Specifications Compared
| Spec | MI355X | RTX 5090 |
|---|---|---|
| TDP | 1400W | 575W |
| VRAM | 288 GB | 32 GB |
| CUDA Cores | Not published | 21,760 |
| FP4 (dense) | 10,100 TFLOPS | 1,676 TFLOPS |
| FP8 (dense) | 5,000 TFLOPS | 419 TFLOPS |
| Memory Type | HBM3e | GDDR7 |
| Architecture | CDNA 4 | Blackwell |
| FP16 (dense) | 2,500 TFLOPS | 209.5 TFLOPS |
| Form Factors | OAM | PCIe |
| INT8 (dense) | 5,000 TOPS | 838 TOPS |
| Interconnect | Infinity Fabric, PCIe 5.0 | PCIe 5.0 |
| Tensor Cores | Not published | 680 |
| FP32 Performance | 157.3 TFLOPS | 104.8 TFLOPS |
| FP64 Performance | 78.6 TFLOPS | 1.6 TFLOPS |
| Memory Bandwidth | 8,000 GB/s | 1,792 GB/s |
| FP4 (with sparsity) | Not published | 3,352 TFLOPS |
| FP8 (with sparsity) | 10,100 TFLOPS | 838 TFLOPS |
| FP16 (with sparsity) | 5,000 TFLOPS | 419 TFLOPS |
| INT8 (with sparsity) | 10,100 TOPS | 1,676 TOPS |
Performance Analysis
The FP16 dense figure of 2500 TFLOPS on the MI355X versus 209.5 TFLOPS on the RTX 5090 indicates the MI355X can sustain substantially larger matrix operations during training and inference passes. When sparsity is applied the MI355X FP16 figure rises to 5000 TFLOPS against 419 TFLOPS on the RTX 5090 so the same ratio holds for sparse workloads. Memory bandwidth of 8000 GB/s on the MI355X compared with 1792 GB/s on the RTX 5090 permits batch sizes several times larger before data movement becomes the limiter. FP32 performance stands at 157.3 TFLOPS for the MI355X and 104.8 TFLOPS for the RTX 5090 so the MI355X maintains an advantage even in single precision kernels.
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.
MI355X
MI355X is not offered on-demand by any provider we track right now. See the MI355X rental page for last-seen listed prices and a price alert.
Notify me when MI355X drops below a price
One email when the cheapest in-stock on-demand price per GPU-hour falls below your threshold.
QuantaCloud
Comparing providers? We broker across all of them.
Stop tab-switching between pricing pages. Tell us what you need, 16+ GPUs reserved or cluster capacity, and we return one quote at partner rates within 24 hours.
When to Choose the MI355X
The MI355X suits deployments that require 288 GB HBM3e capacity and 8000 GB/s bandwidth for models that exceed 32 GB footprints. Its 2500 TFLOPS FP16 dense rating and 5000 TFLOPS FP16 with sparsity rating support large scale training and inference where memory capacity directly limits throughput.
When to Choose the RTX 5090
The RTX 5090 fits environments constrained to 575W TDP and PCIe form factor cards. Its 209.5 TFLOPS FP16 dense rating and 419 TFLOPS FP16 with sparsity rating remain adequate for workloads that fit within 32 GB GDDR7 memory and do not require Infinity Fabric interconnect.
Use Cases
The MI355X supplies 288 GB HBM3e memory and 2500 TFLOPS FP16 dense performance that accommodate models too large for 32 GB GDDR7.
The MI355X 8000 GB/s bandwidth and 5000 TFLOPS FP16 with sparsity rating sustain higher token throughput than the RTX 5090 1792 GB/s and 419 TFLOPS FP16 with sparsity rating.
The MI355X 288 GB capacity permits fine tuning of models that exceed the 32 GB limit of the RTX 5090.
The RTX 5090 32 GB GDDR7 memory and 575W TDP suffice for typical diffusion workloads while fitting standard PCIe servers.
The MI355X 157.3 TFLOPS FP32 rating and 8000 GB/s bandwidth accelerate memory bound scientific kernels beyond the RTX 5090 104.8 TFLOPS FP32 rating.
Frequently Asked Questions
What is the FP16 dense performance of each GPU?▾
The MI355X reaches 2500 TFLOPS FP16 dense and the RTX 5090 reaches 209.5 TFLOPS FP16 dense. These numbers determine raw throughput for dense matrix operations in training and inference.
Which GPU offers higher memory bandwidth?▾
The MI355X provides 8000 GB/s bandwidth compared with 1792 GB/s on the RTX 5090. Higher bandwidth supports larger batch sizes before memory stalls occur.
What TDP values are listed for the MI355X and RTX 5090?▾
The MI355X TDP is 1400W and the RTX 5090 TDP is 575W. Power delivery and cooling requirements therefore differ substantially between the two cards.
Does the MI355X support Infinity Fabric?▾
The MI355X lists Infinity Fabric interconnect in addition to PCIe 5.0 while the RTX 5090 lists only PCIe 5.0. Multi GPU scaling therefore follows different topologies on each platform.
Which is cheaper to rent, the MI355X or the RTX 5090?▾
Cloud rental prices for both the MI355X and RTX 5090 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 MI355X have compared to the RTX 5090?▾
The MI355X has 288 GB of HBM3e memory. The RTX 5090 has 32 GB of GDDR7 memory.
Can I find MI355X and RTX 5090 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 MI355X and the RTX 5090?▾
The MI355X uses the CDNA 4 architecture (2025) while the RTX 5090 uses Blackwell (2025). The MI355X delivers 11.9x the dense FP16 throughput (2,500 vs 209.5 TFLOPS, both without sparsity) and 4.5x the memory bandwidth of the RTX 5090.
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 MI355X and the RTX 5090. 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 MI355XEvery current offer by provider, daily price history and a price alert.
- Rent RTX 5090Every 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.