BTC
ETH
USDT
XMR
LTC
GPU comparison · CDNA 4 vs Hopper
AMD MI355X vs NVIDIA H100.
AMD CDNA 4 against NVIDIA Hopper. The MI355X has 3.6× the memory (288 GB of HBM3E against 80 GB of HBM3), 2.4× the bandwidth, about 2.5× the dense FP8 throughput and FP4, for $370 less per GPU per month. The H100 runs the CUDA ecosystem unchanged.
- 288 vs 80 GBGPU memory
- 8 vs 3.35 TB/smemory bandwidth
- $1,409 vs $1,779per GPU, per month
MI355X vs H100: which one to rent.
Choose the MI355X if
- You serve large models or long contexts: 288 GB per GPU.
- Your frameworks support ROCm and you want the lower price: $1,409 against $1,779 a month.
- You want the lower price per GB of GPU memory: $4.89 against $22.24 a month.
Choose the H100 if
- Your code relies on CUDA kernels, TensorRT-LLM or CUDA-only libraries.
- Your models fit in 80 GB and you want the most mature software support.
- You prefer a market price with many public references (our MI355X reference is indicative).
MI355X vs H100 specs, side by side.
Figures from NVIDIA’s and AMD’s published specifications, per GPU. Tensor figures are peak dense throughput.
| Per GPU | AMD MI355X | NVIDIA H100 | H100 / MI355X |
|---|---|---|---|
| Architecture | CDNA 4 (gfx950), TSMC 3 nm and 6 nm | Hopper, TSMC 4N | — |
| Form factor | OAM module, liquid-cooled | SXM5 module | — |
| GPU memory | 288 GB HBM3E (8 stacks of 36 GB) | 80 GB HBM3, 5,120-bit | 0.28× |
| Memory bandwidth | 8 TB/s | 3.35 TB/s | 0.42× |
| GPU-to-GPU link | Infinity Fabric, 7 links of 153.6 GB/s | NVLink 4, 900 GB/s per GPU | 0.84× |
| FP4 tensor (dense) | 10.1 PFLOPS | Not supported | — |
| FP8 tensor (dense) | 5,000 TFLOPS | 1,979 TFLOPS | 0.4× |
| FP16 / BF16 tensor (dense) | 2,500 TFLOPS | 989 TFLOPS | 0.4× |
| FP64 | 78.6 TFLOPS | 34 / 67 TFLOPS | 0.43× |
| Max power | 1,400 W | Up to 700 W, configurable | — |
| GPUs per server | 1, 2, 4, 8 | 1, 2, 4, 8 | — |
| Per GPU in our servers | 32 vCPU · 384 GB RAM · 3.84 TB | 20 vCPU · 200 GB RAM · 2 TB | — |
Sources: AMD Instinct MI355X · ROCm MI350 series · NVIDIA H100. Full sheets: MI355X · H100
MI355X vs H100 rental price.
At GPUCenter, the MI355X rents for $1,409 and the H100 for $1,779 per GPU per month, 30% below the market median of public on-demand prices, paid in crypto with no KYC. Hourly figures are the monthly price divided by 730 hours.
| GPUCenter | AMD MI355X | NVIDIA H100 | H100 / MI355X |
|---|---|---|---|
| Price per GPU, per month | $1,409 | $1,779 | 1.26× |
| Equivalent per GPU-hour | $1.93 | $2.44 | — |
| Per GB of GPU memory, per month | $4.89 | $22.24 | 4.5× |
| Per PFLOPS of dense FP8, per month | $282 | $899 | 3.2× |
| Largest server | 8× · $11,272/mo | 8× · $14,232/mo | — |
| Market median, per GPU | $2,022* | $2,548 | — |
| Below the median | −30% | −30% | — |
* Indicative market reference: no provider publishes an on-demand price for this GPU yet. Same price per GPU at every server size, no hourly metering. How we compare prices · Configure an MI355X server · Configure a H100 server
Which models fit on each.
Smallest server that holds each model, by precision. Rule of thumb with headroom for the KV cache, 92% of GPU memory usable.
| Model (total parameters) | MI355X · FP8 | H100 · FP8 | MI355X · 4-bit | H100 · 4-bit |
|---|---|---|---|---|
| Qwen3.5-9B | 1× | 1× | 1× | 1× |
| Gemma 4 31B | 1× | 1× | 1× | 1× |
| Llama 3.3 70B | 1× | 2× | 1× | 1× |
| gpt-oss-120b | 1× | 2× | 1× | 2× |
| DeepSeek-V4-Flash | 2× | 8× | 1× | 4× |
| Qwen3.5-397B-A17B | 2× | 8× | 1× | 4× |
| DeepSeek-R1 (671B) | 4× | — | 2× | 8× |
| Kimi K2.6 (1T) | 8× | — | 4× | — |
FP8 ≈ 1.2 bytes and 4-bit ≈ 0.65 bytes per parameter. — = larger than the biggest server of that GPU. Size another model · How much VRAM does an LLM need?
MI355X vs H100 FAQ.
More on each GPU: AMD MI355X · NVIDIA H100.
Is the MI355X faster than the H100?
On paper it has 2.4× the memory bandwidth and about 2.5× the dense FP8 throughput. Actual speed depends on how well your software is optimized for ROCm; we do not publish benchmarks.
Does vLLM run on the MI355X?
Yes: vLLM and SGLang ship ROCm builds for AMD Instinct GPUs, and PyTorch runs with its ROCm build.
Why is the MI355X cheaper than the H100?
Our prices follow the market: each GPU is priced 30% below the market reference we found for it, and that reference is lower for the MI355X. No provider publishes an on-demand price for the MI355X yet, so its market reference is indicative.
What fits on one GPU?
By our rule of thumb, about 61B parameters in FP8 on one H100, and about 220B in FP8 or a 405B model in 4-bit on one MI355X.
Other GPU comparisons.
Rent the MI355X or the H100.
Dedicated servers with 1 to 8 GPUs, one monthly price, paid in crypto. Online in under 10 minutes, no KYC.