NVIDIA GeForce RTX 4090 vs NVIDIA A100 SXM4 80 GB
Comparative analysis of NVIDIA GeForce RTX 4090 and NVIDIA A100 SXM4 80 GB videocards for all known characteristics in the following categories: Essentials, Technical info, Video outputs and ports, Compatibility, dimensions and requirements, API support, Memory. Benchmark videocards performance analysis: PassMark - G2D Mark, PassMark - G3D Mark, 3DMark Fire Strike - Graphics Score, CompuBench 1.5 Desktop - Face Detection (mPixels/s), CompuBench 1.5 Desktop - T-Rex (Frames/s), CompuBench 1.5 Desktop - Video Composition (Frames/s), CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s), Geekbench - OpenCL, CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s).
Differences
Reasons to consider the NVIDIA GeForce RTX 4090
- Videocard is newer: launch date 1 year(s) 10 month(s) later
- 2x more core clock speed: 2235 MHz vs 1095 MHz
- Around 79% higher boost clock speed: 2520 MHz vs 1410 MHz
- 2.4x more pipelines: 16384 vs 6912
- A newer manufacturing process allows for a more powerful, yet cooler running videocard: 4 nm vs 7 nm
- Around 52% better performance in Geekbench - OpenCL: 317227 vs 208996
Specifications (specs) | |
Launch date | 20 Sep 2022 vs 16 Nov 2020 |
Core clock speed | 2235 MHz vs 1095 MHz |
Boost clock speed | 2520 MHz vs 1410 MHz |
Pipelines | 16384 vs 6912 |
Manufacturing process technology | 4 nm vs 7 nm |
Benchmarks | |
Geekbench - OpenCL | 317227 vs 208996 |
Reasons to consider the NVIDIA A100 SXM4 80 GB
- 472.2x more texture fill rate: 609.1 GTexel/s vs 1,290 GTexel/s
- Around 13% lower typical power consumption: 400 Watt vs 450 Watt
- 3.3x more maximum memory size: 80 GB vs 24 GB
- Around 21% higher memory clock speed: 1593 MHz (3.2 Gbps effective) vs 1313 MHz, 21 Gbps effective
Texture fill rate | 609.1 GTexel/s vs 1,290 GTexel/s |
Thermal Design Power (TDP) | 400 Watt vs 450 Watt |
Maximum memory size | 80 GB vs 24 GB |
Memory clock speed | 1593 MHz (3.2 Gbps effective) vs 1313 MHz, 21 Gbps effective |
Compare benchmarks
GPU 1: NVIDIA GeForce RTX 4090
GPU 2: NVIDIA A100 SXM4 80 GB
Geekbench - OpenCL |
|
|
Name | NVIDIA GeForce RTX 4090 | NVIDIA A100 SXM4 80 GB |
---|---|---|
PassMark - G2D Mark | 1295 | |
PassMark - G3D Mark | 38568 | |
3DMark Fire Strike - Graphics Score | 36499 | |
CompuBench 1.5 Desktop - Face Detection (mPixels/s) | 472.899 | |
CompuBench 1.5 Desktop - T-Rex (Frames/s) | 95.851 | |
CompuBench 1.5 Desktop - Video Composition (Frames/s) | 221.169 | |
CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s) | 4429.59 | |
Geekbench - OpenCL | 317227 | 208996 |
CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s) | 0 |
Compare specifications (specs)
NVIDIA GeForce RTX 4090 | NVIDIA A100 SXM4 80 GB | |
---|---|---|
Essentials |
||
Architecture | Ada Lovelace | Ampere |
Code name | AD102 | GA100 |
Launch date | 20 Sep 2022 | 16 Nov 2020 |
Place in performance rating | 8 | 29 |
Technical info |
||
Boost clock speed | 2520 MHz | 1410 MHz |
Core clock speed | 2235 MHz | 1095 MHz |
Manufacturing process technology | 4 nm | 7 nm |
Peak Double Precision (FP64) Performance | 1,290 GFLOPS (1:64) | 9.746 TFLOPS (1:2) |
Peak Half Precision (FP16) Performance | 82.58 TFLOPS (1:1) | 77.97 TFLOPS (4:1) |
Peak Single Precision (FP32) Performance | 82.58 TFLOPS | 19.49 TFLOPS |
Pipelines | 16384 | 6912 |
Pixel fill rate | 443.5 GPixel/s | 225.6 GPixel/s |
Texture fill rate | 1,290 GTexel/s | 609.1 GTexel/s |
Thermal Design Power (TDP) | 450 Watt | 400 Watt |
Transistor count | 76300 million | 54200 million |
Video outputs and ports |
||
Display Connectors | 1x HDMI 2.1, 3x DisplayPort 1.4a | No outputs |
Compatibility, dimensions and requirements |
||
Form factor | Triple-slot | IGP |
Height | 61 mm, 2.4 inches | |
Interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
Length | 304 mm, 12 inches | |
Recommended system power (PSU) | 850 Watt | 800 Watt |
Supplementary power connectors | 1x 16-pin | None |
Width | 137 mm, 5.4 inches | |
API support |
||
DirectX | 12 Ultimate (12_2) | |
OpenCL | 3.0 | 3.0 |
OpenGL | 4.6 | |
Shader Model | 6.7 | |
Vulkan | ||
Memory |
||
Maximum RAM amount | 24 GB | 80 GB |
Memory bandwidth | 1,008 GB/s | 2039 GB/s |
Memory bus width | 384 bit | 5120 bit |
Memory clock speed | 1313 MHz, 21 Gbps effective | 1593 MHz (3.2 Gbps effective) |
Memory type | GDDR6X | HBM2e |
High bandwidth memory (HBM) |