RTX 5080
See rental prices ↓Current median: $0.65/hr across 3 providers (observed), 0.00 (0.0%) vs yesterday. Typical range: $0.60–$0.67/hr (P10–P90). 30-day trend unlocks after 30 days of captures
TDP 360W · 275.4 TFLOPS · 16GB VRAM · 960 GB/s bandwidth
The RTX 5080 has a 360W TDP, 275.4 TFLOPS of FP16 compute, 16GB of GDDR7 VRAM, 960 GB/s memory bandwidth.
Strong mid-range Blackwell. 16GB GDDR7 limits model size but excellent FP32 per dollar. Lower TDP than 5090.
RTX 5080 currently rents for a median of $0.65/hr across 3 providers (observed on-demand prices). RTX 5080 has 16GB of memory. Compare its specs and listed cloud rental prices below. The workload must fit in VRAM with room for runtime memory; a listed price does not confirm stock.
RTX 5080 Overview
The RTX 5080 brings Blackwell architecture to a more attainable consumer tier, with 16GB of GDDR7, 275.4 TFLOPS of FP16/BF16 throughput, and a 360W board power target. It is a fast accelerator for compact models, but its memory capacity—not arithmetic capability—sets the practical boundary for many AI jobs.
GDDR7 delivers 960 GB/s of bandwidth, a substantial improvement for a 16GB-class card and useful for responsive inference on models that fit. Capacity remains the constraint: 16GB is comfortable for smaller models and quantized workloads, while larger FP16 models and large KV caches quickly require a different tier of hardware.
The RTX 5080 suits experimentation, small-model fine-tuning, image generation, and economical inference fleets where individual requests fit on one device. It lacks ECC and NVLink, so it is not a replacement for professional or HBM-equipped GPUs in reliability-critical or memory-scaled deployments.
Memory
Compute Performance
Hardware Specifications
Chip
Processors
Clocks
Memory
Power
Relative Performance
Relative to highest-spec GPU in database
Limitations
Live Cloud PricingOn-demand hourly rates
| Provider | On-Demand | Spot | Region | Availability |
|---|---|---|---|---|
BESTSwissGPU | $0.28/hr | — | CH | unknown |
| $0.68/hr | — | EU/US | unknown | |
| $0.65/hr | — | Various | unknown | |
| $0.59/hr | — | US | unknown |
Job Cost Estimator
At median $0.65/hr across 3 providers (observed on-demand). Spot pricing typically 40–70% lower.
Estimates only. Actual cost depends on provider, region, availability, and utilization. Use the GPU cost calculator for custom inputs.
Compare RTX 5080 vs…
Use Case Guidance
LLM Model Size Guidance
Estimates only. Actual capacity depends on context length, KV cache, and framework overhead.
Related Guides
LLM APIs Running on This GPU Class
Providers that serve frontier LLM inference on Blackwell-class hardware.
RTX 5080 vs Alternatives — Spec Comparison
| Spec | RTX 5080 this | RTX PRO 6000 Blackwell | A100 80GB | A100 40GB |
|---|---|---|---|---|
| VRAM | 16GB GDDR7 | 96GB GDDR7▲ | 80GB HBM2e | 40GB HBM2e |
| Memory Bandwidth | 960 GB/s | 1792 GB/s | 2000 GB/s▲ | 1555 GB/s |
| FP16 TFLOPS | 275.4 | 250 | 312▲ | 312▲ |
| BF16 TFLOPS | 275.4 | 250 | 312▲ | 312▲ |
| FP8 TFLOPS | — | 500▲ | — | — |
| INT8 TOPS | 551 | 500 | 624▲ | 624▲ |
| TDP | 360W | 300W | 400W▲ | 300W |
| Process Node | TSMC 4NP | TSMC 4NP | TSMC 7nm▲ | TSMC 7nm▲ |
| Architecture | GB203▲ | GB202 | GA100 | GA100 |
| Release Year | 2025▲ | 2025▲ | 2020 | 2020 |
| Max model (FP16) | ~8B params | ~48B params▲ | ~40B params | ~20B params |
| Max model (INT4) | ~32B params | ~192B params▲ | ~160B params | ~80B params |
Compare RTX 5080
Related GPUs
RTX 5080 Frequently Asked Questions
What is the RTX 5080 TDP?
The RTX 5080 has a TDP (Thermal Design Power) of 360W.
How many TFLOPS does the RTX 5080 have?
The RTX 5080 delivers 275.4 TFLOPS of FP16 compute and 275.4 TFLOPS BF16.
How much VRAM does the RTX 5080 have?
The RTX 5080 has 16GB of GDDR7 VRAM.
What is the RTX 5080's memory bandwidth?
The RTX 5080 has a memory bandwidth of 960 GB/s.
RTX 5080 User Reviews
RTX 5080 User Reviews
Ready to rent?
Compare RTX 5080 prices across 130+ providers
Live on-demand & spot rates · monthly cost estimates · availability status