Re: Pentium (R) Pro processor press release (fwd)

Michael Sweger (swegerm@baileys-emh5.army.mil)
Fri, 03 Nov 95 13:50:08 EDT


Hello,

Thanks for the info. However, the DEC alphas according to INFOWORLD
will have a 50% increase in speed (they didn't specify 50% in what),
this could mean an increase from 300mhz to 450mhz. They did mention
that the specint will be approx 500.


I would like to add, what good is a fast microprocessor when the
current ISA/EISA bus speeds ( and I presume the same for the PCI buses
coming out) are no faster than 33mhz or less with DRAM access speeds
still being at 60-70 nano-sec with no memory interleaving (i.e. a
two way interleave would get two memory values for each access and
make the 60 ns memory appear as if it were 30 ns. This is an old
super computer memory trick). These are the current bottlenecks
in home PCs and which are important for multimedia applications that
require high I/O bandwidth, even hough the CPU's can crunch numbers
faster. The onboard cache helps some, but I can't store a
multi-megabyte compressed MPEG in this cache alone and therefore I
need to depend on a fast disk with high I/O bus speeds.


I'm waiting for the day when memory access times available in PC's
approach 20ns or less: although RAM and DRAMs with these times are
already available, but costly. I'm waiting for the day when all
buses (disk,video, etc) run at a speed of at least 50% of the CPU
clock speed. Therefore, with each increase in clock speed, the
bus speed would also increase for the PC being sold to the consumer.


I know what PC's can be designed to, if only the PC manufacturers
would do it!


Does anybody know of a PC that is being sold that runs at a
bus clock speed greater than 33mhz?

I realize alot of manufacturers spec their bus speeds only in terms
of bytes/sec. This can be misleading considering that typical buses
can be 16 and 32 bits wide ( 2 bytes and 4 bytes wide). Thus, if I
took a 33mhz bus and made it 16 bits this would be (33mhz*2 bytes= 66
Mbytes/sec) vs. a 32 bit bus (33mhz*4 bytes= 132 Mbytes/sec) and the
numbers look great. But when are the bus clock speeds going to
increase?


Mike Sweger,
swegerm@baileys-emh5.army.mil

______________________________ Reply Separator _________________________________
Subject: Pentium (R) Pro processor press release (fwd)
Author: eabu927@ea.oac.uci.edu at DDN/Internet
Date: 11/3/95 6:24 AM

>From @listproc.mail.cornell.edu:owner-CU-SEEME-L@cornell.edu
X-Envelope-From: @listproc.mail.cornell.edu:owner-CU-SEEME-L@cornell.edu
Received: from listproc.mail.cornell.edu by host.baileys-emh2.army.mil id
aa10965 ; 3 Nov 95 2:48 EST
Received: from localhost.mail.cornell.edu (localhost [127.0.0.1]) by
listproc.mai l.cornell.edu (8.6.12/8.6.9) with SMTP id BAA20067; Fri, 3 Nov 1995
01:25:28 -050 0
Received: from cornell.edu (cornell.edu [132.236.56.6]) by
listproc.mail.cornell. edu (8.6.12/8.6.9) with SMTP id QAA00340 for
<CU-SEEME-L@listproc.mail.cornell.ed u>; Thu, 2 Nov 1995 16:25:56 -0500
Received: (from daemon@localhost) by cornell.edu (8.6.12/8.6.12) id QAA22375 for
CU-SEEME-L@listproc.mail.cornell.edu; Thu, 2 Nov 1995 16:29:23 -0500
Received: from taurus.oac.uci.edu (taurus.oac.uci.edu [128.200.80.23]) by
cornell .edu (8.6.12/8.6.12) with SMTP id QAA22353 for <CU-SEEME-L@cornell.edu>;
Thu, 2 N ov 1995 16:29:21 -0500
Received: by taurus.oac.uci.edu id AA15487
(5.65c/IDA-1.4.4 for CU-SEEME-L@cornell.edu); Thu, 2 Nov 1995 13:29:06 -0800
Message-Id: <Pine.SUN.3.91.951102132655.5275r-100000@taurus.oac.uci.edu>
Date: Thu, 2 Nov 1995 13:29:05 -0800 (PST)
Reply-To: eabu927@ea.oac.uci.edu
Sender: owner-CU-SEEME-L@cornell.edu
Precedence: bulk
From: Claudiu Dan <eabu927@ea.oac.uci.edu>
To: <CU-SEEME-L@cornell.edu>
Subject: Pentium (R) Pro processor press release (fwd)
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-To: Logos1994@aol.com, CU-SEEME-L@cornell.edu, dave sims <sims@usa.acsys.com>
X-PH: V4.1@cornell.edu (Cornell Modified)
X-Listprocessor-Version: 7.2(a) -- ListProcessor by CREN


---------- Forwarded message ----------
Date: Thu, 2 Nov 1995 08:49:46 +0200 (EET)
From: STUDENT-L@lorien.creol.ucf.edu
To: student-l@ulise.cs.pub.ro, MSD <STUDENT-L-SERVER@MSD-project.pub.ro>
Subject: Pentium (R) Pro processor press release (fwd)

Forwarded message:
>From owner-p6info@mailbag.intel.com Thu Nov 2 02:20:50 1995
Sender: owner-p6info@cs.pub.ro
Message-Id: <m0tAjng-00080YC@www.jf.intel.com>
Date: Wed, 1 Nov 95 12:30 PST
From: Clif_G_Purkiser@ccm.sc.intel.com
To: p6info@mailbag.intel.com
Subject: Pentium (R) Pro processor press release
Sender: owner-p6info@mailbag.intel.com
Precedence: bulk



PENTIUM(r) PRO PROCESSOR TO POWER NEXT WAVE OF HIGH-PERFORMANCE
WORKSTATIONS AND SERVERS BASED ON INTEL ARCHITECTURE

Speeds Range from 150 MHz to 200 MHz

SANTA CLARA, CA. November 1, 1995 -- Intel Corporation today introduced the
next generation of the Intel Architecture, the Pentium(r) Pro processor,
debuting at speeds as fast as 200 MHz, with 366 SPECint92 performance.
The 5.5-million transistor chip achieves performance levels that exceed
today's workstations and servers that use proprietary RISC technologies,
signaling the arrival of PC price/performance to these high-end market segments.


< The rest of the text has been deleted to make room for my opinions.>