http://www.mapleprimes.com/posts/36412-GPU-Programming-CUDA-OpenCL-And-Maple en-us 2026 Maplesoft, A Division of Waterloo Maple Inc. Maplesoft Document System Wed, 10 Jun 2026 06:58:25 GMT Wed, 10 Jun 2026 06:58:25 GMT The latest comments added to the Post, GPU Programming, CUDA, OpenCL and Maple http://www.mapleprimes.com/images/mapleprimeswhite.jpg http://www.mapleprimes.com/posts/36412-GPU-Programming-CUDA-OpenCL-And-Maple http://www.mapleprimes.com/posts/36412-GPU-Programming-CUDA-OpenCL-And-Maple?ref=Feed:MaplePrimes:GPU Programming, CUDA, OpenCL and Maple:Comments#comment62073 <p>Since I am planning on actually using this soon on my new machine I would love a windows version.</p> The latest comments added to the Post, GPU Programming, CUDA, OpenCL and Maple 62073 Sat, 07 Nov 2009 06:30:44 Z marvin marvin http://www.mapleprimes.com/posts/36412-GPU-Programming-CUDA-OpenCL-And-Maple?ref=Feed:MaplePrimes:GPU Programming, CUDA, OpenCL and Maple:Comments#comment62074 <p>My real need for CUDA alongside Maple is to do double precision matrix manipulations.&nbsp; What I want to do is speed up the construction of large matrices of type float[8].&nbsp; For that I need to know how you get CUDA to do double-precision and how to interface Maple to that sort of thing.&nbsp; I would love a Windows example that shows me how to do something like</p> <p>1.&nbsp; foo:=Matrix(1700,1700,(i,j)-&gt;myfunc(i,j,stuff....),datatype=float[8], shape=hermitian (or symmetric))</p> <p>where myfunc is an external routine that I have compiled.</p> <p>OR</p> <p>2. M:=Matrix(1700,1700,datatype=float[8]);<br /> &nbsp;&nbsp;&nbsp; foo:=MakeMatrix(args,M);</p> <p>where MakeMatrix fills the matrix M using an external CUDA based function MakeMatrix().</p> <p>It would be really helpful to know how to implement something real world like this.</p> <p>&nbsp;</p> <p>Best,</p> <p>Marvin</p> The latest comments added to the Post, GPU Programming, CUDA, OpenCL and Maple 62074 Tue, 10 Nov 2009 00:56:24 Z marvin marvin http://www.mapleprimes.com/posts/36412-GPU-Programming-CUDA-OpenCL-And-Maple?ref=Feed:MaplePrimes:GPU Programming, CUDA, OpenCL and Maple:Comments#comment62075 <p>I think if you take a look at the Mandelbrot example code, it is very similar to how you would implement your #2.&nbsp; The Mandelbrot code accepts a Maple Matrix and fills it in.&nbsp; If you matrix is triangular, I would suggest simply modifying the if statement that checks that the indicies are in bounds.</p> <p>There will be some differences for windows, however it should not be too hard to for you to figure those out.&nbsp;</p> <p>As for double precision, only the most recent CUDA hardware (compute level 1.3 and higher) supports double precision, and it is slower that single precision, (although still faster than doing it on the host).</p> <p>Darin</p> <p>-- Kernel Developer Maplesoft</p> The latest comments added to the Post, GPU Programming, CUDA, OpenCL and Maple 62075 Tue, 10 Nov 2009 01:19:49 Z dohashi dohashi http://www.mapleprimes.com/posts/36412-GPU-Programming-CUDA-OpenCL-And-Maple?ref=Feed:MaplePrimes:GPU Programming, CUDA, OpenCL and Maple:Comments#comment62076 <p>Actually I was talking to a VP of NVidia on Monday about the hit (which is a factor of 8 for my card).&nbsp; He said that it is all a question of how they apportion the real estate on the card as all processors are already 64-bit.&nbsp; Furthermore he said that the new Fermi board will only take a hit of a factor of 2.&nbsp; Now that is exciting.&nbsp; So, while my new home machine is already a done deal, I am clearly going to wait for my work machine.&nbsp; A factor of 4 is nothing to sneeze at ;^)</p> <p>&nbsp;</p> <p>By the way, what do you think of go? (google's new language)</p> The latest comments added to the Post, GPU Programming, CUDA, OpenCL and Maple 62076 Fri, 13 Nov 2009 05:27:16 Z marvin marvin