Re: [berman@example.com: LINUX on Alpha info]

[TMatsumu@example.com]

Form: Reply
Header: Adaptec
Text: (25 lines follow)
re:

 > Certainly it will do everything more quickly, 

   This is certainly not true.

		Is too.  Tell me something a 64-bit processor will do more slowly than 
		a 32 bit processor.  (I can think of a couple.)  Then tell me how it
		affects George (I think was the one who asked).

Virtually the running of any 16 or 32 bit native application will run 
slower.  

Where you are getting confused is clock speed.  Clock speed is what will 
improve performance of apps not written to utilize 64 bit word sizes.

George is probably not running any native 64 bit applications, so therefore 
this affects George greatly.  Almost nothing will run faster with a 64bit 
processor change only.

DX-2, DX-4, these are not ;64 bit processors, these are double external 
clocked versions of the DX-xx family.  Where did you think they were 64 
bits?


Original text: (74 lines follow)
>From owner-tlug@example.com, on 3/26/96 11:10 AM:
To: tlug@example.com

>>>>> TMatsumu@example.com writes:

   > Certainly it will do everything more quickly, 

   This is certainly not true.

Is too.  Tell me something a 64-bit processor will do more slowly than 
a 32 bit processor.  (I can think of a couple.)  Then tell me how it
affects George (I think was the one who asked).

   Any new 64 bit processor will need applications specifically written for 
64 
   bits, not just recompiled for it, and will also need a 64 bit bus.  Right 


Oh, come now.  You still writing in assembler?  (Come to think of it,
at least for DOS, at Adaptec you probably are.)  How come my 32-bit
linux Ghostscript uses the same source (except for drivers) as the
16-bit MS-DOS version?  Machine dependencies, including word size, are
all in the GCC machine specs.  Yes, you can write programs that depend
on word size, but you don't do it unless you need to.  I mentioned
several of those cases before.  As for the 64-bit bus, if you had it
that would guarantee real improvements on almost everything.  But if a
64-bit processor won't give you improvements simply due to bus
constraints, what's a DX2 or DX4 good for?

   now, the PCI bus on the Alpha is 32 bits.

   NT disk i/o is fairly advanced, I haven't seen anything with Linux that 
   touches the built in mirroring and striping w/parity and w/out parity on 
a 
   dual channel PCI host adapter.

Now this is sensible, although the original poster was not interested
in RAID.  Good disk IO really will hot up a system.

Ol' Craig apparently has

    cd /usr/src
    mv linux linux-`uname -r`
    if batchftp file://SunSITE/linux-(`uname -r` + 1).tar.gz; then
      tar xzf linux-(`uname -r` + 1).tar.gz
      mv linux-`uname -r`/.config linux/
      cd linux
      make zImage
      mv zImage /vmlinuz
      lilo
      shutdown -r now
    fi

in his /etc/rc.d/rc.local.  (Craig, I'm JUST KIDDING, *don't* put that
in your rc.local 'cause it won't work!)  I can think of a few other
reasons for a Pentium 166 or 200 or Alpha (real ones, not just
bragging rights over co and jwt).  But for most people, fast disk I/O
(the 4MB hardware cache on my old 486DX/50MHz makes it a much better
Web-server than the Gateway Pentium 120 I'm using until I get the disk
replaced) and lots of RAM to hold X bitmaps and avoid swapping
processes and to hold hpscan PBMs for xv ;-) is more useful.

Hell, if you had 256MB of ram, wouldn't making a 32MB ramdisk and
putting /usr/lib/gcc-lib and /usr/src/linux in it make a 60MHz Pentium 
faster than a 166MHz PPro with 16MB?

IMHO (as if *I* have any HO).

-- 
                            Stephen J. Turnbull
Institute of Policy and Planning Sciences                    Yaseppochi-Gumi
University of Tsukuba                      http://turnbull.sk.tsukuba.ac.jp/
Tennodai 1-1-1, Tsukuba, 305 JAPAN                 turnbull@example.com

Use Proportional Font: true
Previous From: owner-tlug@example.com
Previous To: tlug@example.com