Quick/Comparisons
16 | Gaming | 51 |
25 | Workstation | 57 |
22 | AI/ML |
Compare with something else?
Compare Intel Arc B390 with another →42 |
100 | Energy Efficiency | 45 |
33 | Quick Comparison Final Score | 51 |
| Panther Lake | Graphics Processor | AD104 |
| 1536 | Cores | 7424 |
| 48 / 24 | TMUs / ROPs | 232 / 80 |
| 1268.79 | Blender GPU | 5009.03 |
| 38 | Forza Horizon 5 | 107 |
| 55 | The Witcher 3 | 152 |
| 60 | Counter-Strike 2 | 200 |
| 49 | Far Cry 6 | 126 |
| 21 | Hogwarts Legacy | 89 |
| 32 | Call of Duty | 132 |
| 25 | Ghost of Tsushima | 82 |
| 33 | Cyberpunk 2077 | 109 |
| 52 | Tomb Raider | 175 |
| 41 | Average FPS | 130 |
| 52 | 1080p High | 157 |
| 41 | 1080p Ultra | 130 |
| 29 | 1440p Ultra | 99 |
| 14 | 4K Ultra | 58 |
| Low | Margin of Error | High |
| 64546 | GB6 Compute Score | 139462 |
| — | Background Blur | 187.1 img/sec |
| — | Face Detection | 139.5 img/sec |
| — | Horizon Detection | 5.71 Gpixels/sec |
| — | Edge Detection | 8.53 Gpixels/sec |
| — | Gaussian Blur | 8.38 Gpixels/sec |
| — | Feature Matching | 1.55 Gpixels/sec |
| — | Stereo Matching | 577.7 Gpixels/sec |
| — | Particle Physics | 16849 FPS |
| — | API | OpenCL |
| 9029 | G3D Mark Score | 22027 |
| 786 | G2D Mark | 721 |
| 70 FPS | DirectX 11 | 215 FPS |
| 50 FPS | DirectX 12 | 96 FPS |
| 3702 Ops/s | GPU Compute | 8635 Ops/s |
| 11871 | GB6 ML Single Precision | 22488 |
| 20120 | GB6 ML Half Precision | 35441 |
| 7879 | GB6 ML Quantized | 16689 |
| 5487 | Image Classification (SP) | 8658 |
| 16864 | Image Segmentation (HP) | 17759 |
| 8788 | Image Super Resolution (Q) | 28281 |
| 25029 | Face Detection (HP) | 37971 |
| 18852 | Pose Estimation (Q) | 113009 |
| 2115 | Text Classification (SP) | 3535 |
| 4266 | Machine Translation (HP) | 4038 |
| 7449 | Object Detection (SP) | 11288 |
| 22207 | Depth Estimation (Q) | 38963 |
| 95039 | Style Transfer (SP) | 236141 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 60 GPixel/s | Pixel Fill Rate | 117-181 GPixel/s |
| 120 GTexel/s | Texture Fill Rate | 340-524 GTexel/s |
| 7.7 TFLOPS | FLOPS (FP32) | 21.8-33.6 TFLOPS |
| System Shared | Memory Type | GDDR6 |
| — | Memory Size | 12 GB |
| — | Memory Clock | 2250 MHz |
| — | Effective Memory Speed | 18000 Mbps |
| — | Bus | 192-bit |
| No | ECC | Yes |
| — | Memory Bandwidth | 432 GB/s |
| PCIe 5.0 | Interface | PCIe 4.0 x16 |
| — | TGP | 60-175 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 3 nm | Fabrication Process | 5 nm |
| — | Die Size | 294 mm² |
| — | Transistor Count | 35.8 billion |
| — | Transistor Density | 121.77 MTr/mm² |
| 12.1 | DirectX | 12 |
| 1.4 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| — | CUDA | 8.9 |
| Yes | Ray Tracing | Yes |
| No | DLSS | DLSS 3 |
| 2.1 | DisplayPort | 1.4a |
| Intel | Vendor | Nvidia |
| Integrated | Build | Discrete |
| January 5, 2026 | Released | March 21, 2023 |
| Laptop | Case | Laptop |
| Gaming | Purpose | Professional |
| Mid-range | Segment | High-end |
| Celestial | Architecture | Ada Lovelace |
| Panther Lake | GPU Codename | AD104 |
| -Intel Core Ultra X9 388Hor above | Recommended CPU | -Intel Core Ultra 9 185Hor above |
| 4242 | Steel Nomad Lite Score | 16926 |
| 6595 | Time Spy | 15516 |
| 26399 | Solar Bay | 68617 |
| 3886 | Port Royal | 10118 |
| 16770 | Fire Strike | 38840 |
| 11707 | Wild Life Extreme | — |
| 68498 | Night Raid | 123923 |
| 300 MHz | Base Clock | — |
| 2500 MHz | Boost Clock | 1465-2260 MHz (configurable) |
| 1536 | Shading Units | 7424 |
| 48 | Texture Mapping Units (TMUs) | 232 |
| 24 | Render Output Units (ROPs) | 80 |
| 12 | Compute Units (Pipelines) | 58 |
| — | Tensor Cores | 232 |
| — | Ray-tracing Cores | 58 |
| 64KB per cluster | L1 Cache | 128KB per cluster |
| 16MB shared | L2 Cache | 48MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |