Report an Error

NVIDIA Tesla D870

NVIDIA Tesla D870 Image NVIDIA G80 Image
The Tesla D870 was a high-end professional graphics card by NVIDIA, launched in May 2007. Built on the 90 nm process, and based on the G80 graphics processor, the card supports DirectX 10.0. The G80 graphics processor is a large chip with a die area of 484 mm² and 681 million transistors. Tesla D870 combines two graphics processors to increase performance. It features 128 shading units, 32 texture mapping units and 24 ROPs per GPU. NVIDIA has placed 3,072 MB GDDR3 memory on the card, which are connected using a 384-bit memory interface per GPU (each GPU manages 1,536 MB). The GPU is operating at a frequency of 600 MHz, memory is running at 800 MHz.
We recommend the NVIDIA Tesla D870 for gaming with highest details at resolutions up to, and including, 1920x1080.
Being a dual-slot card, its power draw is rated at 520 W maximum. Tesla D870 is connected to the rest of the system using a PCIe 1.0 x16 interface. Its price at launch was 7,499 US Dollars.

Graphics Processor

GPU Name: G80
Architecture: Tesla
Process Size: 90 nm
Transistors: 681 million
Die Size: 484 mm²

Graphics Card

Released: May 2nd, 2007
Production Status: End-of-life
Launch Price: 7,499 USD
Bus Interface: PCIe 1.0 x16

Clock Speeds

GPU Clock: 600 MHz
Shader Clock: 1350 MHz
Memory Clock: 800 MHz
1600 MHz effective


Memory Size: 1536 MB x2
Memory Type: GDDR3
Memory Bus: 384 bit x2
Bandwidth: 76.8 GB/s x2

Render Config

Shading Units: 128 x2
TMUs: 32 x2
ROPs: 24 x2
SM Count: 8 x2
Pixel Rate: 14.40 GPixel/s x2
Texture Rate: 19.20 GTexel/s x2
Floating-point performance: 345.6 GFLOPS x2

Board Design

Slot Width: Dual-slot
TDP: 520 W

Graphics Features

DirectX: 10.0
OpenGL: 3.3
OpenCL: n/a
CUDA: 1.0
Shader Model: 4.0


Find graphics card BIOS for this card.

GPU-Z Validation

Find GPU-Z validations for this card.

G80 GPU Notes

Each Streaming Multiprocessor(SM) in the chip architecture contains 8 SPs and 2 SFUs.

Each SP can fulfill up to two single precision operations MAD per clock.

Each SFU can fulfill up to four operations SF per clock (these units can also handle single-precision floating-point multiplications per clock).

The approximate ratio of operations MAD to operations SF is equal 2:1.

The theoretical SP + SFU performance in single-precision floating point operations
[FLOPSsp + sfu, GFLOPS] of the graphics card with shader count [n] and shader frequency [f, GHz], is estimated by the following formula: FLOPSsp+sfu; f × n × 3.

Alternative formula: FLOPSsp+sfu; f × m × ( 8 SPs * 2 (MAD) + 4 * 2 SFUs ). [m] - SM count.

SP - Shader Processor (Unified Shader, CUDA Core)
SFU - Special Function Unit
SM - Streaming Multiprocessor