Guest Blooz1 Posted December 17, 2007 Posted December 17, 2007 It used to be a "given" that programs lagged about 2 years behind hardware advances. Is that gap widening instead of shortening? Article in today's Times:Faster Chips Are Leaving Programmers In Their Dust So, what do ya think? Quote Share this post Link to post Share on other sites More sharing options...
Praz Posted December 17, 2007 Posted December 17, 2007 Most all consumer software utilize only two cores directly, even if advertised as being optimized for quadcore. The work the other two cores do is the result of the operating system. The majority of developers agree that threading for 3 cores is pretty much the limit today. Quote Share this post Link to post Share on other sites More sharing options...
Angry_Games Posted December 17, 2007 Posted December 17, 2007 well, mainstream software programmers anyway scientific, military, R&D, engineering, etc...those specialized softwares will take advantage of as many cores as you can throw at it...there's no limit to massively parallel processing programs. Problem with that is those programs are very centric to a specific use. Unlike OS's that we are used to... Developers might cry that multiple cores are limiting, and they might whine that 3 cores are the limit, but that's a cop-out, and they know it just as well as I know it. Evolution is a necessary step. Instead of crying like girls they need to get off their asses and learn to program for it properly, even if that means moving away from a piece of . OS like Microsoft (any flavor, take your pick) and move to a superior system for such advanced programming like Linux. Of course, there's no money in linux compared to Windows =/ yay! I love being handicapped! Quote Share this post Link to post Share on other sites More sharing options...
Guest Blooz1 Posted December 17, 2007 Posted December 17, 2007 I thought it was kinda sad too. You'd think a company as huge as Microsoft would take their bestest, fattest, smelliest, hairiest, whacked-out programmers whom all the other people HATE because they're so much better and faster and throw 'em in a building out in the woods by themselves. Give 'em an idea, all the Twinkies/Ho Ho's they can eat, Jolt Cola they can drink, and a problem like this! Tell 'em "Hey, we only want to see you 4 times a year - find a solution to this idea" and walk away! Call it the "Skunkworks" or something like that... Quote Share this post Link to post Share on other sites More sharing options...
Angry_Games Posted December 17, 2007 Posted December 17, 2007 Microsoft stifles innovation, it doesn't create it. That's the reason why MS still sucks (what? a competitor's program does something better than our piece of . add-in that comes with windows?? well well well, we are just going to BUY that company and take their technology instead of inventing our own competing technology!) Quote Share this post Link to post Share on other sites More sharing options...
MACarter02 Posted December 17, 2007 Posted December 17, 2007 those specialized softwares will take advantage of as many cores as you can throw at it Matlab, for example, allows you to choose the number of cores it uses for computations On a side note, in the NYTimes website, double-clicking a word brings up its definition. Nice. Quote Share this post Link to post Share on other sites More sharing options...
Praz Posted December 17, 2007 Posted December 17, 2007 Matlab, for example, allows you to choose the number of cores it uses for computations Only for matrix computation and then the matrix has to be large before there is a benefit. Very few multi-core programs truly utilize the cores in an efficient and productive manner. The tools just don't exist to make it feasible from a cost standpoint. From Cleve Moler One way MATLAB does this is with implicit multithreading for matrix computation. But, according to Moler, the matrices have to be rather large before the multicore speedup is really effective. The product also uses multithreaded libraries that exploit multicore, such as ATLAS BLAS, and the Intel and AMD math libraries for matrix computations. But at this point, the matrix math components are just a portion of the entire product, Moler said. "The question is how to do something more complicated that just matrices; and that's difficult," he admitted. Internally the engineers discussed adding multithreading constructs into the product. Some prototypes were built. But eventually they decided it was too complicated and didn't really work well with the MATLAB model. Ultimately, they just weren't convinced that explicit multithreading was the right way to go. http://www.hpcwire.com/hpc/1894516.html Quote Share this post Link to post Share on other sites More sharing options...
Angry_Games Posted December 17, 2007 Posted December 17, 2007 pssst praz kill that newegg banner in your sig...it's the one that doesn't report to our affiliate! (use the new banner we have up on the top) Quote Share this post Link to post Share on other sites More sharing options...
MACarter02 Posted December 17, 2007 Posted December 17, 2007 What does 'rather large' mean? It is very common for scripts that I write to run matrices with sizes around 10,000 through several functions. Is that 'rather large'? Regardless, it "takes advantage of however many cores you throw at it" as angry said. Maybe not completely, but somewhat. Quote Share this post Link to post Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.