hardware requirements

[Wm E. Davidsen Jr]

In article <1990Dec3.162251.19808 at oct1.UUCP> mason at oct1.UUCP (David Mason) writes:
>Did Ken mean a numeric or a graphic co-processor.  I thought he meant
>numeric.  Can anybody tell me if a '387 makes a difference with ISC
>2.2 X?  If yes, then some quantitative numbers would help.

  I added a Cyrix 387sx to the little system I'm using to test V.4. By
running a bunch of "normal" X things and looking at the CPU time used by
the X process, and then repeating the process with the FPU installed, I
noted that it run *five* timesfaster with FPU. Yes I know you can't
believe it, measure it yourself before telling me it can't be so.

  With the FPU float is about 100x faster (I have the figures on that,
too), from 0.026 MFLOP to 2.2MFLOP (roughly). The five to one means that
with the FPU in 4-5% of the time is float ops.

  Now that was with X11R4, X11R3 which I believe you have is far more
float intensive, so I would loook for a big jump in performance.
However, if you don't have enough memory nothing will save you. A 486-25
running ODT took 25 sec to bring up an app with 4MB, 4 sec with 8MB.

  I didn't believe these numbers until I measured them, and I realize
that a number of people who haven't measured them will try to discredit
them. Tough.

  Taiwan board, NEAT chipset, 16MHz, Cyrix 387sx, Dell V.4.
-- 
bill davidsen - davidsen at sixhub.uucp (uunet!crdgw1!sixhub!davidsen)
    sysop *IX BBS and Public Access UNIX
    moderator of comp.binaries.ibm.pc and 80386 mailing list
"Stupidity, like virtue, is its own reward" -me