At the moment, few people care about ATI cards for anything else than video games. This may change in the next few years as better ATI graphics card for HPC are introduced.
Also balancing CPU and GPU usage is even harder ( maybe impossible ? ) as you cannot predict
Computational code is quite regular, so it is usually possible to predict its performance, especially on GPUs which are quite simple architectures (no cache or reordering).
This is a GTK+ application. It only looks as good as your GTK+ theme and font setup. The one in the screenshots of the official website indeed looks a bit ugly.
Unstable, I'd understand. It's there exactly for the purpose of uploading libraries and hold them for the transition period before everything goes at once in testing.
I am assuming that the porting developers are using source code management tools, following a standard, and are adhering to a common design philosophy (coding for exceptions, using a consistent naming standard, comments that address the what of the code (not the how), and indenting code for readability)
In large code bases, it can be rare that everything follows the same naming conventions, indenting style, or even programming style, simply because hundreds of people work on the code, and different teams with different skills edit different parts of it. The only unit where this tends to be true is the file.
As a software developer I've experienced this most of the time, working with code bases much larger than this without any sort of documentation. It's not unusual for single files to be more than 4,000 lines of code and to have thousands of files, with the occasional crazy branching 10 levels deep within a single function.
It's just something you have to deal with as a professional software developer. It's not really a problem. Knowing how to use your tools efficiently comes a long way. I personally use grep and find a lot.
Yes, that's why every industrial and medical CT system comes with GPU reconstruction routines unlike 5 years ago.
I didn't say GPUs were not faster, I said they were not as much faster as people claimed. The OP said that his code went from taking days to taking minutes. That's an acceleration of the order of several thousand times. A GPU is simply not that much faster than a CPU. If it was made so much faster, it's simply that it was rewritten by competent people that knew how to make it fast, while the original CPU version was not.
your ignorant post
alas, you have no clue or experience about this topic
Just so you know, I am the CEO of a company that edits compilers and libraries for parallel computing. We mostly work in two industries : image/video/multimedia/computer vision (a bit of medical imaging too) and banking/insurance/financial. Our people are seasoned computer architecture experts, many of which also sport a phd in various fields, including mathematics, robotics, and computer science. We have strong partnerships not only with NVIDIA, but also with AMD and Intel, which give us future products for evaluation. I myself contribute to the evolution of parallel programming in C++ as an HPC expert at the C++ standards committee. If you feel like you'd want to apply for some consulting to have us help you improve the performance of your filtered backprojection -- I myself have no knowledge of that field, but I assume it's similar to tomography for which we have good results already deployed in the industry --, I'm sure our team would be delighted to help you.
A gpu is at most 20 times faster than a cpu while costing 4 times as much. If your code is so much faster on gpu, it's just because your cpu version was crap and not optimized.
People have different tastes. I personally find that it has the best controller ever made, and that it was full of very good games. All games that were on multiple consoles were better on Dreamcast too: more fluid etc. (Quake III being a prime example)
Entrepreneurs are a tiny fraction of the population, successful ones even less so, and successful ones with hundreds of millions even more. Are you suggesting we design an education system that would only be useful to the lucky 1% of the population?
Slashdot Mirror
You, sir, are confusing normal cars with sports cars.
Only if by "modeled fundamentally differently" you really mean "not modeled at all".
At the moment, few people care about ATI cards for anything else than video games.
This may change in the next few years as better ATI graphics card for HPC are introduced.
The funny thing is that this is still slower than a good CPU matrix multiply.
GCC is GPL too, yet many commercial products use it.
You usually only need to use a compiler for development, licenses like GPL only need to apply once you ship it.
Computational code is quite regular, so it is usually possible to predict its performance, especially on GPUs which are quite simple architectures (no cache or reordering).
Given that the universe is not deterministic, it is pretty hard to justify free will does not exist.
This is a GTK+ application. It only looks as good as your GTK+ theme and font setup. The one in the screenshots of the official website indeed looks a bit ugly.
Try a decent text editor some day, such as geany.
A pregnant woman is less likely to have fun with an assault rifle
I had heard of it as Tsar Bomba, and didn't know Big Ivan was another name for it. My bad.
I'm not a bomb geek but even I know that Big Ivan is not the largest bomb ever made.
You could also try not to overcook your meat, and eat it very rare like it's meant to be eaten.
Isn't that what 'experimental' is for?
That would be debian stable (squeeze), not debian testing (wheezy).
Debian testing has packages which are much more up to date than ubuntu's.
You may also choose to use Debian unstable (sid).
Debian testing (i.e. Wheezy) already gives all the advantages outlined above.
In large code bases, it can be rare that everything follows the same naming conventions, indenting style, or even programming style, simply because hundreds of people work on the code, and different teams with different skills edit different parts of it.
The only unit where this tends to be true is the file.
I'd have to agree on this.
As a software developer I've experienced this most of the time, working with code bases much larger than this without any sort of documentation. It's not unusual for single files to be more than 4,000 lines of code and to have thousands of files, with the occasional crazy branching 10 levels deep within a single function.
It's just something you have to deal with as a professional software developer. It's not really a problem. Knowing how to use your tools efficiently comes a long way. I personally use grep and find a lot.
This is a very insightful post, too bad it is not rated higher.
I didn't say GPUs were not faster, I said they were not as much faster as people claimed.
The OP said that his code went from taking days to taking minutes. That's an acceleration of the order of several thousand times. A GPU is simply not that much faster than a CPU. If it was made so much faster, it's simply that it was rewritten by competent people that knew how to make it fast, while the original CPU version was not.
Just so you know, I am the CEO of a company that edits compilers and libraries for parallel computing. We mostly work in two industries : image/video/multimedia/computer vision (a bit of medical imaging too) and banking/insurance/financial. Our people are seasoned computer architecture experts, many of which also sport a phd in various fields, including mathematics, robotics, and computer science. We have strong partnerships not only with NVIDIA, but also with AMD and Intel, which give us future products for evaluation. I myself contribute to the evolution of parallel programming in C++ as an HPC expert at the C++ standards committee.
If you feel like you'd want to apply for some consulting to have us help you improve the performance of your filtered backprojection -- I myself have no knowledge of that field, but I assume it's similar to tomography for which we have good results already deployed in the industry --, I'm sure our team would be delighted to help you.
A gpu is at most 20 times faster than a cpu while costing 4 times as much. If your code is so much faster on gpu, it's just because your cpu version was crap and not optimized.
There is better than cufft, especially when multi gpu is involved
People have different tastes. I personally find that it has the best controller ever made, and that it was full of very good games.
All games that were on multiple consoles were better on Dreamcast too: more fluid etc. (Quake III being a prime example)
Entrepreneurs are a tiny fraction of the population, successful ones even less so, and successful ones with hundreds of millions even more.
Are you suggesting we design an education system that would only be useful to the lucky 1% of the population?
Because you can't put things in the public domain. Works fall in the public domain once copyright expires, but you cannot force copyright to expire.