NVIDIA GeForce GTS 250 vs NVIDIA GRID K100
Graphics Processor
| GPU model | G92B | GK107 |
|---|---|---|
| GPU variant | G92-426-B1 | — |
| Architecture | Tesla | Kepler |
| Foundry | TSMC | TSMC |
| Manufacturing process | 55 nm | 28 nm |
| Number of transistors | 754 million | 1,270 million |
| Die size | 260 mm² | 118 mm² |
Graphics Card
| Release date | Mar 4th, 2009 | Jun 28th, 2013 |
|---|---|---|
| Generation | GeForce 200 | GRID |
| Production | End-of-life | End-of-life |
| Launch price | 199 USD | 63 USD |
| Interface | PCIe 2.0 x16 | PCIe 3.0 x16 |
| Reviews | 58 in our database | — |
| Predecessor | GeForce 9 | — |
| Successor | GeForce 400 | — |
Clocks
| GPU clock | 675 MHz | 850 MHz |
|---|---|---|
| Shader clock | 1620 MHz | — |
| Memory clock | 1008 MHz 2 Gbps effective | 891 MHz 1782 Mbps effective |
Memory Configuration
| Memory size | 1024 MB | 256 MB |
|---|---|---|
| Memory type | GDDR3 | DDR3 |
| Memory bus width | 256 bit | 128 bit |
| Bandwidth | 64.51 GB/s | 28.51 GB/s |
Render Configuration
| Shading units | 128 | 192 |
|---|---|---|
| TMUs | 64 | 16 |
| ROPs | 16 | 16 |
| SM count | 16 | — |
| Cache L2 | 64 KB | 256 KB |
| SMX count | — | 1 |
| Cache L1 | — | 16 KB (per SMX) |
Performance
| Pixel rate | 10.80 GPixel/s | 3.400 GPixel/s |
|---|---|---|
| Texture rate | 43.20 GTexel/s | 13.60 GTexel/s |
| FP32 (float) performance | 414.7 GFLOPS | 326.4 GFLOPS |
| FP64 (double) performance | — | 13.60 GFLOPS (1:24) |
Dimensions & Outputs
| Slot width | Dual-slot | Dual-slot |
|---|---|---|
| Length | 229 mm 9 inches | — |
| TDP | 150 W | 130 W |
| Suggested PSU | 450 W | 300 W |
| Outputs | 2x DVI | No outputs |
| Power connectors | 1x 6-pin | — |
| Board number | P365 | — |
API Support & Features
| DirectX | 11.1 (10_0) | 12 (11_0) |
|---|---|---|
| OpenGL | 3.3 | 4.6 |
| OpenCL | 1.1 | 3.0 |
| Vulkan | — | 1.1 |
| CUDA | 1.1 | 3.0 |
| Shader model | 4.0 | 5.1 |
