GeForce RTX 5090 vs GTX 960
VS
Nvidia GeForce RTX 5090
GB202-300
Nvidia GeForce GTX 960
GM206
We compared two discrete desktop gaming GPUs: the GeForce RTX 5090 32 GB with 170 pipelines and 21760 shaders against the 10 years and 1 month older GTX 960 2 GB that utilizes 8 pipelines and 1024 shaders. Here you will find complete details about specs, efficiency, performance tests, and more.
Review
Gaming
Performance in DirectX, OpenCL, and Vulkan games
Workstation
Perf. in 3D modeling, video editing and rendering apps
AI/ML
Capabilities for machine learning and AI-related tasks
Energy Efficiency
Power consumption efficiency in different scenarios
NanoReview Final Score
Overall video card score
Value for money (Beta)
Key differences
Reasons to consider the GeForce RTX 5090
- Performs significantly better (up to 24.2x) in 3DMark Steel Nomad Lite
- 43.7x higher maximum theoretical performance (104.8 vs 2.4 TFLOPS)
- Shows 15.9x higher average frame rate in modern games at QHD resolution – 207 vs 13 FPS
- Manufactured using a more efficient 4 nm process technology
- Ray-tracing support with 170 dedicated RT cores
Gaming Performance
Frame rate comparison across popular AAA titles at different resolutionsGames
| Forza Horizon 5 | 336 247 233 187 |
29 14 13 - |
| The Witcher 3 | 364 327 237 123 |
24 18 13 - |
| Counter-Strike 2 | 363 265 202 99 |
46 29 19 - |
| Far Cry 6 | 237 202 190 121 |
31 27 16 10 |
| Hogwarts Legacy | 249 199 166 97 |
18 12 10 - |
| CoD: Modern Warfare III | 334 322 260 184 |
30 24 16 12 |
| Ghost of Tsushima | 221 185 175 113 |
15 - - - |
| Cyberpunk 2077 | 261 233 160 77 |
14 11 - - |
| Shadow of the Tomb Raider | 329 302 238 128 |
28 26 15 - |
| 1080p High | 299 | 26 |
| 1080p Ultra | 254 | 19 |
| 1440p Ultra | 207 | 13 |
| 4K Ultra | 125 | - |
| Margin of Error | Low | Low |
Benchmarks
Graphics cards’ performance in recent benchmarking apps3D Mark
Steel Nomad Lite Score
| Time Spy | 46873 | 2302 |
| Solar Bay | 236689 | - |
| Port Royal | 38202 | - |
| Fire Strike | 87569 | 7720 |
| Wild Life Extreme | 108751 | 4543 |
| Night Raid | 206648 | 31121 |
GeekBench 6 OpenCL
GB6 Compute Score
| Background Blur | 333 img/sec | 51.7 img/sec |
| Face Detection | 298.5 img/sec | 26.1 img/sec |
| Horizon Detection | 21.7 Gpixels/sec | 0.84 Gpixels/sec |
| Edge Detection | 33.7 Gpixels/sec | 1.36 Gpixels/sec |
| Gaussian Blur | 34.5 Gpixels/sec | 1.26 Gpixels/sec |
| Feature Matching | 2.53 Gpixels/sec | 0.14 Gpixels/sec |
| Stereo Matching | 2850 Gpixels/sec | 64.9 Gpixels/sec |
| Particle Physics | 59146.9 FPS | 1825 FPS |
| API | OpenCL | OpenCL |
Passmark Graphics
G3D Mark Score
| G2D Mark | 1410 | 677 |
| DirectX 11 | 340 FPS | 42 FPS |
| DirectX 12 | 179 FPS | 28 FPS |
| GPU Compute | 24383 Ops/s | 2801 Ops/s |
Blender
Blender GPU
Artificial Intelligence Tests
Performance in machine learning and artificial intelligence tasksGeekBench 6 ML
GB6 ML Single Precision
GB6 ML Half Precision
GB6 ML Quantized
| Image Classification (SP) | 18948 | 1492 |
| Image Segmentation (HP) | 55109 | 2183 |
| Image Super Resolution (Q) | 58478 | 3618 |
| Face Detection (HP) | 114685 | 3654 |
| Pose Estimation (Q) | 344729 | 11306 |
| Text Classification (SP) | 4464 | 1315 |
| Machine Translation (HP) | 8184 | 2085 |
| Object Detection (SP) | 27802 | 1927 |
| Depth Estimation (Q) | 85572 | 5491 |
| Style Transfer (SP) | 687080 | 22683 |
| Framework | ONNX | ONNX |
| Backend | DirectML | DirectML |
Specifications
Technical specifications of GeForce RTX 5090 and GTX 960General
| Vendor | Nvidia | Nvidia |
| Build | Discrete | Discrete |
| Released | January 7, 2025 | January 22, 2015 |
| Launch price (MSRP) | $1999 | $199 |
| Case | Desktop | Desktop |
| Purpose | Gaming | Gaming |
| Segment | High-end | Mid-range |
| Architecture | Blackwell 2.0 | Maxwell 2.0 |
| GPU Codename | GB202-300 | GM206 |
| Recommended CPU | - Intel Core Ultra 9 285K or above | - |
Desktop GPU rating (2nd and 90th place)
Graphics Processing Unit
| Base Clock | 2017 MHz | 1127 MHz |
| Boost Clock | 2407 MHz | 1178 MHz |
| Shading Units | 21760 | 1024 |
| Texture Mapping Units (TMUs) | 680 | 64 |
| Render Output Units (ROPs) | 176 | 32 |
| Compute Units (Pipelines) | 170 | 8 |
| Tensor Cores | 680 | No |
| Ray-tracing Cores | 170 | No |
| L1 Cache | 128KB per cluster | 48KB per cluster |
| L2 Cache | 96MB shared | 1MB shared |
| Instructions Per Cycle | 2 IPC | 2 IPC |
Raw Performance
| Pixel Fill Rate | 424 GPixel/s | 38 GPixel/s |
| Texture Fill Rate | 1637 GTexel/s | 75 GTexel/s |
FLOPS (FP32)
Physical
| Interface | PCIe 5.0 x16 | PCIe 3.0 x16 |
| TGP | 575 W | 120 W |
| Manufacturing | TSMC | TSMC |
| Fabrication Process | 4 nm | 28 nm |
| Die Size | 750 mm² | 228 mm² |
| Transistor Count | 92.2 billion | 2 billion |
| Transistor Density | 122.93 MTr/mm² | 8.77 MTr/mm² |
Memory
| Memory Type | GDDR7 | GDDR5 |
| Memory Size | 32 GB | 2 GB |
| Memory Clock | 1750 MHz | 1753 MHz |
| Effective Memory Speed | 28000 Mbps | 7000 Mbps |
| Bus | 512-bit | 128-bit |
| ECC | No | No |
Memory Bandwidth
API
| DirectX | 12.2 | 12 |
| Vulkan | 1.4 | 1.3 |
| OpenGL | 4.6 | 4.6 |
| OpenCL | 3.0 | 3.0 |
| CUDA | 12.0 | 5.2 |
| Ray Tracing | Yes | No |
| DLSS | DLSS 4 | No |
| DisplayPort | 2.1b | 1.2 |
Cast your vote
Total votes: 35
