NVIDIA Tesla K8: specs and benchmarks

NVIDIA Tesla K8

Summary

NVIDIA started Tesla K8 sales 16 September 2014. This is Kepler architecture desktop card based on 28 nm manufacturing process and primarily aimed at designers. 8 GB of GDDR5 memory clocked at 5 GHz are supplied, and together with 256 Bit memory interface this creates a bandwidth of 160.0 GB / s.

Compatibility-wise, this is 1-slot card attached via PCIe 2.0 x16 interface. Its manufacturer default version has a length of 241 mm. 1x 6-pin power connector is required, and power consumption is at 100 Watt.

It provides poor gaming and benchmark performance at 16.06 percent of a leader's which is NVIDIA Quadro RTX 6000.

General info

Of Tesla K8's architecture, market type and release date.
Place in performance rating
 
Architecture
Kepler
 
Code name
GK104
 
Type
Workstation
 
Release date
16 September 2014
 

Technical specs

Tesla K8's general performance parameters such as number of shaders, GPU core clock, manufacturing process, texturing and calculation speed. These parameters indirectly speak of Tesla K8's performance, but for precise assessment you have to consider its benchmark and gaming test results.
Pipelines
1536
 
Core clock speed
693 MHz
 
of 1620 (Quadro RTX 5000)
Boost Clock
811 MHz
 
of 1815 (Quadro RTX 5000)
Transistor count
3,540 million
 
of 21100
Manufacturing process technology
28 nm
 
Power consumption (TDP)
100 Watt
 
of 900 (Tesla S2050)
Texture fill rate
103.8 GTexel / s
 
Floating-point performance
2,491 gflops
 
of 16384 (Radeon Pro Duo)

Compatibility, dimensions and requirements

Information on Tesla K8's compatibility with other computer components. Useful when choosing a future computer configuration or upgrading an existing one. For desktop video cards it's interface and bus (motherboard compatibility), additional power connectors (power supply compatibility).
Interface
PCIe 2.0 x16
 
Length
241 mm
 
Supplementary power connectors
1x 6-pin
 

Memory

Parameters of memory installed on Tesla K8: its type, size, bus, clock and resulting bandwidth. Note that GPUs integrated into processors don't have dedicated memory and use a shared part of system RAM.
Memory type
GDDR5
 
Maximum RAM amount
8 GB
 
Memory bus width
256 Bit
 
of 8192 (Radeon Pro Duo)
Memory clock speed
5000 MHz
 
of 14000
Memory bandwidth
160.0 GB / s
 
of 1024 (Radeon Pro Duo)

Video outputs and ports

Types and number of video connectors present on Tesla K8.
Display Connectors
No outputs
 

API support

APIs supported by Tesla K8, sometimes including their particular versions.
DirectX
12.0 (11_0)
 
OpenGL
4.6
 

Benchmark performance

Non-gaming benchmarks performance of Tesla K8. Note that overall benchmark performance is measured in points in 0-100 range.

Overall benchmark performance

16.06

Octane Render OctaneBench

38

Game benchmarks

Let's see how good Tesla K8 is for gaming. Particular gaming benchmark results are measured in FPS. Comparisons with system requirements are included, but remember that sometimes official requirements may reflect real performance inaccurately.

Relative perfomance

Overall Tesla K8 performance compared to its nearest competitors. All GPU types participate.

AMD equivalent

We believe that the nearest equivalent to Tesla K8 from AMD is Radeon PRO WX 3100, which is slower by 2% and lower by 2 positions in our rating.

Advices and comments

Here you can ask a question about Tesla K8, agree or disagree with our judgements, or report an error or mismatch.