NVIDIA GeForce RTX 3090
Compare NVIDIA GeForce RTX 3090 against another machine in the Arena
Reference: GeForce RTX 30 series on Wikipedia
- Class
- GPU
- Memory
- 24 GB GDDR6X
- Runs (Q4_K_M)
- —
- Bandwidth
- 936 GB/s
- TDP
- 350 W
- Released
- 2020-09-24
- Price (US)
- $1274 used as of 2026-08
- Price (Canada)
- CA$1900 used sourced, as of 2026-08
Sourcing and disambiguation notes
Discontinued by NVIDIA; no reliable new-retail channel (only sporadic, unconfirmed inflated 'new' marketplace listings, not verified unused stock). Used: resaleprices.com fair range $1,200-1,300, market average $1,274, 344 active US listings, 2026-08-22 -- an eBay category browse confirms a similar range. Canadian price: eBay.ca active pre-owned listings this week, pre-tax, no shipping. Representative of a $1,825-2,000 CAD cluster (Founders Edition pre-owned $1,825; ASUS ROG Strix pre-owned $1,900-2,000; EVGA FTW3 Ultra pre-owned $1,999). Card is long past EOL and out of stock new at Canadian retailers (Canada Computers, Memory Express).
Specification sources
- https://www.nvidia.com/en-us/geforce/graphics-cards/30-series/rtx-3090/ — vendor, 2020-09-01
- https://images.nvidia.com/aem-dam/en-zz/Solutions/geforce/ampere/pdf/NVIDIA-ampere-GA102-GPU-Architecture-Whitepaper-V1.pdf — vendor, 2020-09-01
Measurements
Decode is token generation — the speed you feel while an answer streams. Prefill is prompt processing — the wait before it starts. Why bandwidth predicts decode speed.
No records yet for this hardware. Know of a published benchmark on it? Submit the link.
What verified, single-source and estimated mean, and the same rows with every filter and sort in the benchmarks explorer.
Speed over time
What it can run
Fit is arithmetic, not a measurement — how it is computed. Against 24 GB; models that fit are listed largest first.
No modeled quant fits in 24 GB.