NVIDIA Tesla K8 vs NVIDIA GeForce GTX 660 Rev. 2

Comparative analysis of NVIDIA Tesla K8 and NVIDIA GeForce GTX 660 Rev. 2 videocards for all known characteristics in the following categories: Essentials, Technical info, Video outputs and ports, Compatibility, dimensions and requirements, API support, Memory. Benchmark videocards performance analysis: CompuBench 1.5 Desktop - Face Detection (mPixels/s), CompuBench 1.5 Desktop - T-Rex (Frames/s).

NVIDIA Tesla K8

NVIDIA GeForce GTX 660 Rev. 2

Differences

Reasons to consider the NVIDIA Tesla K8

Around 26% higher texture fill rate: 103.8 GTexel / s vs 82.56 GTexel / s
Around 60% higher pipelines: 1536 vs 960
Around 26% better floating-point performance: 2,491 gflops vs 1,981 gflops
Around 40% lower typical power consumption: 100 Watt vs 140 Watt
4x more maximum memory size: 8 GB vs 2 GB

Texture fill rate	103.8 GTexel / s vs 82.56 GTexel / s
Pipelines	1536 vs 960
Floating-point performance	2,491 gflops vs 1,981 gflops
Thermal Design Power (TDP)	100 Watt vs 140 Watt
Maximum memory size	8 GB vs 2 GB

Reasons to consider the NVIDIA GeForce GTX 660 Rev. 2

Around 41% higher core clock speed: 980 MHz vs 693 MHz
Around 27% higher boost clock speed: 1032 MHz vs 811 MHz
Around 20% higher memory clock speed: 6008 MHz vs 5000 MHz

Core clock speed	980 MHz vs 693 MHz
Boost clock speed	1032 MHz vs 811 MHz
Memory clock speed	6008 MHz vs 5000 MHz

Compare benchmarks

GPU 1: NVIDIA Tesla K8
GPU 2: NVIDIA GeForce GTX 660 Rev. 2

Name	NVIDIA Tesla K8	NVIDIA GeForce GTX 660 Rev. 2
CompuBench 1.5 Desktop - Face Detection (mPixels/s)		39.216
CompuBench 1.5 Desktop - T-Rex (Frames/s)		1.173

Compare specifications (specs)

	NVIDIA Tesla K8	NVIDIA GeForce GTX 660 Rev. 2
Essentials
Architecture	Kepler	Kepler
Code name	GK104	GK104
Launch date	16 September 2014	13 September 2014
Place in performance rating	not rated	1506
Type	Workstation	Desktop
Technical info
Boost clock speed	811 MHz	1032 MHz
Core clock speed	693 MHz	980 MHz
Floating-point performance	2,491 gflops	1,981 gflops
Manufacturing process technology	28 nm	28 nm
Pipelines	1536	960
Texture fill rate	103.8 GTexel / s	82.56 GTexel / s
Thermal Design Power (TDP)	100 Watt	140 Watt
Transistor count	3,540 million	3,540 million
Video outputs and ports
Display Connectors	No outputs	2x DVI, 1x HDMI, 1x DisplayPort
Compatibility, dimensions and requirements
Interface	PCIe 2.0 x16	PCIe 3.0 x16
Length	241 mm	241 mm
Supplementary power connectors	1x 6-pin	1x 6-pin
API support
DirectX	12.0 (11_0)	12.0 (11_0)
OpenGL	4.6	4.6
Memory
Maximum RAM amount	8 GB	2 GB
Memory bandwidth	160.0 GB / s	144.2 GB / s
Memory bus width	256 Bit	192 Bit
Memory clock speed	5000 MHz	6008 MHz
Memory type	GDDR5	GDDR5