RX 9070 image benchmarks

37 reported configurations · 4 sources · 37 measurements

Documented results

24 selected configurations

Selected across sources, not a performance ranking. Match model weights, resolution, steps, software and timing scope before comparing.

Source	Model	GPUs	Output	Measurements
View ReportUL Procyon	Stable Diffusion 1.5FP16Model reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine UL Procyon AI Image Generation Backend ONNX Runtime / DirectML OS Windows Parallelization Not reported Memory 16 GB per GPU Reported 2025-06-04 Full record →	512 × 512Batch 4100 stepsBatch 4	2.8 s/imageTime / image · reported Whole workflow · Timing details missing
View ReportDiffusion workflow	Stable Diffusion XLFP8Model reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine SHARK / Shortfin / IREE Backend ROCm / HIP OS Ubuntu Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	Resolution not reportedSteps not reportedBatch not reported	3.7 s/imageTime / image · reported Whole workflow · Timing details missing
View Reportstable-diffusion.cpp / sd2R	FLUX.1 devQ4_K_SModel reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine sd2R / stable-diffusion.cpp Backend Vulkan OS Not reported Parallelization Single GPU Memory 16 GB per GPU Reported Not reported Full record →	768 × 76810 stepsBatch not reported	13.72 sReported time · reported Timing details missing
View ReportSD WebUI	diffusers/john6666/wai-nsfw-illustrious-v120-sdxlQuant not reportedFamily reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine SD.Next 9ea16a30 Backend ROCm / HIP OS Linux Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	512 × 512Batch 1Steps not reportedBatch 1	4.86 image-steps/sPipeline throughput · reportedHarness computes steps times batch divided by process_images wall time; not sampler iterations per second or images per second.
View ReportUL Procyon	Stable Diffusion XLFP16Model reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine UL Procyon AI Image Generation Backend ONNX Runtime / DirectML OS Windows Parallelization Not reported Memory 16 GB per GPU Reported 2025-06-04 Full record →	1024 × 1024Batch 1100 stepsBatch 1	21.4 s/imageTime / image · reported Whole workflow · Timing details missing
View ReportDiffusion workflow	Stable Diffusion XLINT8Model reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine SHARK / Shortfin / IREE Backend ROCm / HIP OS Ubuntu Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	Resolution not reportedSteps not reportedBatch not reported	4.2 s/imageTime / image · reported Whole workflow · Timing details missing
View Reportstable-diffusion.cpp / sd2R	FLUX.1 devQ4_K_SModel reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine sd2R / stable-diffusion.cpp Backend Vulkan OS Not reported Parallelization Single GPU Memory 16 GB per GPU Reported Not reported Full record →	1024 × 102410 stepsBatch not reported	24.84 sReported time · reported Timing details missing
View ReportSD WebUI	diffusers/john6666/wai-nsfw-illustrious-v120-sdxlQuant not reportedFamily reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine SD.Next 9ea16a30 Backend ROCm / HIP OS Linux Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	512 × 512Batch 2Steps not reportedBatch 2	0.86 image-steps/sPipeline throughput · reportedHarness computes steps times batch divided by process_images wall time; not sampler iterations per second or images per second.
View ReportDiffusion workflow	Stable Diffusion XLFP16Model reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine IREE SDXL scripts Backend ROCm / HIP OS Not reported Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	Resolution not reportedSteps not reportedBatch not reported	0.263 s/forwardCalculated · Model forward time · reported · Model componentExact ms-to-seconds conversion; isolated UNet, not image latency.reported milliseconds / 1000 Model component · Timing details missing
View Reportstable-diffusion.cpp / sd2R	FLUX.1 devQ4_K_SModel reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine sd2R / stable-diffusion.cpp Backend Vulkan OS Not reported Parallelization Single GPU Memory 16 GB per GPU Reported Not reported Full record →	2048 × 102410 stepsBatch not reported	42.7 sReported time · reported Timing details missing
View ReportSD WebUI	diffusers/john6666/wai-nsfw-illustrious-v120-sdxlQuant not reportedFamily reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine SD.Next 9ea16a30 Backend ROCm / HIP OS Linux Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	512 × 512Batch 4Steps not reportedBatch 4	1.04 image-steps/sPipeline throughput · reportedHarness computes steps times batch divided by process_images wall time; not sampler iterations per second or images per second.
View ReportDiffusion workflow	Stable Diffusion XLFP8Model reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine IREE SDXL scripts Backend ROCm / HIP OS Not reported Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	Resolution not reportedSteps not reportedBatch not reported	0.151 s/forwardCalculated · Model forward time · reported · Model componentExact ms-to-seconds conversion; isolated UNet, not image latency.reported milliseconds / 1000 Model component · Timing details missing
View Reportstable-diffusion.cpp / sd2R	FLUX.1 devQ4_K_SModel reference ↗Image → image	Radeon RX 90701 GPU Configuration Engine sd2R / stable-diffusion.cpp Backend Vulkan OS Not reported Parallelization Single GPU Memory 16 GB per GPU Reported Not reported Full record →	768 × 76810 stepsBatch not reported	8.16 sReported time · reported Timing details missing
View ReportSD WebUI	v1-5-pruned-emaonly.safetensorsQuant not reportedFamily reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine Stable Diffusion WebUI f680b4e7 Backend ZLUDA OS Windows Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	512 × 512Batch 1Steps not reportedBatch 1	3.03 image-steps/sPipeline throughput · reportedHarness computes steps times batch divided by process_images wall time; not sampler iterations per second or images per second.
View Reportstable-diffusion.cpp / sd2R	FLUX.1 devQ4_K_SModel reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine sd2R / stable-diffusion.cpp Backend Vulkan OS Not reported Parallelization Single GPU Memory 16 GB per GPU Reported Not reported Full record →	1024 × 102410 stepsBatch not reported	24.9 sReported time · reported Timing details missing
View ReportSD WebUI	v1-5-pruned-emaonly.safetensorsQuant not reportedFamily reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine Stable Diffusion WebUI f680b4e7 Backend ZLUDA OS Windows Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	512 × 512Batch 2Steps not reportedBatch 2	3.08 image-steps/sPipeline throughput · reportedHarness computes steps times batch divided by process_images wall time; not sampler iterations per second or images per second.
View Reportstable-diffusion.cpp / sd2R	FLUX.2 Klein 4BQuant not reportedModel reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine sd2R / stable-diffusion.cpp Backend Vulkan OS Not reported Parallelization Single GPU Memory 16 GB per GPU Reported Not reported Full record →	768 × 7684 stepsBatch not reported	13.58 sReported time · reported Timing details missing
View ReportSD WebUI	v1-5-pruned-emaonly.safetensorsQuant not reportedFamily reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine Stable Diffusion WebUI f680b4e7 Backend ZLUDA OS Windows Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	512 × 512Batch 4Steps not reportedBatch 4	3.04 image-steps/sPipeline throughput · reportedHarness computes steps times batch divided by process_images wall time; not sampler iterations per second or images per second.
View Reportstable-diffusion.cpp / sd2R	FLUX.2 Klein 4BQuant not reportedModel reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine sd2R / stable-diffusion.cpp Backend Vulkan OS Not reported Parallelization Single GPU Memory 16 GB per GPU Reported Not reported Full record →	1024 × 10244 stepsBatch not reported	32.51 sReported time · reported Timing details missing
View ReportSD WebUI	animagine-xl-4.0-optQuant not reportedFamily reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine SD.Next 341cbd9d Backend ZLUDA OS Windows Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	512 × 512Batch 1Steps not reportedBatch 1	2.44 image-steps/sPipeline throughput · reportedHarness computes steps times batch divided by process_images wall time; not sampler iterations per second or images per second.
View Reportstable-diffusion.cpp / sd2R	FLUX.2 Klein 4BQuant not reportedModel reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine sd2R / stable-diffusion.cpp Backend Vulkan OS Not reported Parallelization Single GPU Memory 16 GB per GPU Reported Not reported Full record →	2048 × 10244 stepsBatch not reported	45.01 sReported time · reported Timing details missing
View ReportSD WebUI	animagine-xl-4.0-optQuant not reportedFamily reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine SD.Next 341cbd9d Backend ZLUDA OS Windows Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	512 × 512Batch 2Steps not reportedBatch 2	2.52 image-steps/sPipeline throughput · reportedHarness computes steps times batch divided by process_images wall time; not sampler iterations per second or images per second.
View Reportstable-diffusion.cpp / sd2R	FLUX.2 Klein 4BQuant not reportedModel reference ↗Image → image	Radeon RX 90701 GPU Configuration Engine sd2R / stable-diffusion.cpp Backend Vulkan OS Not reported Parallelization Single GPU Memory 16 GB per GPU Reported Not reported Full record →	768 × 7684 stepsBatch not reported	8.08 sReported time · reported Timing details missing
View ReportSD WebUI	animagine-xl-4.0-optQuant not reportedFamily reference ↗Text → image	Radeon RX 90701 GPU Configuration Engine SD.Next 341cbd9d Backend ZLUDA OS Windows Parallelization Not reported Memory 16 GB per GPU Reported Not reported Full record →	512 × 512Batch 4Steps not reportedBatch 4	2.48 image-steps/sPipeline throughput · reportedHarness computes steps times batch divided by process_images wall time; not sampler iterations per second or images per second.

Explore all 37 configurations →

All image benchmarks · Methodology and source review