Product Compare
NVIDIA RTX 4000 Ada Generation vs NVIDIA B300
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Aug 2023 | Sep 2025 |
| Generation | Workstation Ada | Server Blackwell |
| Type | Desktop | Desktop |
| Bus Interface | PCIe 4.0 x16 | PCIe 5.0 x16 |
| Clock Speeds | ||
| Base Clock | 1500 MHz | 1665 MHz |
| Boost Clock | 2175 MHz | 2032 MHz |
| Memory Clock | 2250 MHz | 2000 MHz |
| Memory | ||
| Memory Size | 20GB | 144GB |
| Memory Type | GDDR6 | HBM3e |
| Memory Bus | 160bit | 4096bit |
| Bandwidth | 360.0GB/s | 4.10TB/s |
| Render Config | ||
| SM Count | 48 | 148 |
| Shading Units | 6144 | 18944 |
| TMUs | 192 | 592 |
| ROPs | 64 | 24 |
| Tensor Cores | 192 | 592 |
| RT Cores | 48 | |
| L1 Cache | 128 KB (per SM) | 256 KB (per SM) |
| L2 Cache | 48 MB | 50 MB |
| Theoretical Performance | ||
| Pixel Rate | 139.2 GPixel/s | 48.77 GPixel/s |
| Texture Rate | 417.6 GTexel/s | 1202.9 GTexel/s |
| FP16 (half) | 26.73 TFLOPS | 1231.8 TFLOPS |
| FP32 (float) | 26.73 TFLOPS | 76.99 TFLOPS |
| FP64 (double) | 417.6 GFLOPS | 1202.9 GFLOPS |
| Graphics Processor | ||
| GPU Name | AD104 | GB110 |
| Architecture | Ada Lovelace | Blackwell Ultra |
| Foundry | TSMC | TSMC |
| Process Size | 5 nm | 5 nm |
| Transistors | 35.8 billion | 104 billion |
| Die Size | 294 mm² | Unknown |
| Board Design | ||
| TDP | 130W | 1400W |
| Suggested PSU | 300 W | 1800 W |
| Outputs | 4x DisplayPort 1.4a | No outputs |
| Power Connectors | 1x 16-pin | |
| Graphics Features | ||
| DirectX | 12 Ultimate (12_2) | N/A |
| OpenGL | 4.6 | N/A |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.3 | N/A |
| CUDA | 8.9 | 10.3 |
| Shader Model | 6.8 | N/A |