B300
NVIDIA · Blackwell Ultra
288 GBHBM3e
8 TB/sBandwidth
4.5 PFFP8 dense
$4,899 /GPU/mo$6,552 · −25%
AMD · CDNA 3 · OAM · 8-GPU UBB
Rent the AMD Instinct MI300X by the month.
The AMD Instinct MI300X offers 192 GB of HBM3 at 5.3 TB/s and 2.6 PF of dense FP8, from the lowest monthly price of any HBM GPU in our catalog. It is the cost-efficient way to serve 70B to 405B models.
192 GBHBM3 per GPU
5.3 TB/sMemory bandwidth
2.6 PFFP8 dense
Infinity FabricInterconnect
Why the MI300X
What the MI300X is best at
- 192 GB HBM3. 2.4× the memory of an H100 at a lower monthly price.
- 5.3 TB/s bandwidth. Higher than H200, ideal for memory-bound decoding.
- Lowest-priced HBM GPU. Our cheapest GPU with high-bandwidth memory, from $1,749 per month.
- Inference of 70B–405B models
- Batch embedding and RAG
- Cost-sensitive fine-tuning
vLLM, SGLang, PyTorch and JAX run natively on ROCm; our ROCm image is ready to use.
Pricing
MI300X monthly price by server size
One flat price per GPU, the same from 1 to 8 GPUs. The market median is the public on-demand list price of specialist GPU clouds × 730 hours.
| Server | CoreGPU / month | Market median / month | You save / month | You save / year | ≈ per GPU-hour |
|---|---|---|---|---|---|
| 1× MI300X | $1,749 | $602 | $7,224 | $2.40 | |
| 2× MI300X | $3,498 | $1,204 | $14,448 | $2.40 | |
| 4× MI300X | $6,996 | $2,408 | $28,896 | $2.40 | |
| 8× MI300X | $13,992 | $4,816 | $57,792 | $2.40 |
Prices in USD, excluding taxes. Market median checked Nov 4, 2026 across specialist GPU clouds; hyperscalers excluded. All GPU prices
Compare
MI300X vs H100
MI300X has 192 vs 80 GB and more FP8 throughput for less money; H100 is the pick when you depend on CUDA-only kernels.
| MI300X | H100 | |
|---|---|---|
| Architecture | CDNA 3 | Hopper |
| GPU memory | 192 GB HBM3 | 80 GB HBM3 |
| Memory bandwidth | 5.3 TB/s | 3.35 TB/s |
| FP8 dense | 2.6 PF | 1.98 PF |
| FP4 dense | n/a | n/a |
| Interconnect | Infinity Fabric · 896 GB/s | NVLink 4 · 900 GB/s |
| Power (TDP) | 750 W | 700 W |
| Price per GPU / month | $1,749 | $1,999 |
| Price per GB of memory | $9.11 | $24.99 |
| Price per FP8 PFLOPS | $673 | $1,010 |
| Availability | 41 available | Only 3 left |
Specifications
AMD Instinct MI300X technical specs
Per-GPU figures from the vendor datasheet. Tensor throughput is dense (without sparsity).
| Model | AMD Instinct MI300X |
|---|---|
| Architecture | CDNA 3 |
| Form factor | OAM · 8-GPU UBB |
| GPU memory | 192 GB HBM3 |
| Memory bandwidth | 5.3 TB/s |
| FP8 Tensor (dense) | 2.6 PF |
| FP4 Tensor (dense) | Not supported |
| Interconnect | Infinity Fabric · 896 GB/s |
| Power (TDP) | 750 W |
| GPUs per server | 1 to 8 |
| Availability | 41 available |
FAQ
MI300X rental: common questions
Another question? Our engineers answer within one business day.
Contact usHow much does it cost to rent an AMD Instinct MI300X per month?
$1,749 per GPU per month at CoreGPU, about $2.40 per GPU-hour, versus a market median of $2,351 ($3.22/hour × 730 h). A full 8-GPU server is $13,992 per month, taxes excluded.
Is the MI300X available right now?
Yes: 41 available in the current allocation of 96 GPUs. In-stock servers are delivered with root SSH access in under 60 seconds after payment.
How many MI300X GPUs can I rent?
From 1 to 8 GPUs per server, at the same price per GPU. Servers with 8 GPUs come with the full Infinity Fabric mesh. Need more? Rent several servers or ask for a private allocation.
MI300X or H100?
MI300X has 192 vs 80 GB and more FP8 throughput for less money; H100 is the pick when you depend on CUDA-only kernels. See the full MI300X vs H100 comparison.
What do I get with a MI300X server?
Dedicated GPUs (never shared or sliced), root SSH access, your choice of a clean OS or a ready image with ROCm, PyTorch and vLLM, local NVMe scratch, and month-to-month billing with no setup fee.
Alternatives
Other GPUs in stock now
Deploy your MI300X server in under 60 seconds
Pick 1 to 8 GPUs in the configurator above. Month to month, cancel anytime.