Here is a question worth asking: why would you pair the newest Blackwell architecture with a model released in 2022? Because SD 1.5 is still the backbone of countless production pipelines, and the RTX 5080 makes it absurdly fast. We measured the combination on GigaGPU.
43 Images Per Minute
| Metric | Value |
|---|---|
| Iterations/sec | 18.2 it/s |
| Seconds per image | 1.37 sec (25 steps) |
| Images per minute | 43.68 |
| Resolution | 512×512 |
| Sampler | Euler a / DPM++ 2M Karras |
| Performance rating | Excellent |
Benchmark conditions: 25-step generation at 512×512, batch size 1, FP16 precision. Using A1111 WebUI or ComfyUI backend.
Memory Layout
| Component | VRAM |
|---|---|
| Model weights | 3.2 GB |
| Sampling buffer | ~0.6 GB |
| Total RTX 5080 VRAM | 16 GB |
| Free headroom | ~12.8 GB |
Twelve-plus GB of headroom with a 3.2 GB model. You could run SD 1.5 alongside an entire Whisper Large-v3 instance (3.1 GB) and still have 9 GB to spare. Multi-model pipelines on a single card become not just possible but comfortable at this VRAM level.
What Each Image Costs
| Cost Metric | Value |
|---|---|
| Server cost | £0.95/hr (£189/mo) |
| Cost per 1K images | £0.36 |
| Images per £1 | 2778 |
At 36p per thousand, the 5080 achieves the lowest per-image cost of any card we have tested for SD 1.5. That is not a typo — the Blackwell efficiency gains are substantial. Full numbers on the benchmark comparison page.
The Right Workload
High-volume batch generation. Print-on-demand services processing thousands of variations daily. Marketing teams that need near-instant iterations. At 43 images per minute, the 5080 handles all of it. If you need the absolute maximum throughput, the RTX 5090 pushes past 61 img/min — but at nearly double the monthly cost. Balance your needs using our GPU selection guide.
Quick deploy:
docker run --gpus all -p 7860:7860 ghcr.io/ai-dock/stable-diffusion-webui:latest --sd15Also see: SD hosting guide, all benchmarks, PaddleOCR hosting.
Deploy SD 1.5 on RTX 5080
Order this exact configuration. UK datacenter, full root access.
Order RTX 5080 Server
