News & Analysis
Ten technologies that will shake the CE world
EET staff
1/24/2012 10:50 AM EST
GPU compute
In general-purpose computing on graphics processing units, or “GPU compute,” certain computations traditionally handled by a system CPU or application processor are offloaded to the GPU. The addition of programmable pipelines, schedulers and floating-point precision to the graphics rendering pipeline enables GPU-compute technology, but until now a lack of system- and software-level support has hindered its progress. That’s changing with the introduction of APIs and parallel-capable programming languages such as CUDA, DirectX compute, OpenCL, OpenGL Shading Language and Renderscript compute.
Offloading inner parallel loops of programs from the CPU to the GPU can improve performance and save power. The ability of the GPU to lower power consumption, as well as to influence the look and feel of displays, the responsiveness of games and the user interface, makes it potentially more important than the CPU.
The addition of GPU compute to the GPU’s established graphics rendering duties are another step in reducing the CPU to just a housekeeping processor or host. Applications already being computed on GPUs include the physics of moving objects as part of scene calculation prior to rendering; applications that can benefit from GPU compute include math functions, 2- and 3-D field solvers, simulators, encryption, sorting and alignment, and some database functions.
A PowerVR Series 5 GPU can compute the physics of the above scene, including carpet movement, as well as rendering the resultant image. Source: Imagination Technologies Group plc
Enablers of the trend include Nvidia, with its graphics chips and CUDA parallel programming platform; the Khronos industry organization, which provides API definitions such as OpenCL and OpenGL; ARM, with its Mali line of GPUs, including versions (the T604 and T658) that have been architected with GPU compute in mind; and Imagination Technologies, with its PowerVR line of GPU cores.
— Peter Clarke
Next: Ubiquitous Android
In general-purpose computing on graphics processing units, or “GPU compute,” certain computations traditionally handled by a system CPU or application processor are offloaded to the GPU. The addition of programmable pipelines, schedulers and floating-point precision to the graphics rendering pipeline enables GPU-compute technology, but until now a lack of system- and software-level support has hindered its progress. That’s changing with the introduction of APIs and parallel-capable programming languages such as CUDA, DirectX compute, OpenCL, OpenGL Shading Language and Renderscript compute.
Offloading inner parallel loops of programs from the CPU to the GPU can improve performance and save power. The ability of the GPU to lower power consumption, as well as to influence the look and feel of displays, the responsiveness of games and the user interface, makes it potentially more important than the CPU.
The addition of GPU compute to the GPU’s established graphics rendering duties are another step in reducing the CPU to just a housekeeping processor or host. Applications already being computed on GPUs include the physics of moving objects as part of scene calculation prior to rendering; applications that can benefit from GPU compute include math functions, 2- and 3-D field solvers, simulators, encryption, sorting and alignment, and some database functions.
A PowerVR Series 5 GPU can compute the physics of the above scene, including carpet movement, as well as rendering the resultant image. Source: Imagination Technologies Group plc
Enablers of the trend include Nvidia, with its graphics chips and CUDA parallel programming platform; the Khronos industry organization, which provides API definitions such as OpenCL and OpenGL; ARM, with its Mali line of GPUs, including versions (the T604 and T658) that have been architected with GPU compute in mind; and Imagination Technologies, with its PowerVR line of GPU cores.
— Peter Clarke
Next: Ubiquitous Android
|
|
|
|
|
Navigate to related information
chanj
1/24/2012 2:40 PM EST
Spielberg's vision in Minority Report is becoming reality. What's the world is going to be like?
Sign in to Reply
agk
1/25/2012 7:31 AM EST
MEM's doing many wonders. I think soon these sensors will be woven into our fabrics and this will monitor our sitting position in front of the PC's enable to improve our performance and health.
Sign in to Reply
NSK
1/25/2012 2:07 PM EST
First you say that Google's Android will dominate, then you say how Apple's Siri will change everything. I see a bit of a conflict here.
Sign in to Reply
docdivakar
1/25/2012 2:33 PM EST
@NSK: I feel tempted to say neither one! Both aren't open systems -we know iOS isn't but Android isn't either, contrary to the claims! Google controls it but allows the source to be downloaded.
What is truly open is the up & coming Boot-2-Gecko (B2G) from Mozilla. When I met its CEO Gary Kovacs last year, one question I posed to him was-what is the future of browsers in the world of ubiquitous computing? Do they become irrelevant? Each mobile device has its own browser...
His answer was -well, wait and see. I saw a presentation yesterday, at the Stanford Faculty club where some B2G developers showed what it can do.
I hope B2G enjoys the same success as FireFox.
MP Divakar
Sign in to Reply
junko.yoshida
1/25/2012 11:43 PM EST
That is really fascinating. We all want to know more about it!!!
Sign in to Reply
docdivakar
1/30/2012 12:04 AM EST
Junko, I will keep you posted. Had a nice chat with the CTO of Mozilla and also have some presentation materials on B2G.
I understand how Mozilla monetizes FireFox but I am still in the dark about B2G's monetization mnodel.
MP Divakar
Sign in to Reply
junko.yoshida
1/25/2012 11:45 PM EST
we are not saying one is better than others...the idea of Siri-like services will surely spread everywhere over time...
Sign in to Reply
docdivakar
1/25/2012 2:41 PM EST
I generally like the list, at least for technology sake... but I think many are solutions looking for problems. One thing is sure, we are networking the heck out of anything and everything! And losing privacy fast!
MP Divakar
Sign in to Reply
pixies
1/25/2012 8:31 PM EST
And once you expand the connectivity beyond a critical scale intelligence will arise and eventually render humans obsolete.:)
Sign in to Reply
docdivakar
1/30/2012 12:12 AM EST
@pixies: scary(!) thoughts on extrapolation of networking to higher orders. I have read some what on evolutionary and self-organizing networks but I still consider them dependent on human intervention at least at several phases, for now.
Human beings are already being rendered useless on several fronts with the advances in technology. We are supposed to advance in intellectual thought and their application to work life so we can justify the need for human interaction with processes & tools (in short, work!) but that line of argument seems to be struggling for validation, in some sectors. More automation is rendering human interaction with machines & tools unwanted. I honestly don't know where this stops!
MP Divakar
Sign in to Reply
GREAT-Terry
1/25/2012 9:54 PM EST
The advance of MEMS technology is so amazing!
Sign in to Reply
t.alex
1/26/2012 10:15 AM EST
For the ARMing of Windows, i wonder if Microsoft has any way to help developers port their apps to ARM conveniently?
Sign in to Reply
tina_jeffrey
1/27/2012 10:08 AM EST
Just wanted to draw attention to CogniVue's - also founding member of EVA - latest Smart Back-Up Camera Application dewarping, object detection & distance estimation running on a single CV2201 processor - 9x9mm2 incl sys mem dissipating ~250mW. How's that for 'powerful, low-cost, energy efficient processors as key enablers of this technology'. Check it out on http://www.youtube.com/user/cognivue/videos
Sign in to Reply
Attoman
10/31/2012 12:10 PM EDT
All in the eyes.
For some of us in Berkeley it has been a twenty year wait for the perception that MEMS is emergent.
This insures that the actual core inventors get neither credit nor monetary reward.
Sign in to Reply