CMP 170HX 8 GB
Graphics Card CMP 170HX 8 GB — Specifications, Tests & Benchmarks
CMP 170HX 8 GB
GeForce Mining GPUs
•
GA100
•
01.09.2021
•
8 GB
•
HBM2e
•
4096 bit
•
250W
🏆 Top #12300
Performance
Synthetic Score
959
Efficiency
low
The CMP 170HX 8 GB (Ampere) model joined the NVIDIA lineup of graphics accelerators on 01.09.2021. The graphics processor is manufactured using a modern process. The 826 mm² die contains approximately 54.2 billion transistorsistors transistors. The core configuration includes 4480 shader units, 0 RT Cores for ray tracing, and 280 Tensor Cores for AI tasks. The memory subsystem features HBM2e modules with a total capacity of 8 GB. The bus width is 4096 bit bits, and the peak data exchange rate is 1525.76 GB/s. The GPU's base clock speed is 1140 MHz, and in Boost mode, it can automatically increase up to 1410 MHz. The card's power consumption (TDP) is stated at 250 W. The manufacturer recommends using a power supply unit with a wattage of at least 600 W.
General Information
| Release Date | 01.09.2021 |
| Usage Class | Office Use |
| Market Segment | Mobile / Laptop |
| Series | GeForce Mining GPUs |
| GPU Variant | GA100-105F-A1 |
| Integrated Graphics | Yes |
GPU
| Architecture | Ampere |
| Codename | GA100 |
| Process Node | 7 nm (TSMC) |
| Transistor Count | 54,200,000.0 million |
| Die Size | 826 mm² |
| Transistor Density | 65,599,999 million / mm² |
Rendering Configuration
| Shading Units (ALU) | 4480 |
| Texture Mapping Units (TMU) | 280 |
| Render Output Units (ROP) | 128 |
| Tensor Cores | 280 |
| Compute Units (SM/CU) | 70 |
Clock Speeds
| Base Clock | 1140 MHz |
| Boost Clock | 1410 MHz |
| Memory Clock (Actual) | 1458 MHz |
| Memory Clock (Effective) | 2900 MHz |
Memory
| Memory Capacity | 8 GB |
| Memory Type | HBM2e |
| Memory Bus | 4096 bit bit |
| Memory Bandwidth | 1525.76 GB/s |
Cache Memory
| L1 Cache | 13440 KB |
| L2 Cache | 8 MB |
Board Design & Power
| Interface | PCIe 1.0 x4 |
| Slot Width | 0 |
| Power Consumption (TDP) | 250 W |
| Recommended PSU | 600 W |
| Power Connectors | 2x 8-pin |
API Support
| OpenCL | |
| CUDA |
Theoretical Performance
| Pixel Fill Rate | 180500 GPixel/s |
| Texture Fill Rate | 394800 GTexel/s |
| FP16 Performance (Half) | 50.53 TFLOPS (4:1) |
| FP32 Performance (Float) | 12.63 TFLOPS |
| FP64 Performance (Double) | 6.32 TFLOPS (1:2) |
Synthetic Tests & Benchmarks
| TPU Index Score | 21.12% |
| Performance Tier | low |
Всі тести
Тестів поки немає.
Comments
0Please log in to your account to leave a comment.
Log In or RegisterNo comments yet. Be the first to comment!
Are you sure you want to delete this comment? This action cannot be undone.