AMD MI250X 128GB
Performance
Peak theoretical throughput by precision type
| Precision | Peak |
|---|---|
FP64 64-bit floating point | 48TFLOPS |
FP32 32-bit floating point | 48TFLOPS |
TF32 | No verified data available |
BF16 Brain Float 16 | 383TFLOPS |
FP16 16-bit floating point | 383TFLOPS |
FP8 | No verified data available |
FP6 | No verified data available |
FP4 | No verified data available |
INT8 8-bit integer | 383TOPS |
INT4 4-bit integer | 383TOPS |
Every figure here is dense. The vendor documents no structured (2:4) sparsity mode for this part, so there is no second number to quote.
Dense peak divided by accelerator TDP. Board power only: excludes host CPUs, networking, cooling and facility overhead. TDP for this part: 560 W.
The MI250X 128GB in the GPU landscape
Peak FP16 TFLOPS (dense) against TDP, single-GPU parts tracked by Flopper
Higher and further left is better: more half-precision throughput for less power.
Specifications
Architecture
CDNA 2
Form Factor
OAM
Launch Year
2021
Process Node
6nm
Memory
128 GB
Bandwidth
3,277 GB/s
TDP
560 W
Max power (Flopper estimate)
~644 W est. Flopper estimate: 560 W TDP x 1.15. The vendor publishes no maximum board power for this part.
Spec Confidence
Official
Full Specifications
| Memory | |
|---|---|
| Memory | 128 GB |
| Memory Type | No verified data available |
| Bandwidth | 3.3 TB/s |
| Interface Width | No verified data available |
| Power & Thermal | |
| TDP | 560 W |
| Max power (Flopper estimate) | ~644 W est. |
| Enterprise Features | |
| Sparsity | No |
| General | |
| Form Factor | OAM |
| Architecture | CDNA 2 |
| Process Node | 6nm |
| Launch Year | 2021 |
Datasheet & Resources
Data Provenance
Every figure traced to a source
Primary Source
- Document
- AMD Instinct MI250X
- Publisher
- AMD
- Published
- 2021-11-08
Data Quality
- Spec confidence
- Official
- Clock basis
- Boost
- Core precisions with figures
- 5 of 9
- Normalization
- All values in TFLOPS
Compare cloud providers offering on-demand GPU instances for AI training, inference, and HPC workloads.
Browse GPU Cloud ProvidersSimilar GPUs
Frequently Asked Questions
How many TFLOPS does the AMD MI250X have?
The AMD MI250X delivers 48 TFLOPS FP32 and 383 TFLOPS FP16 at peak. Flopper does not currently have a verified FP8 throughput figure for it.
What is the power consumption of the AMD MI250X?
The AMD MI250X has a TDP (Thermal Design Power) rating of 560 watts.
How much memory does the AMD MI250X have?
The AMD MI250X is equipped with 128 GB of memory with 3,277 GB/s of memory bandwidth.
What architecture is the AMD MI250X based on?
The AMD MI250X is based on the CDNA 2 architecture, launched in 2021.
How much does the AMD MI250X cost to rent?
Flopper is not tracking any live rental listings for the AMD MI250X. The provider directory lists the clouds that rent comparable accelerators on demand.
When did the AMD MI250X come out?
The AMD MI250X launched in 2021 on the CDNA 2 architecture.
Stay Updated on GPU Releases
Get notified when new GPUs are added or specifications are updated.
No spam, unsubscribe anytime.