NVIDIA Jetson TK1 GPU vs NVIDIA Riva TNT
Graphics Processor
| GPU model | GK20A | NV4 |
|---|---|---|
| Architecture | Kepler 2.0 | Fahrenheit |
| Foundry | TSMC | TSMC |
| Manufacturing process | 28 nm | 350 nm |
| Number of transistors | unknown | 7 million |
| Die size | unknown | 90 mm² |
Integrated Graphics
| Release date | Oct 15th, 2014 | — |
|---|---|---|
| Generation | Tegra | — |
| Production | End-of-life | — |
| Interface | IGP | — |
Clocks
| Base clock | 756 MHz | — |
|---|---|---|
| Boost clock | 951 MHz | — |
| Memory clock | System Shared | 110 MHz |
| GPU clock | — | 90 MHz |
Memory Configuration
| Memory size | System Shared | 16 MB |
|---|---|---|
| Memory type | System Shared | SDR |
| Memory bus width | System Shared | 128 bit |
| Bandwidth | System Dependent | 1.760 GB/s |
Render Configuration
| Shading units | 192 | — |
|---|---|---|
| TMUs | 8 | 2 |
| ROPs | 4 | 2 |
| SMX count | 1 | — |
| Cache L1 | 16 KB (per SMX) | — |
| Cache L2 | 128 KB | — |
| Pixel shaders | — | 2 |
| Vertex shaders | — | 0 |
Performance
| Pixel rate | 3.804 GPixel/s | 180.0 MPixel/s |
|---|---|---|
| Texture rate | 7.608 GTexel/s | 180.0 MTexel/s |
| FP32 (float) performance | 365.2 GFLOPS | — |
| FP64 (double) performance | 11.41 GFLOPS (1:32) | — |
Dimensions & Outputs
| Slot width | IGP | Single-slot |
|---|---|---|
| TDP | 8 W | unknown |
| Outputs | No outputs | 1x VGA |
| Suggested PSU | — | 200 W |
| Power connectors | — | None |
API Support & Features
| DirectX | 12 (11_0) | 5.0 |
|---|---|---|
| OpenGL | ES 3.1 | 1.2 |
| OpenCL | — | — |
| Vulkan | 1.1 | — |
| CUDA | 3.2 | — |
| Shader model | 5.1 | — |
| Pixel shader | — | — |
| Vertex shader | — | — |
Other Features
Graphics Card
| Release date | — | Mar 23rd, 1998 |
|---|---|---|
| Generation | — | Riva |
| Production | — | End-of-life |
| Interface | — | AGP 2x |
| Predecessor | — | NV1 |
| Successor | — | TNT2 |
