{"data":{"accelerator":{"count":1,"unit":"GPU"},"accelerator_backend":"nvidia","aliases":["rtx 4080"],"captured_at":"2026-08-27","commercial":{"availability":{"provenance":{"captured_at":"2026-08-31T11:49:17.000Z","sources":[{"captured_at":"2026-08-31T11:49:17.000Z","kind":"retailer","publisher":"newegg","url":"https://www.newegg.com/nvidia-founders-edition-900-1g136-2560-000-geforce-rtx-4080-16gb-graphics-card/p/1FT-0004-00844?item=9SIB4VJKCS6246"}]},"reason":"fresh-in-stock-retailer-observation","scope":"at-least-one-us-retailer","state":"available"},"channel_status":"legacy_vendor_catalog","current_stock":true,"current_street_price":{"amount":1799,"currency":"USD"},"msrp":{"amount":1199,"currency":"USD"},"prices":[{"amount":1199,"as_of":"2022-11","captured_at":"2026-08-27T04:55:00Z","currency":"USD","kind":"MSRP","region":"US","scope":"manufacturer_launch_price","source":{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"},"unit":"one_time"},{"amount":1799,"as_of":"2026-08-31T11:49:17.000Z","captured_at":"2026-08-31T11:49:17.000Z","configuration":"NVIDIA GeForce RTX 4080 Founders Edition Graphics Card 16GB GDDR6X","currency":"USD","kind":"street_price","region":"US","scope":"current_in_stock_retail_listing","source":{"captured_at":"2026-08-31T11:49:17.000Z","kind":"retailer","publisher":"newegg","url":"https://www.newegg.com/nvidia-founders-edition-900-1g136-2560-000-geforce-rtx-4080-16gb-graphics-card/p/1FT-0004-00844?item=9SIB4VJKCS6246"},"unit":"one_time"}]},"compute":{"architecture":"NVIDIA Ada Lovelace","cuda_cores":9728,"stats":{"bf16":{"dense":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":97.5},"structured_2_4":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":195}},"fp16":{"dense":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":194.9},"structured_2_4":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":389.8}},"fp32":{"dense":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known","unit":"tflops","value":49},"structured_2_4":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"reason":{"code":"not-published","detail":"No structured 2:4 fp32 throughput was published for this exact accelerator."},"state":"unknown","unit":"tflops"}},"fp4":{"unknown":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"reason":{"code":"not-supported-by-architecture","detail":"The vendor architecture documentation for this accelerator does not provide a hardware path for this precision."},"state":"not_applicable","unit":"tflops"}},"fp8":{"dense":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":389.9},"structured_2_4":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":779.8}},"int4":{"dense":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":779.8},"structured_2_4":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":1559.6}},"int8":{"dense":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":389.9},"structured_2_4":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":779.8}},"tf32":{"dense":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":48.7},"structured_2_4":{"provenance":{"captured_at":"2026-08-31T15:30:01Z","sources":[{"captured_at":"2026-08-31T15:30:01Z","kind":"vendor","publisher":"NVIDIA","url":"https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf"}]},"state":"known","unit":"tflops","value":97.4}}}},"facts":{"accelerator.count":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known","unit":"GPU"},"commercial.availability":{"provenance":{"captured_at":"2026-08-31T11:49:17.000Z","sources":[{"captured_at":"2026-08-31T11:49:17.000Z","kind":"retailer","publisher":"newegg","url":"https://www.newegg.com/nvidia-founders-edition-900-1g136-2560-000-geforce-rtx-4080-16gb-graphics-card/p/1FT-0004-00844?item=9SIB4VJKCS6246"}]},"reason":"fresh-in-stock-us-retailer-observation","state":"known"},"commercial.channel_status":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known"},"commercial.current_stock":{"provenance":{"captured_at":"2026-08-31T11:49:17.000Z","sources":[{"captured_at":"2026-08-31T11:49:17.000Z","kind":"retailer","publisher":"newegg","url":"https://www.newegg.com/nvidia-founders-edition-900-1g136-2560-000-geforce-rtx-4080-16gb-graphics-card/p/1FT-0004-00844?item=9SIB4VJKCS6246"}]},"reason":"fresh-in-stock-us-retailer-observation","state":"known"},"commercial.current_street_price":{"provenance":{"captured_at":"2026-08-31T11:49:17.000Z","sources":[{"captured_at":"2026-08-31T11:49:17.000Z","kind":"retailer","publisher":"newegg","url":"https://www.newegg.com/nvidia-founders-edition-900-1g136-2560-000-geforce-rtx-4080-16gb-graphics-card/p/1FT-0004-00844?item=9SIB4VJKCS6246"}]},"reason":"fresh-in-stock-us-retailer-observation","state":"known"},"commercial.msrp":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known","unit":"one_time"},"commercial.prices":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known"},"compute.architecture":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known"},"compute.cuda_cores":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known"},"memory.bandwidth_gb_per_s":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known"},"memory.vram_gb":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known"},"memory.vram_type":{"provenance":{"captured_at":"2026-08-27T04:55:00Z","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}]},"state":"known"}},"family":"rtx-4080","id":"rtx-4080-16gb","kind":"discrete","memory":{"bandwidth_gb_per_s":717,"cpu_memory_gb":null,"vram_gb":16,"vram_type":"GDDR6X"},"name":"GeForce RTX 4080","product_names":[],"products":["rtx-4080"],"schema_version":"local-ai-registry/v1","sources":[{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/compare/?section=compare-specs"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/"},{"captured_at":"2026-08-27T04:55:00Z","kind":"vendor","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-delivers-quantum-leap-in-performance-introduces-new-era-of-neural-rendering-with-geforce-rtx-40-series"}],"vendor":"nvidia","recipe_count":2,"relationships":{"models":[{"api":"/api/v1/models/qwen3-5-9b","href":"/models/qwen3-5-9b","id":"qwen3-5-9b","name":"Qwen3.5-9B"}],"prices":[{"api":"/api/v1/prices/rtx-4080--gb","href":"/prices/rtx-4080--gb","id":"rtx-4080--gb","name":"RTX 4080 · GB"},{"api":"/api/v1/prices/rtx-4080--jp","href":"/prices/rtx-4080--jp","id":"rtx-4080--jp","name":"RTX 4080 · JP"},{"api":"/api/v1/prices/rtx-4080--pl","href":"/prices/rtx-4080--pl","id":"rtx-4080--pl","name":"RTX 4080 · PL"},{"api":"/api/v1/prices/rtx-4080--us","href":"/prices/rtx-4080--us","id":"rtx-4080--us","name":"RTX 4080 · US"}],"recipes":[{"api":"/api/v1/recipes/qwen35-9b-q4km-rtx4080-llamacpp-tp1","href":"/recipes/qwen35-9b-q4km-rtx4080-llamacpp-tp1","id":"qwen35-9b-q4km-rtx4080-llamacpp-tp1","engine":"llama.cpp","hardware_count":1,"status":"validated"},{"api":"/api/v1/recipes/qwen359b-exl3-4bpw-rtx4080-tabbyapi-tp1","href":"/recipes/qwen359b-exl3-4bpw-rtx4080-tabbyapi-tp1","id":"qwen359b-exl3-4bpw-rtx4080-tabbyapi-tp1","engine":"tabbyapi","hardware_count":1,"status":"validated"}]}},"meta":{"source":"registry"}}