🚀 go-pugleaf

I'll put a few of these together. By the way the original post is mine
and was re-printed from another list. It was intended to let PC people
know what's going on in the worl of PCs not to fuil a Pentium Pro vs
Power PC debate.

> : The 200Mhz Pentium Pro Spec benchmarks (8.09 SPECint95) show that on 32
> : bit code, it's integer performance beats anything now available,
> : including the 300Mhz Dec Alpha (7.33 Specint95) and 133Mhz Power PC 604
> : (4.07 SPECint95).
> 
> Well, besides the fact that the 604 is available at 150 MHz, you're
> comparing "preliminary SPECs" versus SPECmarks in a selling computer.

> Let's talk actual systems, not estimated processor benchmarks.
> January 1996 Byte, Page 140 Graph "Raw Performance: The VIS-less Ultra I".

Preliminary Yes, in that no system is currently shipping with these parts
in them. But they were run on a real, but probably souped up system.
Still the difference in performance between PowerPC and Pentium Pro is nearly
2x and I would bet real system performance is not going to be half of the
performance of Intel's reference system.

> Come back when there are real systems to test.

Whats the point. It does not take long following discussions such
as this one to realize that most people quote benchmarks when they
support their view and argue with them when they don't.

> : It even beats the estimated performance of any Power
> : PC chip yet talked about including the new PowerPC 620 (225 Specin92).
> 
> The 620?  How about the 604e?  The current line on the 620 is that it's
> being redesigned, and the last I heard, the 604e should be out sometime
> this year in a faster (200 MHz+) model that should give performance about
> even with the real-world PPro 200 MHz machines.

Not what I've heard. I didn't mention the 604e because I did not have
the numbers in front of me to back up any statements I might make. I
don't like vague unsupported statements like the one above. At the risk
of making my own vague statements, the information I've seen from Mot and
IBM places the 604e well below the 200Mhz Pentium Pro. If you have other
information please post it. I would be interested in seeing it.

It's real clear that the Power PC guys have badly missed their mark. They
stated at the start of the consortium that to make headway into the Windows PC
market they would need to have twice the performance at half the cost. They
have done neither, and now the first silicon of the 620 has had so many problems
they pushed the release date back 6 months. By there own estimates, their fallback,
the 604e is not going to even match the Pentium Pro on performance. These guys
are in trouble.

> ...or it's possible that the analysts are just a little bit wrong.

I've seen two reports, one from Microprocessor Report is quite detailed.
It discused the subject in a level of detail I am not equiped to comment
on. The two primary supporting pieces of evedence are the architecture of
the clock distribution network. It seems the placement and type of clock
buffers result in much lower skew than on previous processors. The other
was the placement and interconnection of functional units. Although I can't
recall the details I believe that they have arranged the architecture with
an unusualy high number of redundant functional units so that highly
highly interconnected blocks can be placed very close together. The result
is that the interconnect between functional units is very fast.

> WHY are the Pentium Pro SPECint/fpXX scores so high in comparison to the
> PowerPC and other CPU's, when the BYTEMark scores are so low comparing
> the same systems?

Prior to the release of the Pentium Pro Intel only refered to 133Mhz and 150Mhz
parts. These are the ones that the Byte benchmarks are based on. When the part
was released it was released at much higher clock frequencies, 150Mhz and 200Mhz.
Give Byte some time and they will release new benchmarks. I am as interested
to see how these will turn out as anyone.

One of the reasons I generated the original post (to a PC list) is that it
seems the press have given up on the Pentium Pro. If you watch Computer Shopper,
Byte, or PC Magazine there is very little said on the subject. It's almost as
if these magazines ran their initial benchmarks with 133Mhz Pentium Pro's,
found lack luster 16 bit code performance and decided the processor was
irrelevent. The unexpected release of 200Mhz processors and the accompanying
performance increase stike me as very exceiting news for PC owners, yet have
received little notice in the press.

-- 
David Scidmore
davids@globaldialog.com