Why so little parallelism?
|
From: Andrew Reilly on 2 Nov 2006 19:01 On Thu, 02 Nov 2006 13:11:15 -0700, Eugene Miya wrote: > In article <4qtfa0Foj3v6U1(a)individual.net>, > Andrew Reilly <andrew-newspost(a)areilly.bpc-users.org> wrote: >>How are the developers, however smart, going to express their algorithms >>without introducing sequential dependencies, however inadvertently? >>(What is an algorithm, without sequential dependencies?) Don't you need an >>appropriate language, and perhaps a plausible parallel machine >>abstraction, before you start on the compiler? How would your language be >>different from Verilog or VHDL or Occam? What would be different, this >>time? > > Take a look and try what friends did on SISAL. That's nice and all, but what's stopping it from being used instead of MPI or OpenMP on the big, parallel iron? Just inertia? SISAL is twenty years old, now. [There does appear to be an active SISAL implementation on sourceforge, though. Reckon it's worth having a look at.] >>Or are you, perhaps, hinting at the "High Productivity" DARPA project, or >>one of Sun, IBM or Cray's sub-projects, each of which, I assume, has >>working answers to my previous questions? > > They mostly work on hardware. Cray Chapel; IBM X10; Sun Fortress: http://crd.lbl.gov/~parry/hpcs_resources.html Opinions? Cheers, -- Andrew
From: BDH on 2 Nov 2006 19:25 > Feynman was a "wizard of the highest order". Yes. > > Oh to have considering going out with his goddauther. > > Please tell! Yes. You know, there are more people nowadays, and by definition more people with whatever IQ. But the amount of genius around seems constant at best.
From: BDH on 2 Nov 2006 19:42 > Hmmm, no. > I've worked on several benchmark teams and have to sweat various issues. > If you want brutal, non-trivial slow problems outside cryptanalysis, > I can think of many. A favorite might be Golomb rulers. > No floating point. Can fit in memory. Golomb rulers are already found with massively parallel computation. I guess I'm not sure what you mean. > On the contrary what von Neumann wrot was a very good simple system for > his time. Fine, it's unfair to blame him like that. Some people definitely have a higher opinion of him than I do though. > I recall the special APL character set > (the Culler-Fried keyboard was little better and its was QWERTY). > I extracted from an officemate involved with the CDC Star-100 that the > direction of evaluation was a poor choice. But its influence died off too. Why? > Part of the problem which kills simple languages like these are the > kinds of numeric applications which have volumetric and border (edges, > faces, vertices, and hyper-structures (4D and higher Ds) which are > exception cases which then go to irregular geometries, etc. That can be turned into arrays too.
From: Eugene Miya on 2 Nov 2006 19:00 >> Feynman was a "wizard of the highest order". In article <1162513538.821962.326490(a)h54g2000cwb.googlegroups.com>, BDH <bhauth(a)gmail.com> wrote: >Yes. I have to agree a bit with what Gell-Mann wrote about him creating an air/eire about him. Not that I ever met him more than a couple of times. I spent enough time in Laurtisen to pass by his secretary's office occasionally and also see one of his series of photos lecturing. >> > Oh to have considering going out with his goddauther. >> Please tell! >Yes. I just thought this one woman I went out with was one of the finest women I have ever had the chance to know. Turned out her dad was a Caltech physics prof and colleague of Feynman's who I first met about 4 years or so earlier at the time of his Los Alamos lecture at UCSB. Who would have thought that I'd take a job at JPL, then live next door to Tech, then work for Tech for the Lab (well one is part of the other). Anyways this CD is available in the revised book Classic Feynman (I can hear Rik Smoody ask the question about safes at the end (Rik went to Tech for a year or 2 before burn out)). I chanced to meet her in a PCC intro geology class before grad school. My friends at the Lab remarked "Where did you find her?" at the Section hot tub party we used to hold (one of my then superviser is now JPL's Director, go figure [an amazing Lebanese guy]). So the lesson is do things outside computing. >You know, there are more people nowadays, and by definition more people >with whatever IQ. But the amount of genius around seems constant at >best. Life is far more than intelligence. It doesn't tell you about the physical and emotional problems (complexities) people have especially women. I'm not going to get into her problems and her life. The US is very fortune to have Caltech. It's intensity is like no other campus in the US that I have visited (it's the basis for the film Real Genius by Martha Cooliege (sp)). I include MIT and Stanford (differ qualities: I'm amazed that I know Hennessy and go that one page for his book: go buy 4th ed.). But Caltech has never had a great CS Dept. in part because it's an institution (It's The Institute) of physicists (I mean Carver well, and it's where the precursor to Mathematica got its start, and the home of hypercubes, and I'd get asked questions about CS by my friends who run the place (one friend was #2 man there)). Oh and Connection Machines had their start where when Danny move from there to MIT (I am pretty sure that Danny was in a von Neumann bottleneck lecture given by Ivan Sutherland). --
From: Eugene Miya on 2 Nov 2006 19:11
In article <ch2q14-1d1.ln1(a)osl016lin.hda.hydro.com>, Terje Mathisen <terje.mathisen(a)hda.hydro.com> wrote: >Feynman was a "wizard of the highest order". Even after he had figured >something out and explained it, there was no way for mere mortals to see >how they could ever have done the same. Oh that was you who said that. Hmmmm what can I say? Caltech collects geniuses like Feynman. He was clearly unique, but so are and were a fair number of other people who went to, worked at, or were part of Tech. I could never have gotten in. I am not smart enough. They may have also passed their their prime (friends hope not). I took a class from a JPL co-worker (hard to use that word as he was so much more advanced that I was: Blinn). Most of you will have never heard Jim because this isn't comp.graphics. He had a hobby of collecting Pythagorean theorem proofs (he's got over 100). He's also a master of 4x4 matrix multiplication. That says nothing of non-computational fields. You should visit Tech for half a day the next time you come to California Terje. -- |