NVIDIA GeForce RTX 4080 SUPER 16GB Specs, Benchmarks & Pricing

• Release Date: January 31, 2024
• MSRP: $999 USD
• GPU Architecture: ada
• Hardware-Accelerated GEMM Operations:
  FP64 FP32 TF32 BF16 FP16 FP8 FP4 INT8 INT4
• CUDA Compute Capability : 8.9

Specifications for NVIDIA GeForce RTX 4080 SUPER

Gaming Benchmarks for NVIDIA GeForce RTX 4080 SUPER

Rankings based on 805,051 real-world benchmark results from OpenBenchmarking.org

References

• https://images.nvidia.com/aem-dam/Solutions/geforce/ada/nvidia-ada-gpu-architecture.pdf
• https://images.nvidia.com/aem-dam/Solutions/Data-Center/l4/nvidia-ada-gpu-architecture-whitepaper-v2.1.pdf
• https://www.nvidia.com/en-us/geforce/graphics-cards/40-series/rtx-4080-family/
• https://www.nvidia.com/en-us/geforce/news/geforce-rtx-4080-4070-ti-4070-super-gpu/
• https://www.techpowerup.com/review/nvidia-geforce-rtx-4080-super-founders-edition/
• https://www.bestgpusforai.com/gpu-comparison/5090-vs-4080-super
• https://pcpartpicker.com/product/9crqqs/nvidia-founders-edition-geforce-rtx-4080-super-16-gb-video-card-900-1g136-2555-000
• https://videocardz.com/newz/nvidia-geforce-rtx-4080-super-listed-with-ad103-gpu-and-new-device-id
• https://www.tomshardware.com/pc-components/gpus/here-are-all-the-nvidia-rtx-4080-super-cards-currently-listed-including-specs-photos-and-pricing
• https://www.techpowerup.com/gpu-specs/asus-rog-strix-rtx-4080-super-gaming.b11501
• https://www.asus.com/motherboards-components/graphics-cards/tuf-gaming/tuf-rtx4080s-o16g-gaming/
• https://rog.asus.com/us/graphics-cards/graphics-cards/rog-strix/rog-strix-rtx4080s-o16g-white/
• https://www.msi.com/Graphics-Card/GeForce-RTX-4080-SUPER-16G-VENTUS-3X-OC
• https://www.msi.com/Graphics-Card/GeForce-RTX-4080-SUPER-16G-GAMING-X-TRIO
• https://www.bestbuy.com/site/msi-nvidia-geforce-rtx-4080-super-16gb-ventus-3x-oc-16gb-gddr6x-pci-express-4-0-graphics-card-black/6571812.p?skuId=6571812
• https://www.bhphotovideo.com/c/product/1805075-REG/gigabyte_gv_n408sgaming_oc_16gd_geforce_rtx_4080_super.html
• https://www.bestbuy.com/site/gigabyte-nvidia-geforce-rtx-4080-super-windforce-v2-16gb-gddr6x-pci-express-4-0-graphics-card-black/6573357.p?skuId=6573357
• https://www.bestbuy.com/site/pny-geforce-rtx-4080-super-verto-overclocked-16gb-graphics-card-with-triple-fans-black/6574654.p?skuId=6574654
• https://www.amazon.com/ZOTAC-Graphics-IceStorm-Advanced-ZT-D40820D-10P/dp/B0CSBDFS55
• https://www.amazon.com/ZOTAC-Graphics-IceStorm-Advanced-ZT-D40820F-10P/dp/B0D17XR7SF
• https://www.amazon.com/ZOTAC-Graphics-IceStorm-Advanced-ZT-D40820Q-10P/dp/B0CSBDKXK3
• https://www.techpowerup.com/gpu-specs/palit-rtx-4080-super-gamingpro.b11560
• https://www.techpowerup.com/gpu-specs/gainward-rtx-4080-super-phoenix.b11550
• https://www.gainward.com/main/vgapro.php?id=1211&lang=en

Notes

1. MSRP of $999 USD sourced from NVIDIA's official announcement. A Founders Edition was available at this price, representing a $200 price reduction from the original RTX 4080's $1199 MSRP.
2. fp16TFLOPS of 418.79 represents "Peak FP16 Tensor TFLOPS with FP16 Accumulate" using the "with sparsity" metric. Dense performance is 209.40 TFLOPS, sparse is 418.79 TFLOPS. Following the same pattern as other Ada GPUs, NVIDIA publishes the sparse performance.
3. int8TOPS of 417.79 represents non-sparse (dense) INT8 performance, derived from sparse INT8 of 835.58 TOPS ÷ 2 per NVIDIA Ada architecture whitepaper. With 2:4 structured sparsity, INT8 performance is 835.58 TOPS. Dense value used for consistent cross-vendor comparison.
4. fp8TFLOPS of 417.79 represents non-sparse (dense) FP8 performance. NVIDIA's Ada Lovelace 4th-generation Tensor Cores process FP8 and INT8 at identical per-clock throughput (both are 8-bit formats) per the NVIDIA Ada GPU Architecture whitepaper; this equal-rate relationship is confirmed by official NVIDIA per-SKU datasheets (e.g. RTX 4090, L40, L40S, L4 all publish identical FP8 and INT8 dense/sparse figures). Dense FP8 = dense INT8 = 417.79 TFLOPS. Dense value used for consistent cross-vendor comparison.
5. Ada Lovelace has no FP4 hardware path (FP4 tensor core support was introduced in Blackwell) — fp4TFLOPS is left null and is a confirmed hardware negative, not a research gap, so it is not added to unverifiedSpecs.
6. fp32TFLOPS of 52.2 is TF32 Tensor Core throughput without sparsity. It equals FP32 vector (CUDA core) throughput, computed as 10,240 CUDA cores × 2 × 2.55 GHz boost = 52.2 TFLOPS; NVIDIA's RTX 4080 family page gives 52 shader TFLOPS. With sparsity TF32 is 104.4 TFLOPS. On GeForce Ada, dense TF32 runs at the FP32 vector rate, as Tables 3 to 6 of NVIDIA's RTX Blackwell whitepaper show, for example the RTX 4080 at 48.7 (https://images.nvidia.com/aem-dam/Solutions/geforce/blackwell/nvidia-rtx-blackwell-gpu-architecture.pdf).
7. Power based on what NVIDIA refers to as "TGP (Total Graphics Power)"
8. The RTX 4080 SUPER uses the fully enabled AD103 chip (all 80 SMs), whereas the original RTX 4080 uses 76 of 80 SMs
9. Memory bandwidth calculated as 256-bit interface × 23 Gbps = 736 GB/s
10. NVIDIA Founders Edition part number (NVPN) is 900-1G136-2555-000, sourced from PCPartPicker and Newegg product listings
11. Board identifier is PG139 SKU 355 with AD103-400 GPU (fully enabled), and Device PCI ID is 2703, sourced from VideoCardz
12. Third-party manufacturer identifiers sourced from official product pages, retail listings (Best Buy, Newegg, Amazon, B&H Photo), and TechPowerUp GPU Database
13. ASUS uses model number format with variant suffix (e.g., ROG-STRIX-RTX4080S-O16G-GAMING for OC, ROG-STRIX-RTX4080S-16G-GAMING for non-OC)
14. MSI uses internal 912-V511-XXX part numbers for inventory management in addition to public SKUs
15. Gigabyte uses GV-N408S prefix for all RTX 4080 SUPER models
16. Zotac uses ZT-D40820X-10P format where X varies by model (D=Trinity Black, F=AMP, Q=Trinity White OC)
17. Palit and Gainward (sister companies) both released RTX 4080 SUPER models; some Gainward models have both numeric and alphanumeric identifiers