🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.sys.intel
Article #1444

Re: 486SL dropped, vendors test Cyrix

#1444
From: Bill Henning
Date: Sun, 10 Jan 1993 21:27
47 lines
2613 bytes
In article <C0HsJp.ALy@inews.Intel.COM> rskinner@mipos2.intel.com (Rod Skinner) writes:
>
>Machine				Processor/MHz		SYSmark92
>Compaq LTE Lite/25c		Intel386 sl-25MHz	51.90
>NEC PowerMate 3x25a		Intel386 sx-25MHz	51.60
>PC Brand LeaderBook 486LC/25	Cyrix 486slc-25MHz	48.69
>TI TravelMate WinSLC		Cyrix 486slc-25MHz	47.23
>Aspen Personal 486		Cyrix 486slc-25MHz	43.65
>Keydata Keynote 486slc-25	Cyrix 486slc-25MHz	43.55
>Compaq Prolinea 3/25 2s		Intel386 sx-25MHz 	41.35 **
>Micro Express NB486slc/25	Cyrix 486slc-25MHz	41.15
>Grid 1755/486slc		Cyrix 486slc-25MHz	37.31
>
>** no cache (on or off chip)

A SYSmark92 of 48.69 for the PC Brand LeaderBook 486LC/25 (the fastest
486SLC figure you posted) vs. the SYSmark92 of 51.90 for the Compaq
LTE Lite/25c (486SL) makes for a difference of about 7%.

This is somewhat misleading, as this is not a purely CPU benchmark.

>It just means that you need to be careful when buying a machine that has
>the numbers "486" on the name plate.  The system may have 486 in the
>name and may execute "all the 486 instructions" but the thing that is
>missing is 486 performance.  That may change for the Cyrix 486DLC but I
>don't have any application performance results that would allow me to
>may any conclusions other than to expect Cyrix 486DLC performance 
>similar to the Intel386 DX at equivalent frequencies.

Rod, I can't let that pass without a comment. Intel builds some truly
excellent CPU's, and I am looking forward to the Pentium (R) but please
do not equate a Cyrix 486DLC with an Intel 386DX at the same clock speed.
Even a 486SLC beats a 386DX at the same clock speed according to some
benchmarks I've seen.

I believe that a 486DX, or SX, or SL is faster then the 486DLC at the
same clock speed - makes a lot of sense, 8k cache vs. 1k cache ought
to mean 80%+ hit ratio vs. 60%+ hit ratio, but I refuse to believe 
that a 486DLC even with the only 1k cache will not beat a 386DX at
the same clock frequency, probably by a factor of 2.

>rod
>
>Rod Skinner				I speak for myself only.
>Intel Corp
>2200 Mission College Blvd RN4-21          PH (408) 765-4474
>Santa Clara, CA 95052                     FAX (408) 765-4920
Technical Details
Message-ID: <C0nqAF.B42@wimsey.bc.ca>
Path: arc1-ca.pugleaf.net!sparky!uunet!van-bc!bhenning
References: <C0E0zB.DwG@inews.Intel.COM> <1993Jan6.135936.23923@crd.ge.com> <C0HsJp.ALy@inews.Intel.COM>

Show Full Headers
Complete Article Headers
Newsgroups: comp.sys.intel
Path: arc1-ca.pugleaf.net!sparky!uunet!van-bc!bhenning
From: bhenning@wimsey.bc.ca (Bill Henning)
Subject: Re: 486SL dropped, vendors test Cyrix
Organization: Wimsey Information Services
Date: Sun, 10 Jan 1993 21:27:51 GMT
Message-ID: <C0nqAF.B42@wimsey.bc.ca>
References: <C0E0zB.DwG@inews.Intel.COM> <1993Jan6.135936.23923@crd.ge.com> <C0HsJp.ALy@inews.Intel.COM>
Lines: 52