Tesla M2070

Specifications of the NVIDIA Tesla M2070 GPU

Name / Brand / Architecture
Manufacturer:	NVIDIA
Model:	Tesla M2070
Reference card?:	Yes
Target market segment:	Server / High Performance Computing
Die name:	GF100
Architecture:	Fermi
Fabrication process:	40nm
Transistors:	3.1 billion
Bus interface:	PCI-E 2.0 x 16
Launch date:	July 2011
Frequency
Graphics clock:	575 MHz
Processor clock:	1.15 GHz
Memory specifications
Memory size:	6144 MB
Memory type:	GDDR5
Memory clock:	783 MHz
Memory clock (effective):	3132 MHz
Memory interface width:	384-bit
Memory bandwidth:	150.34 GB/s
Cores / Texture
CUDA:	2.0
CUDA cores:	448
ROPs:	48
Texture units:	56
RAMDACs:	400 MHz
Electric characteristics
Maximum power draw:	225 W
Performance
Pixel fill rate:	27.6 Gigapixels/s
Texture fill rate:	32.2 Gigatexels/s
Single presision compute power:	1030.4 GFLOPS
Double precision compute power:	515.2 GFLOPS
External connectors
Power connectors:	2 x 6-pin
Dimensions
Length:	10.51 inches (267 mm)
Width:	Dual-Slot
Other features / Support
Open CL support:	1.1
OpenGL support:	4.3
DirectX support:	11.0
Shader model:	5.0
Notes
Tesla M2000-series have no graphics output and are designed for GPGPU applications only.
- data is not confirmed.