Is it possible to use OpenCL for PowerVR SGX530 GPU device? I have to write image recognition software that can run on Droid X smartphone. I would greatly appreciate it if someone could provide links, references, citations, sample code.
It seems it is, but it depends on the SOC vendor, have a look at this:
http://www.imgtec.com/forum/forum_posts.asp?TID=194
Imagination Technologies say that the gpu has OpenCL 1.0 embedded capabilities, but it depends on the SOC vendor whether a driver is available or not
I found that it is not possible to use OpenCL. I'd have to rewrite my algorithm to OpenGL and use shaders and vertexes - then I can gain "General Purpose" programming (welcome back to past, about 4-5 years back to be more exact).
Take a look at the following thread elaborates on what is possible and not possible to do up to date (14th of Nov,2010):
link text
I've seen this example from the folks from Nokia:
http://www.hotchips.org/archives/hc21/1_sun/HC21.23.2.OpenCLTutorial-Epub/HC21.23.270.Pulli-OpenCL-in-Handheld-Devices.pdf
So I ask myself, is there any SDK from any mobile platform/os out there that I could use to test some of my desktop apps to an embedded app? I'd really, really appreciate to be able to program opencl on mobile/tablets systems. Vertex/fragment shader are not much of help because their specs for embedded systems do not include all the extensions we would need to rewrite our opencl code to shader.
Related
Can we use openCL on consoles like Xbox One and PS4 for General purpose GPU? If yes, can we use openCL framework like ArrayFire - http://arrayfire.com/ ?
No.
While the AMD GPU hardware on the most recent version of each console is similar enough to desktop hardware that has OpenCL support, the console vendors do not offer OpenCL as a programming API (unless it's not public information and only available under NDA). If enough game devs asked for it perhaps it would happen.
I was also looking for it too! And I've found this discution: here
Apparently, up to now, the only way out is to use HLSL for Xbox, and PSSL on Playstation. According to a guy on the topic above. They (the APIs) are different, but very similar in such a way that shall be possible to write a code that compiles under both platform using some preprocessors.
PS.: I'm also looking for a good answer for that question. So, if you've found anything around, please post here, I'd love to know ^^
In my last question I asked whether there was a better way to rotate images than I had thought of. I ended up discovering jpegtran and have since found libjpeg-turbo.
Now I am looking for a better way to resize the images (jpegs) than imagemagick and graphicsmagick.
Is there a specialized commandline tool to resize the images in a more efficient way than imagemagick or graphicsmagick? Maybe the resizing can be done on the GPU using opencl or opengl?
The provided hardware is the same as in the other post:
Intel Atom D525 (1,8 Ghz)
Mobility Radeon HD 5430 Series
4 GB of RAM
SSD Vertility 3
Check this link out: http://leocharre.com/articles/faster-image-resizing-in-linux/
In particular the author mentions that imgresize is faster than imagemagick, and epeg is extremely fast.
epeg (http://www.systhread.net/texts/200507epeg1.php) seems quite well documented for generating thumbnails. If the quality is good enough, this could be the solution.
OpenCL is a standard for cross-platform, parallel programming of modern processors found in personal computers, servers and handheld/embedded devices. It's directly supported by ATI. You'll need to get AMD APP SDK (formerly known as AMD Stream SDK) to get GPU support (also check out this getting started guide).
Take a look at Intel's IPP - Integrated Performance Primitives. It's a multi-threaded software library of functions for multimedia and data processing applications. Among other features, it's has functions to resize images (bilinear, nearest neighbor, etc). Unfortunately, it is not free (cheapest version costs $199).
VIPS is a free image processing system. It claims that compared to most image processing libraries, VIPS needs little memory and runs quickly, especially on machines with more than one CPU. See the Speed and Memory Use page for a simple benchmark against other similar systems.
You can actually do a lot of bulk processing like this with GIMP's CLI options.
http://www.gimp.org/tutorials/Basic_Batch/
There is also djpeg and cjpeg from the Independent JPEG Group which can rescale and image to an M/N fraction. Not perfect but very fast.
Simply use FFMpeg.exe. It can resize , convert , change quality and so on.
And also it works with almost all known types of videos/audios/pictures.
It works in linux/unix too, and there is open source code for it written in C++.
You can get it Here (for Windows/compiled exe) or Here (source code and so on).
If you are developing a program, I recomend you to use standard GDIPlus library.
It does everything with pictures.
I'm looking for comparisons between OpenCL and DirectCompute, but I haven't found anything. OpenCL's advantages of being cross-platform and having a wider range of supported GPUs don't matter to me. I'm fine with coding on Windows against DX11 GPUs only. Assuming that, what are the pros and cons of each API?
I know this question was raised before, but I'm looking for more details.
I'm not interested in CUDA, since I don't want to restrict myself to only Nvidia hardware.
Probably the biggest difference for a coder is that DirectCompute is programmed by a language which is similar to HLSL, and OpenCL is programmed via a C-like language.
Another difference to consider is that, generally, for commodity level GPUs, the DirectX support is better (faster and less buggy) than OpenGL support on Windows. This may translate to more stable support for DirectCompute, but really, this is just speculation.
Well the major advantage of OpenCL is that it is not just limited to graphics cards. You can make use of your multicore CPU, Graphics Card and potentially any number of other hardware acceleration devices (DSPs etc) all from the same program.
I'm not sure if DirectCompute allows that freedom.
The OpenCL cross-platform-ness is not just a detail, as the host code (the one calling the OpenCL API and submitting kernels) can itself be cross-platform (see link text, link text...).
Write once, run on any GPGPU, anywhere.
Otherwise the OpenCL tooling is really getting better, with an ATI Stream plugin for Visual Studio, the NVidia & ATI SDKs that contains tons of samples, etc...
Another option now is C++ AMP which gives you modern C++ syntax without a need for a seperate compiler while still preserving hardware portability. Please follow links from here for more info and feel free to post questions as you have them: http://blogs.msdn.com/b/nativeconcurrency/archive/2011/09/13/c-amp-in-a-nutshell.aspx
I use OpenCL because i can easily port my App to Linux but with DirectCompute this is not possible.
I think also that the performance of the OpenCL implementation will increase with time (that it comes at the same Level like CUDA for NVidia Cards) and also that the (driver)bugs will (hopefully ;) ) be eliminated with time.
I would like to solder a microcontroller, control buttons and an DVI/HDMI output and program this in a way, that I can store images on it and let them display as a dia-show via the outputs.
It doesn't have to have a lot of storage capacity, 128Mb would be enough.
but I don't know how to start, because I haven't done anything like this before.
My aim is to present some important images to friends by just taking this hardware, connecting it to a TV screen and showing these photos. If should be able to switch the photos manually (using a button) or automatically in a dia-show.
It should support several TV resolutions and it should be connectable to my PC (USB prefered), so that I can upload and delete photos.
So where to start and how to do that?
Thank you in advance, Andreas
If your aim is just to show some photos, there are assuredly simpler and more cost effective ways to do so; devices exist which do more or less exactly what you are proposing.
If your aim is to learn about microcontrollers and this is a project your are taking up to further that, I would recommend looking into the Arduino: http://www.arduino.cc/ or a similar kit based micro, and growing your project from that.
Microcontroller + low level language will be a huge pain to work with, particularly if you wish to handle various file formats and screen resolutions. Get a full-blown computer with an OS instead - something like http://en.wikipedia.org/wiki/PC/104
If your goal is purely to be able to display photos then I would recommend using a digital camera with video out capabilities.
If your aim is to learn about electronics and microcontrollers I would start with a good book and an Arduino board. Note that writing microcontroller code to handle file systems, image formats and video output is non-trivial. Simpler projects may be a better starting point as they are more accessible resulting in quicker progress, less frustration and more motivation!
The engineering field is a interesting field. You can start with the web site "www.microchip.com". You will need a high end device consider the PIC32MX795L512, there is a nice starter kit for it, "Ethernet Starter Kit" http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=2615&dDocName=en545713. This kit has the on board debugger & programmer to do all the hard work.
You get sample projects with the package, you can program using ansi c programming.
IDE : MPLAB which is free, and the C32 compiler has a student/lite version.
Arduino also has a board with the same device.
I personally like "www.techtoys.com.hk", they have device compatible with Microchip boards like techtoys.com.hk/PIC_boards/PIC32STK%20SSD1963%20EVK/PIC32STK%20SSD1963%20EVK%20R1A.htm, or this techtoys.com.hk/PIC_boards/PIC2432EVK-RD4/PIC2432%20EVK%20RD4.htm where this board you will need a debugger/programmer like the low cost PIC Kit 3 "microchip.com/pickit3".
The trouble is you need to write the HDMI video library yourself, there are some VGA libraries available but they are only black and white and very hard to get color with these analog images. The rest of the libraries are already there, USB MSD(flash drive), SD Card, pictures (jpg) etc.
microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=2680&dDocName=en547784
Feel free to contact me if you need some help, I might be able to help with the HDMI library.
It's a lot of fun to play with these toys.
Regards
Lucas
B-Eng Digital Engineering.
imlucanio#yahoo.com (no spamming)
Remember to add the http and www to the web links.
It sounds like you want an iPod. That is a dead simple thing to work with and it does everything you want. Otherwise, very complicated. I'd suggest the BeagleBoard and embedded Linux. Yes, it warrants that level of complexity.
The options for small microcontrollers just aren't there. The Arduino is very popular and yes, you can interface an SD card to it. That'd be your storage. Yes, you can put a digital potentiometer on it, that can be your interface. I've seen some video overlays that do simple text, but never any JPEG display (too much processing required). And certainly no 24 bit color (so that the output would actually look good) - that would take WAY too many pins to do correctly (and the Arduino doesn't have a D/A converter! You'd have to rig something up that would suck). And even then, all of the options for TV out weren't HDMI, but RCA (the old red/white/yellow cables).
So in short, no. Get a computer. That's what can do the job.
Besides Arduino, what other ways are there to learn hardware programming in a hands-on way? Are there any nifty kits available, either a pre-assembled robot, that you can program to move a certain way, or do certain things, or anything similar to that?
Atmel AVR and the PIC both have experiment boards that you can use solder stuff on to, usually they have a couple of buttons and some lights pre-soldered to the area. This let's you program/flash the microprocessor and play with the output pins. You can either write the programs in assembly or C.
Parallax have a number of kits. They have two product lines suited for "playing around", Basic Stamp and something called Propeller. The former is a small microprocessor that runs programs written in Basic (a tad disgusting ;)) and the latter runs something called Spin or assembly (well after compilation obviously.)
I would go with either AVR or the PIC. I've done PIC but I've heard good things about AVR, they seem to ship with better software.
At first look Microsoft's VPL sounds good, but when it comes to actually LEARNING how hardware works it goes a LONG way to hide those details from you. As a matter of fact it is pretty much designed for people who don't program, and is distastful to someone who's actually written embedded software. IF you just want to make stuff happen and not delve into the details it's fine, but if you want to get down to the metal like programming the "Arduino" boards it's not for you.
If you're used to something like the Arduino then something like the PIC will be an easy transistion. SparcFun Electronics has all sorts of DIY type projects and hardware available. If you have a decent bookstore around your area, I would suggest looking for "Circuit Cellar" magazine. It has articles on a monthly basis with project for someone looking to get into hardware projects, everything from homebrew Software Defined Radio to FPGA based 3D graphics. (Raytracing actually) Usually the authors describe the project in an article and "WHY" they made the decisions they did, a description and schematics of the hardware and provide a link to source code.
Cypress Semiconductor has one of the most interesting embedded processors on the market and several high quality dev boards for sale. The PSoC includes the ability to not only configure the software, but also to "drop in" software configured hardware such Analog to digital converters, serial I/O, Digital to Analog and Various amps and filters. It's a REALLY cool concept, and the "touch sensor" capability of the PSoC were actually used in several models of the IPod.
One thing about programming these little micros is they don't have a lot between you and the hardware, you get to see how things really work. It doesn't matter whether you're talking about an 8-bit microcontroller or a quad-core Pentium programming hardware is largely the same concept. You write to a memory mapped register for some piece of hardware like a serial controller, and the hardware responds in someway. If you program a baudrate generator in a PIC or PC it's largely the same idea, you write a value that will be used as a division factor from a given clock to achive a given baudrate. The numbers and names maybe different, but the concepts is the same. On a PC you may have to map to the PCI address of the card, which adds a some complications, but if you looked underneath the OS you would see that that was done just by writing values to registers simalar to programming a PIC to use a different "Page" of memory. Is it worth learning an 8-bitter? Well, there are approximately $5 billion dollars in sales of the little 8-bit micros today with projection only showing growth in that market in the future. I saw one reference that state the average car has 25 Microcontrollers in it. That's not too bad.
I haven't played with it much, but the iRobot looks pretty cool.
The ability to simulate how your robot will work which some of the other answers mentioned is nice, but there's nothing like seeing a real-life robot do what you programmed it to do. That, to me, is what really makes robots fun and cool.
There's the .NET Micro Framework.
It's incredibly simple to use/setup and there's lots of hardware being made to target this framework.
You should take a look at Microsoft Robotics Developer Studio which supports many different kits.
I have always been curious about gumstix. It seems more professional than arduino, and it aims at the Linux programmer. I cannot give you a real suggestion, as I've never played with it, but I would definitely go with one of this toys if I had to do and learn some cool hardware programming.