Re: Too good to believe?

Perry Harrington (
Sun, 3 May 1998 17:39:37 -0700 (PDT)

> sounds *too* good. I use "make -j install" all the time on my 4-processor
> pentium pro 200mhz machine from ALR. It takes exactly 3 minutes to
> complete. No way a dual is going to run in 1/4th that time...

It takes 4 minutes to build 2.0.34 on a UP PPro 200/256 at work. This includes:

Standard stuff

> I also had, until about 2 weeks ago, a dual PII/300, and it was slower
> than my quad-pro machine.
> >
> > him2: Just an IDE drive, but with multithreading enabled...
> >
> > me1- [I am about to put together a dual PPro166/512 box with the 256MB
> > me1- of main memory in a SuperMicro P6DOF board [Orion chipset, the
> > me1- 4-way interleaved memory.]
> >
> > him1: The P2 is designed to run Windows 95, not a full 32 bit OS like Unix
> > him1: or NT, conversely a PPro runs '95 rather poorly.
> >
> this is nuts. The PII is the same architecture as the Ppro, except that

Ahh, but it isn't entirely. I have it on good faith (from a Borland hacker
friend) that intel "fixed" 16 bit code on the P2. Apparently it runs 16
bit code a lot faster than the PPro. However this is all moot, since I've
got Linux installed on all 11 of my PPros at work :)

> the L2 cache is not as fast. The Pro runs the L2 at 200mhz, the PII/300
> (or any PII) runs the L2 at clock/2. But internally they look the
> same, at least according to the stack of manuals I have from each,
> published by intel. The PII's do have the mmx stuff, but that doesn't
> do a thing at present...

Perry Harrington       Linux rules all OSes.    APSoft      ()
email: 			Think Blue. /\

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to