Cores |
|---|
| Media Engines |
|---|
| Product Name |
|---|
| Ras Features |
|---|
| Software Platform |
|---|
| Sr Iov Support |
|---|
| Stream Processors |
|---|
| Tf32 Tflops Dense |
|---|
| Tf32 Tflops Sparse |
|---|
| SKU | MI325X |
|---|
| Tagline | |
|---|---|
| Supported AI data types | |
| Memory capacity (marketing) | |
| Memory bandwidth (marketing) | |
| Platform compatibility | |
| Video codec acceleration footnote | |
| I/O Dies (IODs) | |
| Memory clock | |
| Scale-up AMD Infinity Fabric™ Links | |
| Scale-out network bandwidth | |
| Core architecture | |
| AMD Infinity Fabric links | |
| Platform | |
| Accelerated compute dies (XCDs) | |
| L1 cache per CU | |
| L2 cache | |
| AMD Infinity Cache | |
| Decoders (architecture) | |
| Memory (architecture) | |
| Virtualization (architecture) | |
| Coherent shared memory bandwidth between GPUs | |
| Platform memory (8-GPU) | |
| Supported AI & ML frameworks | |
| Copyright |
