Difference between revisions of "Jetson/Graphics Performance"

From eLinux.org
Jump to: navigation, search
(Graphics Performance)
 
(Checkpoint save)
Line 1: Line 1:
== The [[Jetson TK1]] GPU ==
+
== The [[Jetson TK1]] SOC ==
The [[Tegra/SoCs#Tegra124|Tegra K1]] SOC in [[Jetson TK1]] is targeted for mobile <abbr title="General Purpose Computing on Graphics Processing Units">GPGPU</abbr> embedded solutions as well as general purpose processor for Tablets, Laptops, Desktop, and Super Phones.   
+
The [[Tegra/SoCs#Tegra124|Tegra K1]] <abbr title="System-On-a-Chip">SOC</abbr> in the [[Jetson TK1]] is targeted for embedded <abbr title="General Purpose Computing on Graphics Processing Units">GPGPU</abbr> applications as well as general purpose use in Tablets, Laptops, Desktop, and Super Phones.   
  
 
== Competitive Positioning ==
 
== Competitive Positioning ==
The GPGPU capability is probably greatest competitive advantage for the TK1 SOC. Nvidia claim Tegra TK1 reaches 326 GFLOPS of performance, compared to [http://kyokojap.myweb.hinet.net/gpu_gflops/|an estimated 200 GFLOPS] for the competing SnapDragon 805.
+
The Tegra K1 SOC GPU provides both excellent graphics and GPGPU performance.  The GPGPU capability is probably greatest competitive advantage. Nvidia claims Tegra TK1 can attain 326 GFLOPS, whereas its closest contemporary competitor, the SnapDragon 805, may achieve [http://kyokojap.myweb.hinet.net/gpu_gflops/|an estimated 200 GFLOPS].
 +
Imagination Technologies [http://www.macrumors.com/2014/02/24/rogue-series6xt-ios-graphics/|have announced] a PowerVR GX6650 which favors <abbr title="Floating Point 16-bit">FP16</abbr> operations.  At best it might be competitive with the Tegra K1 in GPGPU.  However, it had not yet been sampled or bench marked, and may not ship until 2015.
  
NVidia have a strong ecosystem and experience in the GPGPU space.  The same Kepler architecture used in the K1 is used to power many of the worlds fastest super computers.  NVidia [http://www.anandtech.com/show/7622/nvidia-tegra-k1/3|leap-frogs the competition] with superior drivers refined over years for Desktop and Super Computer use. Notable capabilities include OpenGL ES 3.0, CUDA 6, OpenCL 1.2, and OpenGL 4.4, and a bevy of other GPGPU tools  
+
NVidia have a strong ecosystem and experience in the GPGPU space.  NVidia [http://www.anandtech.com/show/7622/nvidia-tegra-k1/3|has leapt over the competition] by using the same Kepler GPU architecture that is used to power desktop GPUs and many of the worlds fastest super computers. This allows them to leverage existing (and well-tested!) code to provide superior breadth of capabilities, including OpenGL ES 3.0, CUDA 6, OpenCL 1.2, OpenGL 4.4, and a bevy of other GPGPU tools  
 
 
Imagination Technologies [http://blog.imgtec.com/powervr/powervr-gx6650-redefining-performance-mobile-192-cores|have made compromises] to their support of <abbr title="Floating Point 32-bit">FP32</abbr> and FP64 in their latest GPU. So whereas they claim the same GFLOP potential as the TK1 Kepler GPU, the
 
  
 
== Frames Per Seconds (FPS) Comparisons ==
 
== Frames Per Seconds (FPS) Comparisons ==
Coming soon.
+
More details soon.
  
 
== Power Consumption ==
 
== Power Consumption ==

Revision as of 13:28, 21 June 2014

The Jetson TK1 SOC

The Tegra K1 SOC in the Jetson TK1 is targeted for embedded GPGPU applications as well as general purpose use in Tablets, Laptops, Desktop, and Super Phones.

Competitive Positioning

The Tegra K1 SOC GPU provides both excellent graphics and GPGPU performance. The GPGPU capability is probably greatest competitive advantage. Nvidia claims Tegra TK1 can attain 326 GFLOPS, whereas its closest contemporary competitor, the SnapDragon 805, may achieve estimated 200 GFLOPS. Imagination Technologies announced a PowerVR GX6650 which favors FP16 operations. At best it might be competitive with the Tegra K1 in GPGPU. However, it had not yet been sampled or bench marked, and may not ship until 2015.

NVidia have a strong ecosystem and experience in the GPGPU space. NVidia leapt over the competition by using the same Kepler GPU architecture that is used to power desktop GPUs and many of the worlds fastest super computers. This allows them to leverage existing (and well-tested!) code to provide superior breadth of capabilities, including OpenGL ES 3.0, CUDA 6, OpenCL 1.2, OpenGL 4.4, and a bevy of other GPGPU tools

Frames Per Seconds (FPS) Comparisons

More details soon.

Power Consumption

According to my testing, graphics performance on the Jeston TK1 is about on par with Intel HD 4600 graphics, but with superior OpenGL and GPGPU capabilities. Power draw for graphic-intensive tasks has been surprisingly low. It appears that typical graphical applications probably default to lower-power FP16 operations. GPGPU tasks, by comparison, do increase power draw significantly.

More details soon.