…
H100 SXM vs Instinct MI300X
Cloud rental price, compute per dollar and specs, side by side.
Open in the interactive comparer →
H100 SXM$1.66lowest on-demand per GPU·h
$1.68 per PFLOP·h · 28 providers
Instinct MI300X$1.71lowest on-demand per GPU·h
$1.31 per PFLOP·h · 10 providersWhat actually differs
- The Instinct MI300X has 1.3× more dense FP16/BF16 throughput than the H100 SXM (0.989 PF vs 1.31 PF).
- The Instinct MI300X has 1.6× more memory bandwidth than the H100 SXM (3,350 vs 5,300 GB/s), which usually limits LLM token generation. Memory: 80 GB vs 192 GB.
- Both cost about the same per hour at the lowest on-demand prices ($1.66 vs $1.71).
- Per unit of compute, the Instinct MI300X is the better deal: $1.31 vs $1.68 per PFLOP-hour of dense FP16/BF16.
| Spec / price | H100 SXM | Instinct MI300X |
|---|---|---|
| Memory | 80 GB HBM3 | 192 GB HBM3better |
| Memory bandwidth | 3,350 GB/s | 5,300 GB/sbetter |
| FP16 / BF16 (dense) | 0.989 PF | 1.31 PFbetter |
| FP8 (dense) | 1.98 PF | 2.61 PFbetter |
| FP4 (dense) | — | — |
| INT8 (dense) | 1.98 PF | 2.61 PFbetter |
| TF32 (dense) | 0.495 PF | 0.654 PFbetter |
| FP32 (dense) | 0.067 PF | 0.163 PFbetter |
| FP64 (dense) | 0.067 PF | 0.163 PFbetter |
| Power | 700 W | 750 W |
| Architecture | Hopper (2022) | CDNA 3 (2023) |
| Lowest on-demand | $1.66better | $1.71 |
| Median on-demand | $3.19 | $2.99better |
| Lowest spot | $0.79better | $1.11 |
| FP16/BF16 $ per PFLOP·h | $1.68 | $1.31better |
| Providers | 28 | 10 |
H100 SXM vs Instinct MI300X: questions
Is the H100 SXM faster than the Instinct MI300X?
The Instinct MI300X has 1.3× more dense FP16/BF16 compute than the H100 SXM. The Instinct MI300X has 1.6× more memory bandwidth than the H100 SXM. Real-world speedups depend on the workload: training and prefill scale with compute, while LLM decoding scales mostly with memory bandwidth.
Which is cheaper per FLOP, the H100 SXM or the Instinct MI300X?
Right now the Instinct MI300X is cheaper per unit of compute: $1.31 vs $1.68 per PFLOP-hour at the lowest on-demand prices.
How much does it cost to rent an H100 SXM vs an Instinct MI300X for a month?
Running one GPU around the clock at the lowest on-demand price costs about $1,212 per month for the H100 SXM and $1,248 for the Instinct MI300X.
Which has more memory, the H100 SXM or the Instinct MI300X?
The Instinct MI300X has more memory: 192 GB vs 80 GB.