RTX A5000
Graphics Card RTX A5000 — Specifications, Tests & Benchmarks
RTX A5000
GeForce Ax000
•
GA102
•
12.04.2021
•
24 GB
•
GDDR6
•
384 bit
•
230W
🏆 Top #4341
Performance
Synthetic Score
17,967
PassMark
23,015
Efficiency
high
The RTX A5000 graphics card from NVIDIA is built on the Ampere architecture and was officially introduced on 12.04.2021. At the core of the card is a chip created with modern lithography. The die houses 28.3 billion transistorsistors transistors, and its area is 628 mm². The core configuration includes 8192 shader units, 64 RT Cores for ray tracing, and 256 Tensor Cores for AI tasks. The graphics card is equipped with 24 GB of GDDR6 memory. A 384 bit-bit bus is used, providing bandwidth at 768.00 GB/s. The graphics chip operates at a base clock of 1170 MHz with dynamic acceleration capability up to 1695 MHz. The card's power consumption (TDP) is stated at 230 W. The manufacturer recommends using a power supply unit with a wattage of at least 550 W.
General Information
| Release Date | 12.04.2021 |
| Usage Class | Hardcore Gaming |
| Market Segment | Server / Workstation |
| Series | GeForce Ax000 |
GPU
| Architecture | Ampere |
| Codename | GA102 |
| Process Node | 8 nm (Samsung) |
| Transistor Count | 28,300,000.0 million |
| Die Size | 628 mm² |
| Transistor Density | 45,100,000 million / mm² |
Rendering Configuration
| Shading Units (ALU) | 8192 |
| Texture Mapping Units (TMU) | 256 |
| Render Output Units (ROP) | 96 |
| Tensor Cores | 256 |
| RT Cores (Ray Tracing) | 64 |
| Compute Units (SM/CU) | 64 |
Clock Speeds
| Base Clock | 1170 MHz |
| Boost Clock | 1695 MHz |
| Memory Clock (Actual) | 2000 MHz |
| Memory Clock (Effective) | 16000 MHz |
Memory
| Memory Capacity | 24 GB |
| Memory Type | GDDR6 |
| Memory Bus | 384 bit bit |
| Memory Bandwidth | 768.00 GB/s |
Cache Memory
| L1 Cache | 8192 KB |
| L2 Cache | 6 MB |
Board Design & Power
| Interface | PCIe 4.0 x16 |
| Slot Width | 2 |
| Dimensions (L x W x H) | |
| Power Consumption (TDP) | 230 W |
| Recommended PSU | 550 W |
| Power Connectors | 1x 8-pin |
| Board Number | PG132 SKU 500 |
API Support
| DirectX | 12 Ultimate (12_2) |
| OpenGL | 4.6 |
| OpenCL | |
| Vulkan | 1.4 |
| CUDA | |
| Shader Model |
Theoretical Performance
| Pixel Fill Rate | 162700 GPixel/s |
| Texture Fill Rate | 433900 GTexel/s |
| FP16 Performance (Half) | 27.77 TFLOPS (1:1) |
| FP32 Performance (Float) | 27.77 TFLOPS |
| FP64 Performance (Double) | 433.9 GFLOPS (1:64) |
Synthetic Tests & Benchmarks
| PassMark (G3D Mark) | 23 015 |
| PassMark (G2D Mark) | 1 065 |
| TPU Index Score | 36.46% |
| Performance Tier | high |
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.