900-21030-2220-000 NVIDIA Tesla C2070 Graphics Card Review
The Next-Generation CUDA Architecture GPU
Based on the new NVIDIA CUDA GPU architecture codenamed “Fermi”, the 900-21030-2220-000 NVIDIA Tesla C2070 Graphics Card has built from the ground up for high performance computing. It supports “must have” features for technical and enterprise computing and VR Gaming.
Being a dual-slot graphics card, the NVIDIA Tesla C2070 draws power from 1x 8-pin power connector, with power draw rated at 238 W maximum. Display outputs include: 1x DVI. Tesla C2070 GPU is connected to the rest of the system using a PCI-Express 2.0 x16 interface. The card measures 248 mm in length, and features a dual-slot cooling solution.
It features 448 shading units, 56 texture mapping units, and 48 ROPs. NVIDIA has paired 6 GB GDDR5 memory with the Tesla C2070, which are connected using a 384-bit memory interface. The GPU is operating at a frequency of 574 MHz; memory is running at 747 MHz (3 Gbps effective). The card has 4 GPCs each capable of 8 pixels per clock. This limits complete GPU to 32 pixels per clock and because of that it can’t feed all 48 ROPs when they all require data at the same time.
900-21030-2220-000 Technical Specifications:
| Manufacturer | NVIDIA |
| Part Number | 900-21030-2220-000 |
| Product Type | Desktop Graphics Card |
| Architecture | Fermi |
| Chipset Manufacturer | NVIDIA |
| GPU Model | Tesla C2070 |
| Memory Size | 6 GB |
| Type of Memory | GDDR5 |
| Data-Width of Memory | 384-bit |
| Interface | PCI Express 2.0 x16 |
| Connectors/Ports | DVI-D |
900-21030-2220-000 Key Features:
- ECC Memory
- 448 CUDA Cores
- Asynchronous Transfer
- NVIDIA Parallel Data-Cache
- NVIDIA Giga-Thread Engine
- Desktop Cluster Performance
- Up to 6GB of GDDR5 memory per GPU
Also Read:
Memory4less Official Blog 