Apple A18 Pro GPU vs GeForce RTX 5090
[ VERDICT ]
Winner: GeForce RTX 5090 (96/100)
GeForce RTX 5090 leads overall with a score of 96/100 compared to Apple A18 Pro GPU (28/100), offering superior measured performance and technical specifications.
Why choose Apple A18 Pro GPU:
- + Higher Efficiency score (95 vs 48)
Why choose GeForce RTX 5090:
- + Higher Gaming score (100 vs 11)
Benchmark Performance Comparison
Normalized Lab Scores| Benchmark | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| Efficiency | 95.0 Excellent WINNER | 48.0 Poor |
| Gaming | 11.0 Poor | 100.0 Excellent WINNER |
Detailed Technical Specifications
3D Mark
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| Solar Bay | 7374 | — |
| Wild Life Extreme | 3947 | — |
| Fire Strike | — | 87569 |
| Night Raid | — | 206648 |
| Port Royal | — | 38202 |
| Solar Bay | — | 236689 |
| Time Spy | — | 46873 |
| Wild Life Extreme | — | 108751 |
API
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| DLSS | No | DLSS 4 |
| Ray Tracing | Yes | — |
| CUDA | — | 12.0 |
| DirectX | — | 12.2 |
| DisplayPort | — | 2.1b |
| OpenCL | — | 3.0 |
| OpenGL | — | 4.6 |
| Ray Tracing | — | Yes |
| Vulkan | — | 1.4 |
GeekBench 6 GPU Compute
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| API | Metal | OpenCL |
| Background Blur | 57.1 img/sec | — |
| Edge Detection | 1.21 Gpixels/sec | — |
| Face Detection | 91.4 img/sec | — |
| Feature Matching | 0.45 Gpixels/sec | — |
| Gaussian Blur | 1.5 Gpixels/sec | — |
| Horizon Detection | 0.92 Gpixels/sec | — |
| Particle Physics | 3061 FPS | — |
| Stereo Matching | 84.2 Gpixels/sec | — |
GeekBench 6 ML
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| Backend | GPU | DirectML |
| Depth Estimation (Q) | 13468 | — |
| Face Detection (HP) | 11770 | — |
| Framework | Core ML | ONNX |
| Image Classification (SP) | 3264 | — |
| Image Segmentation (HP) | 4550 | — |
| Image Super Resolution (Q) | 6705 | — |
| Machine Translation (HP) | 5678 | — |
| Object Detection (SP) | 3360 | — |
| Pose Estimation (Q) | 15548 | — |
| Style Transfer (SP) | 31541 | — |
| Text Classification (SP) | 3678 | — |
| Depth Estimation (Q) | — | 85572 |
| Face Detection (HP) | — | 114685 |
| Image Classification (SP) | — | 18948 |
| Image Segmentation (HP) | — | 55109 |
| Image Super Resolution (Q) | — | 58478 |
| Machine Translation (HP) | — | 8184 |
| Object Detection (SP) | — | 27802 |
| Pose Estimation (Q) | — | 344729 |
| Style Transfer (SP) | — | 687080 |
| Text Classification (SP) | — | 4464 |
General
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| Architecture | Apple M GPU | Blackwell 2.0 |
| Build | Integrated | Discrete |
| Case | Laptop | Desktop |
| GPU Codename | Custom | — |
| Purpose | Generic | Gaming |
| Recommended CPU | -Apple A18 Pro (MacBook)or above | — |
| Released | September 9, 2024 | January 7, 2025 |
| Segment | Mid-range | High-end |
| Used in CPUs | -Apple A18 Pro (MacBook) | — |
| Vendor | Apple | Nvidia |
| GPU Codename | — | GB202-300 |
| Launch price (MSRP) | — | $1999 |
| Recommended CPU | — | -Intel Core Ultra 9 285Kor above |
Graphics Processing Unit
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| Boost Clock | 1490 MHz | — |
| Compute Units (Pipelines) | 128 | — |
| Instructions Per Cycle | 2 IPC | — |
| Shading Units | 768 | — |
| Texture Mapping Units (TMUs) | 40 | — |
| Base Clock | — | 2017 MHz |
| Boost Clock | — | 2407 MHz |
| Compute Units (Pipelines) | — | 170 |
| Instructions Per Cycle | — | 2 IPC |
| L1 Cache | — | 128KB per cluster |
| L2 Cache | — | 96MB shared |
| Ray-tracing Cores | — | 170 |
| Render Output Units (ROPs) | — | 176 |
| Shading Units | — | 21760 |
| Tensor Cores | — | 680 |
| Texture Mapping Units (TMUs) | — | 680 |
Memory
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| Bus | 128-bit | 512-bit |
| ECC | No | No |
| Memory Clock | 7500 MHz | — |
| Memory Type | System Shared | — |
| Effective Memory Speed | — | 28000 Mbps |
| Memory Clock | — | 1750 MHz |
| Memory Size | — | 32 GB |
| Memory Type | — | GDDR7 |
Physical
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| Fabrication Process | 3 nm | — |
| Interface | Custom | PCIe 5.0 x16 |
| Manufacturing | TSMC | TSMC |
| Die Size | — | 750 mm² |
| Fabrication Process | — | 4 nm |
| TGP | — | 575 W |
| Transistor Count | — | 92.2 billion |
| Transistor Density | — | 122.93 MTr/mm² |
Raw Performance
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| Texture Fill Rate | 60 GTexel/s | — |
| Pixel Fill Rate | — | 424 GPixel/s |
| Texture Fill Rate | — | 1637 GTexel/s |
Recent User Tests
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| Date | Benchmark | — |
| 📘 2026-04-24->Rennsport | GeekBench | — |
Games
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| 1080p High | — | 299 |
| 1080p Ultra | — | 254 |
| 1440p Ultra | — | 207 |
| 4K Ultra | — | 125 |
| CoD: Modern Warfare III | — | 334322260184 |
| Counter-Strike 2 | — | 36326520299 |
| Cyberpunk 2077 | — | 26123316077 |
| Far Cry 6 | — | 237202190121 |
| Forza Horizon 5 | — | 336247233187 |
| Ghost of Tsushima | — | 221185175113 |
| Hogwarts Legacy | — | 24919916697 |
| Margin of Error | — | Low |
| Shadow of the Tomb Raider | — | 329302238128 |
| The Witcher 3 | — | 364327237123 |
GeekBench 6 OpenCL
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| API | Metal | OpenCL |
| Background Blur | — | 333 img/sec |
| Edge Detection | — | 33.7 Gpixels/sec |
| Face Detection | — | 298.5 img/sec |
| Feature Matching | — | 2.53 Gpixels/sec |
| Gaussian Blur | — | 34.5 Gpixels/sec |
| Horizon Detection | — | 21.7 Gpixels/sec |
| Particle Physics | — | 59146.9 FPS |
| Stereo Matching | — | 2850 Gpixels/sec |
Passmark Graphics
| Specification | Apple A18 Pro GPU | GeForce RTX 5090 |
|---|---|---|
| DirectX 11 | — | 341 FPS |
| DirectX 12 | — | 179 FPS |
| G2D Mark | — | 1410 |
| GPU Compute | — | 24360 Ops/s |
Looking for More Comparisons?
Browse all available head-to-head comparisons or select new devices in the comparison hub.