Highest memory capacity AMD GPU for single-chip model serving.
Weekly median price per GPU per hour · Get the data
By provider shows one card per company, placed where its best offer ranked. By configuration lists every offer.
What you can rent comes first: in stock, then waitlist, then not reported, then out of stock. Priced offers rank ahead of quote-only, and on-demand ahead of other billing types.
Within a group, five factors set the order:
We match the provider, its country, GPU model, form factor, billing type, availability and instance name. Matching is partial and case-insensitive. Everyday words work too, so "interruptible" finds spot and "sold out" finds out of stock.
No providers match .
Heads up: A provider's own page may quote a different figure, for example on tax, region or a promotion. A monthly-only plan shows a derived hourly rate. Verify before provisioning. More on how we price.
One MI325X has 256 GB of VRAM. In practice, that's enough memory for roughly 408B parameters at 4-bit or 110B at 16-bit, assuming a 32K context. Below are some open-weight LLMs, with the estimated memory and GPUs each one needs.
| Model | Memory (INT4 / FP4) | MI325Xs needed | Cost /hr | Cost /mo |
|---|---|---|---|---|
Qwen3.8-27B
27B
|
18 GB
|
1
|
$3.06
|
$2,203
|
Gemma 4 31B
30.7B
|
22 GB
|
1
|
$3.06
|
$2,203
|
GPT-OSS-120B
117B · 5.1B active
|
67 GB
|
1
|
$3.06
|
$2,203
|
GLM-5.3-Flash
320B · 18B active
|
178 GB
|
1
|
$3.06
|
$2,203
|
MiniMax-M3
428B · 23B active
|
239 GB
|
8
|
$24.48
|
$17,626
|
DeepSeek V4.1 Flash
552B · 16B active
|
306 GB
|
8
|
$24.48
|
$17,626
|
Kimi K3
2.8T · 104B active
|
1,544 GB
|
8
|
$24.48
|
$17,626
|
Estimates based on the median on-demand rate. Memory is weights plus FP8 KV cache at 32K context per request (FP16/BF16 in the 16-bit column). GPU counts assume 90% of advertised VRAM is usable. No guarantee of runtime support, usable performance, or that a matching quantized build exists. Pricing methodology.
AMD MI325X · Per GPU
| Compute · dense | |
|---|---|
| FP8 | 2,614.9 TFLOPS 5,229.8 with sparsity |
| FP16 / BF16 | 1,307.4 TFLOPS 2,614.9 with sparsity |
| INT8 | 2,614.9 TOPS 5,229.8 with sparsity |
| FP32 | 163.4 TFLOPS |
| FP64 | 81.7 TFLOPS |
| Precision support | FP8FP16BF16TF32FP32FP64INT8 |
| Memory | |
|---|---|
| Capacity | 256 GB HBM3e |
| Bandwidth | 6,000 GB/s |
| Bus width | 8,192-bit |
| ECC | Yes |
| Silicon | |
|---|---|
| Architecture | CDNA 3 |
| Process | TSMC 5nm / 6nm FinFET |
| Transistors | 153 billion |
| Shader cores | 19,456 Stream processors |
| Matrix cores | 1,216 |
| Compute units | 304 |
| Fabric and host | |
|---|---|
| GPU interconnect | Infinity Fabric 896 GB/s |
| Host interface | PCIe 5.0 x16 |
| Power | |
|---|---|
| Board power | 1,000 W |
| Cooling | Passive |
| Platform | |
|---|---|
| Partitioning | Up to 8 partitions |
Source: official AMD MI325X datasheet.
As of September 25, 2026, the median on-demand price is $3.06 per GPU per hour across 3 providers.
| Billing type | Configs | Median /GPU/hr | Cheapest |
|---|---|---|---|
On-demand | 7 | $3.06 | $2.25 |
Reserved | 14 | $1.95 | $1.48 (12 mo) |
Spot | 1 | Sold out |
1 provider also quotes the MI325X on a custom contract, priced per deal.
The lowest listed on-demand price for the MI325X is $2.25 per GPU per hour from Bentaus, though we haven't verified current stock.
At 720 hours per month, one MI325X costs an estimated $2,203 at the median on-demand price.
GetDeploying currently tracks MI325X configs from 5 providers. See the full price comparison above for every provider and config.
One MI325X runs models up to roughly 408B parameters at 4-bit quantization or 110B at 16-bit, assuming a 32K context. Larger models run across multiple GPUs: the model table above shows the estimated memory and GPU count for popular open-weight LLMs.
256GB HBM3e with 6 TB/s bandwidth. Largest VRAM available for serving massive models without sharding. FP8 support for efficient inference.
Very limited cloud availability. ROCm ecosystem, while rapidly improving, may have gaps that affect compatibility with some training frameworks. Consider Nvidia alternatives if CUDA compatibility is required.
Nvidia H100
80 GB HBM3 · 44% less bandwidth than the MI325X
Nvidia B200
180 GB HBM3e · 28% more bandwidth than the MI325X
Nvidia H200
141 GB HBM3e · 20% less bandwidth than the MI325X
AMD MI300X
192 GB HBM3 · released Q4 2023
Last updated
