Specifications Compared
| Spec | GH200 | MI300X |
|---|---|---|
| TDP | 900W | 750W |
| VRAM | 96 GB | 192 GB |
| CUDA Cores | 16,896 | Not published |
| FP8 (dense) | 1,979 TFLOPS | 2,614.9 TFLOPS |
| Memory Type | HBM3 | HBM3 |
| Architecture | Hopper | CDNA 3 |
| FP16 (dense) | 989 TFLOPS | 1,307.4 TFLOPS |
| Form Factors | SXM | OAM |
| INT8 (dense) | 1,979 TOPS | 2,614.9 TOPS |
| Interconnect | NVLink-C2C, PCIe 5.0 | Infinity Fabric, PCIe 5.0 |
| Tensor Cores | 528 | Not published |
| FP32 Performance | 67 TFLOPS | 163.4 TFLOPS |
| FP64 Performance | 34 TFLOPS | 81.7 TFLOPS |
| Memory Bandwidth | 4,000 GB/s | 5,300 GB/s |
| FP8 (with sparsity) | 3,958 TFLOPS | 5,229.8 TFLOPS |
| FP16 (with sparsity) | 1,979 TFLOPS | 2,614.9 TFLOPS |
| INT8 (with sparsity) | 3,958 TOPS | 5,229.8 TOPS |
Performance Analysis
The FP16 dense rating of 1307.4 TFLOPS on the MI300X exceeds the 989 TFLOPS dense FP16 rating on the GH200 by a factor derived solely from these two figures. This gap indicates greater throughput for training workloads that rely on dense FP16 operations. The FP16 with sparsity rating of 2614.9 TFLOPS on the MI300X likewise exceeds the 1979 TFLOPS with sparsity rating on the GH200. Memory bandwidth of 5300 GB/s on the MI300X compared with 4000 GB/s on the GH200 supports larger batch sizes during both training and inference. FP32 throughput of 163.4 TFLOPS on the MI300X compared with 67 TFLOPS on the GH200 favors the MI300X for precision sensitive scientific tasks.
Current On-Demand Offers
Cheapest secure on-demand offer per provider that is reported in stock right now, per GPU per hour; multi-GPU instances show the whole-instance price alongside. Deploy opens the DeployGPU console for providers on the platform, otherwise the provider's own site. Stock is rechecked every minute.
GH200
GH200 is not offered on-demand by any provider we track right now. See the GH200 rental page for last-seen listed prices and a price alert.
MI300X
1 provider in stock, 1 offer. All MI300X offers, price history and alerts
Notify me when GH200 drops below a price
One email when the cheapest in-stock on-demand price per GPU-hour falls below your threshold.
QuantaCloud
Comparing H-series providers? We broker across all of them.
Hopper stock changes by the hour and prices differ by provider. If you need 16+ GPUs reserved or a cluster in the next 90 days, we quote H-series or B300 inventory at partner rates: one quote, 24h turnaround.
When to Choose the GH200
The GH200 suits deployments that prioritize NVLink C2C interconnect within SXM form factor systems. Its 989 TFLOPS dense FP16 rating remains adequate when paired with existing NVIDIA software stacks. Lower memory capacity of 96 GB HBM3 limits its selection to workloads that fit within this constraint.
When to Choose the MI300X
The MI300X suits deployments that require 192 GB HBM3 memory and 5300 GB/s bandwidth for large model training. Its 1307.4 TFLOPS dense FP16 rating and 163.4 TFLOPS FP32 rating deliver higher arithmetic throughput. The 750W TDP rating further favors the MI300X in power constrained clusters.
Use Cases
The MI300X supplies 192 GB HBM3 memory and 1307.4 TFLOPS dense FP16 performance compared with 96 GB HBM3 memory and 989 TFLOPS dense FP16 performance on the GH200.
The MI300X supplies 5300 GB/s memory bandwidth and 2614.9 TFLOPS dense FP8 performance compared with 4000 GB/s bandwidth and 1979 TFLOPS dense FP8 performance on the GH200.
The MI300X supplies 192 GB HBM3 memory which accommodates larger fine tuning batches than the 96 GB HBM3 memory on the GH200.
Both GPUs provide comparable FP16 with sparsity ratings of 1979 TFLOPS on the GH200 and 2614.9 TFLOPS on the MI300X for image generation workloads.
The MI300X supplies 163.4 TFLOPS FP32 performance compared with 67 TFLOPS FP32 performance on the GH200.
Frequently Asked Questions
How does FP16 dense performance compare between the two GPUs?▾
The MI300X delivers 1307.4 TFLOPS dense FP16. The GH200 delivers 989 TFLOPS dense FP16.
Which GPU offers higher memory bandwidth?▾
The MI300X offers 5300 GB/s memory bandwidth. The GH200 offers 4000 GB/s memory bandwidth.
What are the TDP ratings for the GH200 and MI300X?▾
The GH200 carries a 900W TDP rating. The MI300X carries a 750W TDP rating.
Does the GH200 support NVLink C2C interconnect?▾
The GH200 supports NVLink C2C interconnect. The MI300X supports Infinity Fabric interconnect.
What FP32 performance does each GPU deliver?▾
The MI300X delivers 163.4 TFLOPS FP32. The GH200 delivers 67 TFLOPS FP32.
Which is cheaper to rent, the GH200 or the MI300X?▾
Cloud rental prices for both the GH200 and MI300X vary by provider, configuration, and availability. This page shows live pricing from 25+ providers updated every 60 seconds. Scroll to the Live Cloud Pricing section to compare current rates.
How much VRAM does the GH200 have compared to the MI300X?▾
The GH200 has 96 GB of HBM3 memory. The MI300X has 192 GB of HBM3 memory.
Can I find GH200 and MI300X GPUs available to rent right now?▾
Yes. This page shows real-time availability across 25+ cloud GPU providers. The Live Cloud Pricing section displays only in-stock offers with current pricing.
What is the main difference between the GH200 and the MI300X?▾
The GH200 uses the Hopper architecture (2023) while the MI300X uses CDNA 3 (2023). The MI300X delivers 1.3x the dense FP16 throughput (1,307.4 vs 989 TFLOPS, both without sparsity) and 1.3x the memory bandwidth of the GH200.
Rent these GPUs
Each GPU page lists every current on-demand offer by provider, per GPU-hour, updated every minute.
Related comparisons
How this page is made
- Specifications come from the NVIDIA, AMD and Intel datasheets for the GH200 and the MI300X. Dense and sparse throughput are listed separately.
- Prices and availability are live: every offer shown is an in-stock, on-demand listing from a provider we track, rechecked every minute.
- The written comparison (overview, performance notes, when to choose each card, verdict, use cases and FAQ) was drafted with an AI model from the specification table above and passed an automated check that rejects any figure not in that table. It was last generated on . It contains no prices; those are always read live.
- Read how we collect the data, or report an error on this page.
Next steps
- Rent GH200 Grace HopperEvery current offer by provider, daily price history and a price alert.
- Rent MI300XEvery current offer by provider, daily price history and a price alert.
- GPU price indexHow on-demand prices for the major GPUs have moved, updated daily.
- All GPUsEvery GPU we track, with current lows and provider counts.