NVIDIA Quadro RTX 4000 8GB vs NVIDIA Quadro RTX 8000 48GB

The Quadro RTX 8000 delivers 2.3x the FP16 throughput of the Quadro RTX 4000 (131 vs 57 TFLOPS dense). The Quadro RTX 8000 also carries 40 GB more memory (48 GB vs 8 GB).

Quadro RTX 4000: Turing, 2018 Quadro RTX 8000: Turing, 2018
FP16 dense lead
2.3x
Quadro RTX 8000: 131 vs 57 TFLOPS
VRAM
8 vs 48 GB
40 GB more for the Quadro RTX 8000
Bandwidth
416 GB/s vs 672 GB/s
Quadro RTX 8000 moves data faster
TDP
160 W vs 295 W
Quadro RTX 4000 draws 135 W less

Specifications

Quadro RTX 4000
Quadro RTX 8000
Architecture
TuringTuring
Launch Year
20182018
Form Factor
PCIePCIe
VRAM
8 GB48 GB
Memory Bandwidth
416 GB/s672 GB/s
TDP
160 W295 W
Process Node
12nm12nm

Performance (TFLOPS)

Quadro RTX 4000
Quadro RTX 8000
FP64
0.2 TFlops 0.5 TFlops
FP32
7.1 TFlops 16.3 TFlops
TF32
No verified data available No verified data available
BF16
No verified data available No verified data available
FP16
57.0 TFlops 130.5 TFlops
FP8
No verified data available No verified data available
FP6
No verified data available No verified data available
FP4
No verified data available No verified data available
INT8
No verified data available261.0 TFlops

FLOPS by Precision

What actually differs

Both GPUs launched in 2018: the Quadro RTX 4000 on Turing and the Quadro RTX 8000 on Turing.

Dense throughput for the Quadro RTX 4000 against the Quadro RTX 8000: FP64 0.2 vs 0.5 TFLOPS, FP32 7.1 vs 16 TFLOPS, FP16 57 vs 131 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 8 GB against 48 GB, fed at 416 GB/s versus 672 GB/s. More memory per GPU means larger models fit before you have to shard across cards, which often matters more than raw TFLOPS for inference.

Power budgets are 160 W for the Quadro RTX 4000 and 295 W for the Quadro RTX 8000. At FP32 that works out to 0.04 against 0.06 TFLOPS per watt.

For LLM training and inference, weight the FP8 and FP16 rows and memory capacity most heavily. For scientific and HPC workloads, the FP64 row is the one to read.

Frequently asked questions

Is the Quadro RTX 4000 faster than the Quadro RTX 8000?

At FP16 precision the Quadro RTX 8000 reaches 131 TFLOPS dense against 57 TFLOPS for the Quadro RTX 4000. The performance table on this page lists every published precision for both GPUs.

Which has more memory, the Quadro RTX 4000 or the Quadro RTX 8000?

The Quadro RTX 8000 carries 48 GB of VRAM versus 8 GB for the Quadro RTX 4000. Memory bandwidth is 416 GB/s for the Quadro RTX 4000 and 672 GB/s for the Quadro RTX 8000.

How much power do the Quadro RTX 4000 and the Quadro RTX 8000 draw?

The Quadro RTX 4000 is rated at 160 W TDP and the Quadro RTX 8000 at 295 W. On FP32 throughput per watt, the Quadro RTX 8000 is the more efficient part.

Can I rent the Quadro RTX 4000 or the Quadro RTX 8000 in the cloud?

Yes. Live cloud listings tracked by Flopper start at $0.07 per GPU hour. The rental pricing section on this page lists current providers and rates for both GPUs.

Where to Rent

NVIDIA Quadro RTX 4000 8GB

ProviderConfigurationPrice/GPU-hrChecked
Vast.ai
2× Quadro RTX 4000 8GB Community $0.07 1h ago View →
All Quadro RTX 4000 listings and price history →

NVIDIA Quadro RTX 8000 48GB

ProviderConfigurationPrice/GPU-hrChecked
Vast.ai
1× Quadro RTX 8000 48GB Community $0.25 1h ago View →
All Quadro RTX 8000 listings and price history →

Get Comparison Updates

New GPUs added weekly. Be the first to see how they compare.