Author: Vincent Diepeveen
Date: 05:19:09 04/16/01
Go up one level in this thread
On April 16, 2001 at 07:25:52, Mogens Larsen wrote: >On April 16, 2001 at 07:18:19, Vincent Diepeveen wrote: > >>Diep can run at n processors, where N is any number of >>nodes or processors (as long as there is a high bandwidth network, >>so a 100mbit network is not exactly what i mean :)). >> >>Working now at n x i processors even. Where i is the number of processors >>used within a node and n for the number of nodes. >> >>Note that SOS and Amy are also multiprocessor capable. SOS already ran >>in februari 1999 at 4 processors, long before fritz and junior ran >>parallel :) > >Thank you for the info, I had forgotten about SOS. This certainly proves that it >isn't impossible to implement SMP. A specific encouragement wouldn't exactly >hamper development either. Well Rudolf could get anything to work long before the rest of the world can. In fact many programs are parallel now BECAUSE of Rudolf's effort as he also explained in and around germany how he did it. >Mogens. It is not difficult to implement the form of parallellism as used by Rudolf. Invented by a frenchman who couldn't spell a word english and who wrote an impossible article for JICCA (did anyone proofread it at the time as i'm pretty sure they didn't get his parallel idea?). At the time when i read the article i was pathetically laughing about it actually as i also didn't get the idea of the frenchman. But it appears everyone who can make a chessprogram work under win2000 can also get within an afternoon his program parallel to work. Then some debugging and a day later it works cool. Yet i heart rumours that the K7 dual motherboard is going to be like at least us$ 1000, so that's not encouraging either when it comes out around oktober 2001 or so (if it gets out anyway as it is getting announced every month now), so there is not going to be much competition. Fact is that one sells very little programs because it is a parallel version, so hesitation to get parallel is understandable. Also nearly all home tests, i just have to point to SSDF for example, are done on single cpu computers. Whereas duals are nowadays very affordable when they carry intel logo, still 4 processors are a bit more expensive. Of course the entire design of quad motherboards is expensive, but each processor costs like $2000 or something, for a step in model that is. If you need 4 processors of each $2000 then that definitely is above my current budget as $8000 for something which is slower as soon as that dual K7 motherboard gets out, that's really hard to affort. Also AMD is not really giving much of a hope for a quad k7 motherboard. Where on paper K7 is capable of running 8 processors at the same time, it still is very unclear how they are going to manage that. As the quad motherboard under development right now seems, if i read it well and we still need to see whether this report was a very accurate one, it has no shared memory at all like the quad xeons have! The shared memory is very very slow memory. Way slower as normal RAM (like sdram or edoram) but it is in fact not a quad but a 2 node dual with some slow memory added. So not really encouraging...
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.