Article View: comp.sys.intel
Article #200Re: 486DX2 vs 486DX
From: Rod Skinner
Date: Fri, 07 Aug 1992 16:00
Date: Fri, 07 Aug 1992 16:00
88 lines
4785 bytes
4785 bytes
In article <1992Aug7.024301.20079@schbbs.mot.com> P70008@WACCVM.corp.mot.com writes:
>We are thinking about purchasing a number of i486 machines and had heard
>about problems with the i486DX-50. Are these problems solved?
Old, old problem correct many moons ago.
>Should we go with a i486DX2-50/25 instead. We will be using the
>machines as Workstations (most of them running Windows).
>
BAPCo will have some of the information that would help you
make your decision.
The Business Applications Performance Corporation (BAPCo) is an
organization of companies (AMD, Chips&Tech, Dell, DEC,
Gateway 2000, HP, IBM, Infoworld, Intel, Lotus, Microsoft,
NCR, Novell, NSTL, and Ziff-Davis Labs) developing business
applicaton benchmarks to measure system performance. The
first benchmark suite was recently released and for $795 includes
the installation and measurement tools, the applications being
measured (DOS, Windows 3.0, WFW, Wordperfect, Lotus 1-2-3, Excel,
Quattro Pro, dBASE IV, Paradox, Harvard Graphics, PageMaker,
Borland C++, Microsoft C).
The first BAPCo Newsletter is nearing publication and I have
pulled out the "best of breed" system results submitted by the
member companies and have included them below. These are the
systems results with the best times foreach of the processors
reported so far. I think there will be approximately 40 system
results in the first Newsletter. The BAPCo reporting rules
require that these results be easily reproduceable and therefore
must be on a system that is being currently sold by the manufacturer.
The Report which is published 4 times a year is $250 per year.
Call BAPCo for more information:
Phone (408) 988-7654
Metrics are relative to a standard machine and so bigger is better.
SYSmark92 is the overall metric, but one should characterize the
application workload and choose machines that make sense for them.
CPU FPU CLK Mem Cache SYSmark92 WP SS DB DG DP SD
Intel486DX Int 50 12MB 128KB 187 171 175 188 215 195 303
Intel486DX2 Int 25/50 12MB None 154 139 153 131 196 164 243
Intel486DX Int 33 8MB 128KB 131 124 124 124 141 139 205
IBM486SLC2 CX387SX 20/40 8MB None 100 93 99 84 114 112 148
Intel486SX None 25 12MB None 97 89 92 85 121 108 155
AM386DX None 40 8MB 64KB 84 77 81 83 94 87 125
Intel486SX None 20 8MB 128KB 82 75 78 75 96 88 123
Intel386DX None 33 8MB 64KB 77 71 72 74 95 79 114
IBM386SLC CX387SX 20 8MB None 57 52 58 50 62 67 86
Intel386DX None 25 8MB 64KB 55 51 53 51 60 59 82
AM386SX None 25 10MB None 40 38 35 41 49 44 58
Intel386SX CX387SX 20 8MB None 35 33 35 31 38 40 51
WP - wordprocessing (WordPerfect, WFW)
SS - spreadsheet (1-2-3, Excel, QuattroPro)
DB - data base (dBase, Paradox)
DG - desktop graphics (Harvard Graphics)
DP - desktop publishing (PageMaker)
SD - S/W development (Borland C++, MS C)
SYSmark92 is a weighted mean of the 6 application catogory results.
The weighting was determined by survey results that BAPCo contracted
to have performed. It weighs the things that a typical users
does most frequently most heavily. The survey results were used to
determine the composition of the workloads. They were designed to
match the things that users said they were doing.
Since these benchmarks are based on the same applications that many use
the computer for, they should more accurately reflect the system
performance one would see than small synthetic benchmarks with "unrealistic"
instruction mixes such as Landmark and PC Magazine type of benchmarks.
>If all is OK with the i486DX-50, why doesn't Compaq or IBM have any
>systems?...any and all comments welcome.
We had a Compaq 50 MHz system delivered yesterday. We have a couple
50 MHz machines in the lab. My only comment would be why would one
say they have no 50 MHz machines?
>
>-Bill
Rod Skinner I speak for myself only.
Intel Corp
3065 Bowers Avenue MS NW1-20 PH (408) 765-4474
Santa Clara, CA 95052 FAX (408) 765-4920
Technical Details
Message-ID:
<12808@inews.intel.com>Path:
arc1-ca.pugleaf.net!sparky!uunet!gatech!destroyer!sol.ctr.columbia.edu!eff!iWarp.intel.com|inews!mipos2!rskinnerReferences:
<1992Aug7.024301.20079@schbbs.mot.com>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!gatech!destroyer!sol.ctr.columbia.edu!eff!iWarp.intel.com|inews!mipos2!rskinner From: rskinner@mipos2.intel.com (Rod Skinner) Newsgroups: comp.sys.intel Subject: Re: 486DX2 vs 486DX Message-ID: <12808@inews.intel.com> Date: 7 Aug 92 16:00:19 GMT References: <1992Aug7.024301.20079@schbbs.mot.com> Sender: news@inews.intel.com Reply-To: rskinner@mipos2.UUCP (Rod Skinner) Organization: Microprocessor Component Group, Intel Corp., Santa Clara, CA Lines: 88