Author: stuart taylor
Date: 15:51:30 05/18/05
Go up one level in this thread
On May 18, 2005 at 16:24:15, Robert Hyatt wrote: >On May 18, 2005 at 15:23:48, stuart taylor wrote: > >>On May 18, 2005 at 11:38:48, Robert Hyatt wrote: >> >>>On May 18, 2005 at 06:31:23, stuart taylor wrote: >>> >>>>On May 18, 2005 at 00:04:32, Robert Hyatt wrote: >>>> >>>>>On May 17, 2005 at 20:14:56, Ryan B. wrote: >>>>> >>>>>>On May 17, 2005 at 19:58:35, stuart taylor wrote: >>>>>> >>>>>>> 1). What exactly is the implications of RPM in hard disks? >>>>>>>a). in comparing 42000 rpm to 72000 rpm, >>>>>>In chess none >>>>> >>>>>Makes a +huge+ difference if you use endgame tables. >>>>> >>>>>From someone that does this. :) >>>> >>>>But would the difference between 4200 rpm and 7200 rpm be huge? >>>> >>>>S.Taylor >>>>> >>> >>>4200 rpm is 14 ms per revolution, or 7ms average rotational latency. 7200 rpm >>>is 8ms per revolution, or 4ms average rotational latency. Every random disk I/O >>>will save 3ms with the 7200 RPM drive. It will be very noticable for endgame >>>probes. >> >>Thanks Bob, for this further clarification! >> I never really learned the basics, although I ought to have done by now. I just >>try to fit things together in my mind like a jigsaw puzzle. I'm sure it wouldn't >>take more than 5-6 hours to get a very useful amount of basic knoledge in my >>head, if someone was explaining to me extremely clearly. >> >>However, I never knew that endgame bases take much time to access anyway, so I'm >>not sure how the lack of being less than a half slower would be that bad. 4200 >>as opposed to 7200. >>I have been aware that when things reach the advanced endgame, some programs >>move instantly, every time. So that seemed quite quick, even with some of the >>programs (and hardware) from over 15 years ago. > >Think about this. How much stronger is your engine when its NPS is 2x faster? >Significantly stronger? That is what you are talking about with the endgame >database probes. I have seen positions where I do thousands of probes a second. > If you double the speed, you double the search tree size as well, which >produces better play. Or the same play, if you give it twice as much time? S.T. > > > >>S.Taylor >>> >>> >>> >>> >>>>> >>>>> >>>>>> >>>>>>> >>>>>>>b).what difference does this make to speed of computer functions, and which >>>>>>>computer functions does it affect? >>>>>>> >>>>>>will affect read/write speed. For chess this should not mater. >>>>> >>>>>RPM doesn't really affect bandwidth. In fact, the fastest spinners actually >>>>>have lower sustained bandwidth because clocking the bits becomes a problem. As >>>>>a secondary issue, the fastest disks are not the best capacity drives, for the >>>>>same reason. Best SCSI drives for capacity today are the 7200 and under >>>>>drives... >>>>> >>>>> >>>>> >>>>> >>>>>> >>>>>>>c).Does the new 915 chipset make the rpm slower? (I read that they make >>>>>>>something a little slower, but I couldn't understand exactly what, and if it >>>>>>>causes any handicap). >>>>>>> >>>>>>I really doubt it. >>>>>> >>>>>>>d).One computer salesman told me over the phone that it represents the weakness >>>>>>>or strength of the computer. Does he know what he's talking about? >>>>>>> >>>>>>no >>>>>> >>>>>>>2).What is the difference between the proccessors from Intel Pentium Mobile >>>>>>>technology, if it is 730, 735, 740 till about 770 etc? >>>>>>> >>>>>>speed >>>>>> >>>>>>>Thanks for any answers, information or opinions. >>>>>>> >>>>>> >>>>>>>S.Taylor >>>>>> >>>>>> >>>>>>The Pentium M is a very fast cpu and good for chess however for engines that hit >>>>>>the hash table a lot the slower RAM access of a laptop may slow things down a >>>>>>bit. Might try turning off ETC if it is an option in your favorit chess program >>>>>>when playing on a laptop.
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.