Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!rutgers!mit-eddie!uw-beaver!tektronix!percival!bucket!leonard
From: leonard@bucket.UUCP (Leonard Erickson)
Newsgroups: comp.sys.ibm.pc
Subject: Re: Speeding up old PC's
Message-ID: <1101@bucket.UUCP>
Date: 12 Oct 88 21:55:41 GMT
References: <234e0a42@ralf>
Reply-To: leonard@bucket.UUCP (Leonard Erickson)
Organization: Rick's Home Grown Unix; Portland, OR
Lines: 28

In article <234e0a42@ralf> Ralf.Brown@B.GP.CS.CMU.EDU writes:
<Rule-of-Thumb memory speed: (many memory chips actually perform better)
<
<        max_access_time_ns = (1000 / clock_speed_MHz) * clocks_per_access / 2
<
<where clocks_per_access is FOUR for 808[68] and TWO plus wait states for
<80[23]86 machines.
<
<So, the calculation for an original PC goes:
<
<        1000/4.77 * 4 / 2 = 210 * 4 / 2 = 420ns
<
<and for a 10/1 AT
<
<        1000/10 * 3 / 2 = 100 * 3 / 2 = 150ns

Sorry, but your formula produces bogus answers. As an example, it says that
a 10 Mhz XT clone can use 200 ns RAM. When you compare this with reality,
you find that you need at least 120 ns RAM to get anything resembling
reliable operation.

I could give other examples but the bottom line is that your formula gives
values that are roughly double the "real" answer...
-- 
Leonard Erickson		...!tektronix!reed!percival!bucket!leonard
CIS: [70465,203]
"I used to be a hacker. Now I'm a 'microcomputer specialist'.
You know... I'd rather be a hacker."