From owner-freebsd-hackers  Sun Aug  3 09:31:11 1997
Return-Path: <owner-freebsd-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id JAA01122
          for hackers-outgoing; Sun, 3 Aug 1997 09:31:11 -0700 (PDT)
Received: (from jmb@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id JAA01116;
          Sun, 3 Aug 1997 09:31:08 -0700 (PDT)
From: "Jonathan M. Bresler" <jmb>
Message-Id: <199708031631.JAA01116@hub.freebsd.org>
Subject: Re: Pentium II?
To: tony@dell.com (Tony Overfield)
Date: Sun, 3 Aug 1997 09:31:07 -0700 (PDT)
Cc: cjs@portal.ca, freebsd@atipa.com, tom@sdf.com, hackers@FreeBSD.ORG
In-Reply-To: <3.0.2.32.19970803041915.006a69e4@bugs.us.dell.com> from "Tony Overfield" at Aug 3, 97 04:19:15 am
X-Mailer: ELM [version 2.4 PL24]
Content-Type: text
Sender: owner-freebsd-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

Tony Overfield wrote:
> 
> At 03:10 PM 8/2/97 -0700, Curt Sampson wrote:
> >
> >On Sat, 2 Aug 1997, Tony Overfield wrote:
> >
> >> Since the hit rate of the L1 cache is usually much higher than that of
> the L2 
> >> cache, the effect of the slower L2 cache in the Pentium II is usually
> offset 
> >> by the beneficial effect of not having to access it.  So even at the same 
> >> clock rate, the Pentium II can run faster than the Pentium Pro.
> >
> >Do you have any benchmarks that indicate this? 
> 
> I think many of the benchmarks indicate this.  The benchmarks show, when 
> run at the same clock frequency, that the Pentium II runs at speeds 
> comparable to the Pentium Pro, even though the L2 cache is running at 
> half-speed.  Many folks had claimed that the Pentium II would be much 
> slower because of the half-speed L2 cache.

	oh?  what is the size of your dataset?  what is the data access
	pattern?  without specifing these two items, i cant tell how 
	your are using L1 and L2 cache.

	what we need is a benchmark that has a fixed data access pattern
	and known data set size.  better yet would be one that starts with
	a very small data set and grows the data set till the computer starts
	using disk.  a graph of the results would show the speed of the 
	machine accross all its memory regimes.

	http://www.scl.ameslab.gov/scl/HINT/HINT.html
jmb

> 
> It should be easy to agree that larger L1 caches have higher hit rates.  In 
> turn, higher L1 cache hit rates reduce the demand on the L2 cache.  Whenever 
> the Pentium II is hitting in the L1 cache strictly due to its larger size, it 
> will be faster.  Whenever the Pentium II misses the L1 cache, it will be 
> slower.
> 
> >Or are you just dreaming?
> 
> I sure hope not.  I wouldn't want to waste a dream on this stuff.  :-)
> 
> -
> Tony
> 
>