GPU Benchmarks NVIDIA RTX 3090 vs. NVIDIA RTX 4090


Quick links:
- Benchmarks
- Specifications
- Deep Learning GPU Benchmarks 2024–2025
- 3D, GPU Rendering Benchmarks 2024–2025
- RELION Cryo-EM Benchmarks 2024-2025
- Llama3 70B Inference Benchmark 2024–2025
Best GPUs for deep learning, AI development, compute in 2023–2024. Recommended GPU & hardware for AI training, inference (LLMs, generative AI). GPU training, inference benchmarks using PyTorch, TensorFlow for computer vision (CV), NLP, text-to-speech, etc.
We benchmark NVIDIA RTX 3090 vs NVIDIA RTX 4090 GPUs and compare AI performance (deep learning training; FP16, FP32, PyTorch, TensorFlow), 3d rendering, Cryo-EM performance in the most popular apps (Octane, VRay, Redshift, Blender, Luxmark, Unreal Engine, Relion Cryo-EM).
Our benchmarks will help you decide which GPU (NVIDIA RTX 4090/4080, H100 Hopper, H200, A100, RTX 6000 Ada, A6000, A5000, or RTX 6000 ADA Lovelace) is the best GPU for your needs. We provide an in-depth analysis of the AI performance of each graphic card's performance so you can make the most informed decision possible. We offer deep learning and 3d rendering benchmarks that will help you get the most out of your hardware.
Looking for a GPU workstation or server for AI/ML, design, rendering, simulation or molecular dynamics?
Explore BIZON AI workstations or GPU servers.
Contact us today or explore our various customizable AI solutions.
Featured GPU benchmarks:
- NVIDIA H100 vs. A100 vs. RTX 6000 Ada
- NVIDIA RTX 4090 vs. RTX 4080 vs. RTX 3090
- NVIDIA RTX 3080 vs. 3080 Ti vs. RTX 3090
- NVIDIA RTX 3090 vs. NVIDIA A6000
- NVIDIA A6000 vs. A5000 vs. NVIDIA RTX 3090
- NVIDIA A100 vs. NVIDIA A6000
- NVIDIA RTX 2080 Ti vs. Titan RTX vs Quadro RTX8000
Benchmarks
Deep Learning GPU Benchmarks 2024–2025 [Updated]↑
Resnet50 (FP16)
Resnet50 (FP32)
Resnet152 (FP16)
Resnet152 (FP32)
Inception V3 (FP16)
Inception V3 (FP32)
Inception V4 (FP16)
Inception V4 (FP32)
VGG16 (FP16)
VGG16 (FP32)
3D, GPU Rendering Benchmarks 2024–2025 [Updated]↑
V-Ray
Octane
Redshift
Blender
Luxmark
Unreal Engine
RELION Cryo-EM Benchmarks 2024-2025 [Updated]↑
Total run time
Llama3 70B Inference Benchmark 2024–2025 [Updated]↑
Eval rate
Technical Specifications
Board Design
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
Length | 13 in / 336 mm | 13 in / 336 mm | - |
Outputs | 1x HDMI, 3x DisplayPort | 1x HDMI, 3x DisplayPort | - |
Power Connectors | 1x 12-pin | 1x 16-pin | |
Slot width | Triple-slot | Triple-slot | - |
TDP | 350 W | 450 W | 100 W (29%) |
Clock Speeds
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
Boost Clock | 1695 MHz | 2520 MHz | 825 MHz (49%) |
GPU Clock | 1395 MHz | 2235 MHz | 840 MHz (60%) |
Memory Clock | 19500 MHz | 21200 MHz | 1700 MHz (9%) |
Graphics Card
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
Bus Interface | PCIe 4.0 x16 | PCIe 4.0 x16 | - |
Generation | GeForce 30 | GeForce 40 |
Graphics Features
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
DirectX | 12 Ultimate (12_2) | 12 Ultimate (12_2) | - |
OpenCL | 3 | 3 | - |
OpenGL | 4.6 | 4.6 | - |
Shader Model | 6.6 | 6.7 | |
CUDA | 8.6 | 8.9 |
Graphics Processor
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
Architecture | Ampere | Ada Lovelace | |
Die Size | 628 mm2 | 608 mm2 | -20 mm2 (-3%) |
GPU Name | GA102-300-A1 | AD102-300-A1 | |
Process Size | 8 nm | 5 nm | -3 nm (-37%) |
Transistors | 28300 million | 76300 million | 48000 million (170%) |
Memory
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
Bandwidth | 936.2 GB/s | 1018 GB/s | 81.8 GB/s (9%) |
Memory Bus | 384 bit | 384 bit | - |
Memory Size | 24 GB | 24 GB | - |
Memory Type | GDDR6X | GDDR6X | - |
Render Config
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
ROPs | 112 | 192 | 80 (71%) |
RT Cores | 82 | 128 | 46 (56%) |
Shading Units/ CUDA Cores | 10496 | 16384 | 5888 (56%) |
TMUs | 328 | 512 | 184 (56%) |
Tensor Cores | 328 | 512 | 184 (56%) |
Theoretical Performance
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
FP16 (half) performance | 35.58 TFLOPS | 82.58 TFLOPS | 47 TFLOPS (132%) |
FP32 (float) performance | 35.58 TFLOPS | 82.58 TFLOPS | 47 TFLOPS (132%) |
FP64 (double) performance | 556 GFLOPS | 1290 GFLOPS | 734 GFLOPS (132%) |
Pixel Rate | 189.8 GPixel/s | 483.8 GPixel/s | 294 GPixel/s (155%) |
Texture Rate | 556 GTexel/s | 1290 GTexel/s | 734 GTexel/s (132%) |
Price
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
MSRP | $1,499.00 | $1,599.00 | $100.00 (7%) |
Release Date | Sep 1st, 2020 | Oct 12th, 2022 |
Test bench configuration
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
Hardware | BIZON X5000 More details | BIZON X5500 More details | |
Software | 3D Rendering: Nvidia Driver: 456.38 VRay Benchmark: 5 Octane Benchmark: 2020.1.5 Redshift Benchmark: 3.0.28 Demo Blender: 2.90 Luxmark: 3.1 | 3D Rendering: Nvidia Driver: VRay Benchmark: Octane Benchmark: Redshift Benchmark: Blender: Luxmark: |
Recommended hardware
NVIDIA RTX 3090 | NVIDIA RTX 4090 | Difference | |
---|---|---|---|
Best GPU workstations | BIZON X5500 | BIZON X5500 |