Comparison · GPU vs GPU · read 00:32 UTCH200 vs MI300X
Specifications, today's prices from 69 providers, price per unit of memory and compute, and who rents each cheapest. Better values are highlighted; lower is better for prices.
Side by sideSpecs, prices, value
Highlighted cells are the better of the two: more memory and compute, lower prices, lower cost per unit.
| H200 | MI300X | Difference | |
|---|---|---|---|
| Memory & compute | |||
| VRAM | 141 GB | 192 GB | -27% |
| Memory bandwidth | 4.80 TB/s | 5.30 TB/s | -9% |
| FP16 tensor | 989 TFLOPS | 1,307.4 TFLOPS | -24% |
| FP32 | 67 TFLOPS | 163.4 TFLOPS | -59% |
| Power | 700 W | 750 W | -7% |
| Prices today | |||
| Cheapest on-demand | $2.09 | $1.65 | +27% |
| Median on-demand | $4.80 | $2.71 | +77% |
| Highest on-demand | $13.67 | $6.00 | +128% |
| Cheapest spot | $0.93 | $0.99 | -6% |
| Cheapest reserved | $1.67 | $1.32 | +27% |
| Current offers (n) | 111 | 23 | +383% |
| Providers with a fresh offer | 55 | 14 | +293% |
| Value | |||
| Per GB of VRAM · hour (cheapest) | $0.015 | $0.009 | +72% |
| Per GB of VRAM · hour (median) | $0.034 | $0.014 | +141% |
| Per PFLOP FP16 · hour (median) | $4.853 | $2.074 | +134% |
| Per TB/s · hour (median) | $1.000 | $0.512 | +95% |
| Trend | |||
| 30-day median change | +4.8% | +1.5% | — |
| 90-day median change | +15.7% | +1.5% | — |
90 days · median on-demand price · rebased to 100How the two have moved
Both medians indexed to 100 ninety days ago, so the slopes compare even though the price levels differ.
Cheapest providers today · on-demand · current offersWho rents each one cheapest
- 1
BeamTrust 761× $2.09 - 2
GPUBankTrust 931× $2.49 - 3
Airon AI8× $2.63
- 1GPUBankTrust 931× $1.65
- 2
TensorWaveTrust 648× $1.71 - 3
OpenRelay1× $1.99
Verdict · computed from the numbers aboveChoose one, choose the other
Choose the H200 if…
- you want choice: 55 providers have a fresh offer today against 14.
Choose the MI300X if…
- your model needs more than 141 GB per GPU: 192 GB against 141 GB.
- the hourly price matters most: median $2.71 against $4.80, 44% less.
- you serve LLMs: memory bandwidth per dollar is better ($0.512 per TB/s-hour against $1.000).
Questions about H200 vs MI300X
Which is cheaper to rent, the H200 or the MI300X?
Today the cheapest current on-demand H200 is $2.09 per GPU-hour and the cheapest MI300X is $1.65; medians are $4.80 and $2.71. Cheapest excludes offers marked unavailable and counts one floor per marketplace.
How much memory do the H200 and MI300X have?
141 GB of HBM3e on the H200, 192 GB of HBM3 on the MI300X, at 4.80 and 5.30 TB/s.
Where do these numbers come from?
From 134 current offers read on the providers' own pages and APIs over the last 45 days (reference 00:32 UTC), or in our October 2026 market study for the few providers our collector does not read. Specifications come from manufacturer datasheets. Every price on the GPU pages links to its receipt and shows its read date.