Consumercarts is reader-supported. Learn more
NVIDIA L40 vs NVIDIA Quadro RTX A6000: Comparison
NVIDIA L40 and NVIDIA Quadro RTX A6000 separate most on boost clock: NVIDIA L40 at 2490 MHz, NVIDIA Quadro RTX A6000 at 1800 MHz, 28% in favour of NVIDIA L40. Memory Bandwidth splits them by another 11%, again favouring NVIDIA L40. They match on 11 other specifications, including gpu brand, vram and tier.
Specs
| Specification |  |  |
|---|---|---|
| GPU Brand | NVIDIA | NVIDIA |
| Series | RTX Ada Generation | RTX Ampere Generation |
| Board Partner | NVIDIA Reference (data center/OEM only, passive cooling) | NVIDIA Reference (workstation, no AIB partners) |
| VRAM | 48 GB | 48 GB |
| Tier | Flagship | Flagship |
| Memory Type | GDDR6 | GDDR6 |
| Boost Clock | 2490 MHz | 1800 MHz-28% |
| Ray Tracing Support | true | true |
| TDP | 300 W | 300 W |
| Architecture | Ada Lovelace | Ampere |
| Process Node | TSMC 4N | Samsung 8nm |
| Card Length | 267 mm | 267 mm |
| Slot Width | 2 | 2 |
| Memory Bus Width | 384 bit | 384 bit |
| Memory Bandwidth | 864 GB/s | 768 GB/s-11% |
| Recommended PSU | 550 W | 550 W |
| Power Connector | 1x 16-pin (12VHPWR) | 1x 8-pin |
| MSRP | 8000 | 4650 |
| Launch Date | 2022-10 | 2020-10 |
| Base / Game Clock | 735 MHz | 1410 MHz |
| PCIe Interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
| CUDA Cores | 18176 | 10752 |
| Display Outputs | 4x DisplayPort 1.4a (rarely used in server deployment) | 4x DisplayPort 1.4a |
| Media Engine | 3x NVENC (8th gen), 3x NVDEC, AV1 encode/decode | 2x NVENC (7th gen), 2x NVDEC |
Verdict
NVIDIA L40
8.6 / 10 Expert Rating
A versatile data center GPU that genuinely splits the difference between graphics and AI workloads - since replaced by the L40S for AI-heavy deployments but still a strong all-rounder.
NVIDIA Quadro RTX A6000
7.9 / 10 Expert Rating
Still a genuinely useful 48GB workstation card on the secondhand market, even four years after launch - just expect a real gap versus current RTX PRO Blackwell cards in ray tracing and AI performance.