We need a PowerMac 6100-80, Soon...
Andrew Geweke
Test PowerPC 601 Pentium Pentium PPC 601
Clock Speed 66 MHz 80MHz 66 MHz 150 MHz 150 MHz
---------------------------------------------------- ------- -------
SPECint92 62 77 64.5 147 145
SPECfp92 72 93 56.9 130 174
Power (worst case) 8.5w ? 16w ? ?
Die Size (mm^2) 120 ? 262 ? ?
PLEASE NOTE THE LAST TWO COLUMNS ARE MY OWN WORK.
Okay, Intel's "demonstrated" a 150MHz Pentium. I have no doubt that this is a
sort of fluke and would be _extremely_ hard to reproduce in any sort of
quantity.
OTOH, look at this: Although the MPC601 at 80MHz beats the pants off the
Pentium at 66MHz (a VALID comparison; you compare the highest-clocked chips
being produced in quantity at the time -- PC users have compared the
486DX2/66 with a 68040 at 40MHz for along time now) AND it's a lot cheaper
and cooler, right now machines are very different. A cheap 60MHz Pentium
(slower than the 66MHz PPC) can be had for $3000, and prices are falling. But
the chip itself is $875 or something, while the 80MHz PPC is less than HALF
that.
My point? Apple should (IMHO, of course) produce a PowerMac 6100-80. Put in a
512K L2 cache. If they could get the price down around $2000 with monitor and
keyboard -- or even $2500 -- this machine would kill a lot of Pentium sales.
On March 14, you will _finally_ be able to get a Mac that's faster than ANY
PC you can get (the PowerMac 8100-80) -- but, of course, it'll be far more
expensive than the cheapest Pentiums.
We need something that'll kill the Pentium. Apple has shown its willingness
to cut prices. If PowerPC development continues as scheduled, people will
start to look at it, albeit slowly. If they did this (IMHO) people would turn
their heads very quickly indeed.
---------------------------------------------------------------
Andrew Geweke / Computer Engineering, Michigan State University
"I disapprove of what you say, but I will defend to the death
your right to say it." -- Francois Voltaire, 1694-1778 [attr.]
GAT/CS/E/MU/T d? -p+ c+(++++) !l u e+(*) m+(-) s++/
n+ h(--) f(*) g+(-) w+ t+ r-- y?
Steve Dagley
Andrew Geweke, gewe...@studentg.msu.edu writes:
>Okay, Intel's "demonstrated" a 150MHz Pentium. I have no doubt that this is
>a sort of fluke and would be _extremely_ hard to reproduce in any sort of
>quantity.
Andrew,
Don't forget, IBM "demonstrated" a 99MHz 601 chip at Comdex in November. I
don't think they published any SPEC ratings since it was just supposed to be
a technology demo but I heard it was simultaneously doing 2 windows of video
decompressing at 30fps (320x240 or 640x480 I didn't hear) and speech
recognition.
I think the real PPC competitor for the 150MHz P5 is going to be the PPC 604.
I don't remember the actual numbers but at 100MHz I believe it was supposed
to to be more than twice as fast as the 80MHz 601. They will probably ship
at the same time (or the 604 will be first given Intel's track record with
the P5). Of course, I wouldn't mind a Power Mac 6100/80 as you describe
either but I'll be happy to get one of the 6100/60s ASAP.
Steve
(Durn, March 14th seems like a long time to wait)
Mark Rogowsky
>
> In article <940220113...@geweke.ppp.msu.edu>
> Andrew Geweke, gewe...@studentg.msu.edu writes:
> >Okay, Intel's "demonstrated" a 150MHz Pentium. I have no doubt that this is
> >a sort of fluke and would be _extremely_ hard to reproduce in any sort of
> >quantity.
>
> Andrew,
>
> Don't forget, IBM "demonstrated" a 99MHz 601 chip at Comdex in November. I
> don't think they published any SPEC ratings since it was just supposed to be
> a technology demo but I heard it was simultaneously doing 2 windows of video
> decompressing at 30fps (320x240 or 640x480 I didn't hear) and speech
> recognition.
>
but, of course, I have no benchmark for what it was doing, i.e. fast
compared to what? Look for 99 or 100MHz 601s to be real by mid to late
summer of this year. Look for 604s running in the 120MHz plus range by
early 1995. The P6 (which Andy Grove called 3x faster than the current P5)
is due in late 1995. The 604 will compete very favorably with that chip and
beat it to market by six months or better. The battle lines have been
drawn.
Mark Rogowsky
ro...@forsythe.stanford.edu
Michael Lee
: The demonstration at Comdex was of a 95MHz chip, not 99. It was very fast
: but, of course, I have no benchmark for what it was doing, i.e. fast
: compared to what? Look for 99 or 100MHz 601s to be real by mid to late
: summer of this year. Look for 604s running in the 120MHz plus range by
: early 1995. The P6 (which Andy Grove called 3x faster than the current P5)
: is due in late 1995. The 604 will compete very favorably with that chip and
: beat it to market by six months or better. The battle lines have been
: drawn.
What is the speed of the external bus on PowerMacs? I'm all for faster
CPUs but if the external speed cannot keep up, what would be the point
in a 120MHz CPU?
Philip Machanick
(Michael Lee) wrote:
> What is the speed of the external bus on PowerMacs? I'm all for faster
> CPUs but if the external speed cannot keep up, what would be the point
> in a 120MHz CPU?
This is where a bigger cache becomes necessary. Some Silicon Graphics
models for example run internally at 150MHz with a 50MHz bus, and smooth
over the difference with 1M of fast second-level cache. The bigger the
internal/external speed difference, the bigger the cache is a good general
rule. Another factor is how big the on-chip cache is, since that runs at
full CPU speed.
--
Philip Machanick phi...@concave.cs.wits.ac.za
Department of Computer Science, University of the Witwatersrand
2050 Wits, South Africa
phone 27(11)716-3309 fax 27(11)339-7965
Cyrus Shafai
>What is the speed of the external bus on PowerMacs? I'm all for faster
>CPUs but if the external speed cannot keep up, what would be the point
>in a 120MHz CPU?
Apple will soon (fall 94?) be putting PCI busses in the PowerMacs. They
are rated at 132 Mbytes/sec transfer rate. Someone correct me if I am wrong.
--
-----------------------------------------------------------------------------
Cyrus Shafai (sha...@ee.ualberta.ca) |
Department of Electrical Engineering | "Never share a foxhole with
University of Alberta, Edmonton, Canada | someone braver than youself"