Specifications Compared
| Spec | MI300X | RTX 3090 Ti |
|---|---|---|
| TDP | 750W | 350W |
| VRAM | 192 GB | 24 GB |
| CUDA Cores | Not published | 10,496 |
| FP8 (dense) | 2,614.9 TFLOPS | Not published |
| Memory Type | HBM3 | GDDR6X |
| Architecture | CDNA 3 | Ampere |
| FP16 (dense) | 1,307.4 TFLOPS | Not published |
| Form Factors | OAM | PCIe |
| INT8 (dense) | 2,614.9 TOPS | Not published |
| Interconnect | Infinity Fabric, PCIe 5.0 | NVLink |
| Tensor Cores | Not published | 328 |
| FP32 Performance | 163.4 TFLOPS | 35.6 TFLOPS |
| FP64 Performance | 81.7 TFLOPS | Not published |
| Memory Bandwidth | 5,300 GB/s | 936 GB/s |
| FP8 (with sparsity) | 5,229.8 TFLOPS | Not published |
| FP16 (with sparsity) | 2,614.9 TFLOPS | Not published |
| INT8 (with sparsity) | 5,229.8 TOPS | Not published |
| FP16 (vendor figure, sparsity not specified) | Not published | 35.6 TFLOPS |
Performance Analysis
The MI300X variant delivers FP16 dense performance of 1307.4 TFLOPS against the RTX 3090 Ti variant FP16 figure of 35.6 TFLOPS without sparsity designation. This gap means the MI300X variant supports larger batch sizes during training and inference when memory bandwidth of 5300 GB/s is available to every MI300X variant. FP32 performance reaches 163.4 TFLOPS on the MI300X variant compared with 35.6 TFLOPS on the RTX 3090 Ti variant. The RTX 3090 Ti variant TDP of 350 W shared across its family remains lower than the 750 W TDP shared across the MI300X family. Memory bandwidth of 936 GB/s shared by every RTX 3090 variant limits batch sizes relative to the higher bandwidth shared by every MI300X variant.
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.
MI300X
MI300X is not offered on-demand by any provider we track right now. See the MI300X rental page for last-seen listed prices and a price alert.
RTX 3090 Ti
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| Vast.ai | Finland, FI | 8 | $0.27 | $2.13 | Deploy |
| LeaderGPU | The Netherlands | 8 | $0.29 | $2.29 | Deploy |
2 providers in stock, 10 offers (cheapest per provider shown). All RTX 3090 Ti offers, price history and alerts
Notify me when MI300X 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 MI300X
The MI300X variant with 192 GB HBM3 memory suits scenarios that require high capacity for models exceeding 24 GB. Its OAM form factor and Infinity Fabric interconnect align with deployments that need the FP16 dense throughput of 1307.4 TFLOPS shared by every MI300X variant.
When to Choose the RTX 3090 Ti
The RTX 3090 Ti variant with 24 GB GDDR6X memory fits workloads that operate within lower power limits of 350 W shared across its family. Its PCIe form factor and NVLink interconnect match systems that do not require the higher memory capacity of the MI300X variant.
Use Cases
The MI300X variant provides 192 GB HBM3 memory that accommodates models larger than the 24 GB GDDR6X capacity of the RTX 3090 Ti variant.
FP16 dense performance of 1307.4 TFLOPS shared by every MI300X variant supports higher throughput than the RTX 3090 Ti variant FP16 figure of 35.6 TFLOPS.
Memory bandwidth of 5300 GB/s shared across the MI300X family enables larger batch sizes than the 936 GB/s shared by every RTX 3090 variant.
The RTX 3090 Ti variant TDP of 350 W shared across its family suits consumer scale image generation within 24 GB GDDR6X memory.
FP32 performance of 163.4 TFLOPS on the MI300X variant exceeds the 35.6 TFLOPS of the RTX 3090 Ti variant for compute heavy simulations.
Frequently Asked Questions
How does TDP compare between these variants?▾
TDP of 750 W applies to every MI300X variant. TDP of 350 W applies to every RTX 3090 variant including the 3090 Ti model.
What FP16 performance does the MI300X variant deliver?▾
The MI300X variant delivers FP16 dense performance of 1307.4 TFLOPS. This figure is shared by every MI300X variant and exceeds the RTX 3090 Ti variant FP16 figure of 35.6 TFLOPS.
Which interconnect does the RTX 3090 Ti variant use?▾
The RTX 3090 Ti variant uses NVLink interconnect. This differs from the Infinity Fabric and PCIe 5.0 interconnect on the MI300X variant.
Which is cheaper to rent, the MI300X or the RTX 3090 Ti?▾
Cloud rental prices for both the MI300X and RTX 3090 Ti 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 MI300X have compared to the RTX 3090 Ti?▾
The MI300X has 192 GB of HBM3 memory. The RTX 3090 Ti has 24 GB of GDDR6X memory.
Can I find MI300X and RTX 3090 Ti 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 MI300X and the RTX 3090 Ti?▾
The MI300X uses the CDNA 3 architecture (2023) while the RTX 3090 Ti uses Ampere (2020). The MI300X delivers 18.4x the dense FP16 throughput (1,307.4 vs 71 TFLOPS, both without sparsity) and 5.7x the memory bandwidth of the RTX 3090 Ti.
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 MI300X and the RTX 3090 Ti. 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 MI300XEvery current offer by provider, daily price history and a price alert.
- Rent RTX 3090Every 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.