#Motorola Unveils 68060
26 messages in this thread
> 68060 ……. roughly equally powerful Pentium
?? the processors have always lined up 68000 <=> '286; '020 <=> '386..
for the same clock speed as far as I know. At least Dr Dobbs came to that
conclusion after applying every test known to computerkind.
(they also said you could prove what you liked if you used the right test
🙂
regards, Lu on AutoPilot downunder
Lu-
Motorola skipped one generation – the 050. So the 060 and the Pentium
are roughly comparable.
Paul
Paul,
The rest of the equation I assume would have gone '030<=>'486;
'040<=>Pentium; '060<=>? As the Dobbs article pointed out, it's what
happens at the blunt end (users) that counts. This is where the
instruction set comes into it and you can prove what you like if you use
the right test :). I respect their work the more because they were so
thorough and they had the guts to say `well overall they roughly equate..'
instead of trying to deal in absolutes and the hype of some popular
magazines.
Now if only one of them would come out with twin overhead foxtails and
custom duco….. 🙂
regards, Lu on AutoPilot downunder
Lu-
I don't know quite what you mean, but, very roughly speaking, the 030
was somewhat better than the 386, the 040 was better than an equally
clocked 486 but the 486 clock has been pushed far higher than the 040
clock, and the 060 is roughly comparable to the Pentium at equal clocks,
though again the Pentium has been pushed higher faster.
Paul
I was under the impression that clock speed, in it's basic form, doesn't
necessarily make a machine fast. Now, let me qualify that with "100mhz
DOES make a fast machine", but more via brute force than actually
architectural integrity… but just because a system has a higher clock
rate than another system, doesn't always make it faster.
Chris – Fast and Low on my '020!
…Is it significant that my spell checker doesn't know "Intel"?
Oh, sure, clock speed is -really- only useful in distinguishing between
same-model chips, though it also has ramifications for system design –
witness all those DX4 and DX2 and so on chips out there.
Btw, why on earth is a DX2 clock DOUBLED and a DX4 clock TRIPLED????
Paul
I hadn't really paid that much attention to the DX4, so I hadn't really
noticed… (AAMOF, I really don't pay much attention to the DOS-BOXES
anyway… 'cept maybe the notebooks! I really could find a use for one of
those!)
Something about a chip that runs 4 times faster than its motherboard can
accomodate ALSO falls into the Brute Force category.
Chris – Fast and Low on my '020!
…Is it significant that my spell checker doesn't know "Intel"?
>Btw, why on earth is a DX2 clock DOUBLED and a DX4 clock TRIPLED????
There were already clock tripled 486 class chips on the market. Intel didn't
want their tripled chip to be confused with the others, and knowing that they'd
never make a quadrupled 486, they decided to use the old marketing ploy of
upping the number for no reason. A DX4 is a clock tripled chip with a
marketing-quadrupled number.
Michael @ Soft-Logik Publishing Corp.
"why on earth is a DX2 clock DOUBLED and a DX4 clock TRIPLED???? "
Before someone answers that one, I want to know why it's called a
three-way switch when you have two switches controlling the same light and
a four-way is three switches doing the same thing?
🙂
Christopher
Too late, someone already answered. <g> In case you're interested, it
turned out that IBM already has clock-tripled chips (their Blue
Lightenings come in a clock-tripled flavour) so Intel wanted to get a
marketing jump. BOO! HISS! As to the switches, ya got me mate.
Paul
I presume you're looking for a silly answer, but a "three-way switch",
with two switches controlling the same light, is technically called that
because both throws on each switch are "hot", i.e., "on if the other
switch is in the same position"; thus there are two "on" states (both up
or both down), and one (this logic escapes me) "off" state (one up and one
down).
A "four-way" I don't know about: If Switch A (where you're standing) is
Down, Switch B is Up, and Switch C is Down, you can't turn the light on no
matter what you do, can you?..? :-\
Darned if I know. But I can speculate that it may have something to do with
the fact that a "three-way" switch has 3 connections (as opposed to two on a
regular switch); and that swtiches beyond the first two need to have four
connections, which must make "four-way".
Mebbe it comes from electrical suppliers, whose stock clerks figured that
"way" meant "connection points".
–Jim
Chris, it is still called a four way switch when there is 5-6-7 or any
number of switches greater than 3. 3 way or 4 way refers to the way the
switch is hooked up rather than the number of switches. (I'm an
Electrical Contractor)
…Steve
====>>> Motorola 68030 Inside <<<====
> I was under the impression that clock speed, in it's basic form,
> doesn't necessarily make a machine fast. Now, let me qualify
> that with "100mhz DOES make a fast machine", but more via brute
> force than actually architectural integrity… but just because
> a system has a higher clock rate than another system, doesn't
> always make it faster.
High clocks also imply that the CPU is "simple", this can mean it's cheap to
make.
In the end all that matters is raw speed. There are two basic ways to achieve
this. The "Super Pipelining" camp says make it super simple and push the clock
very high. This is what MIPS does.
The other camp says go "Super Scalar", add in execution units to let more
things happen in parallel. This makes it slightly harder to push the clock
high. PowerPC is in this camp.
The CISC chips are generally limited to "Super Scalar", but get fewer benifits
because the overall complexity don't allow the execution units to do as much as
they should.
In the end, it's easier to throw execution units at something than to raise the
clock. 200Mhz clocks are expensive to control and distribute.
Also, a clever "Super Scalar" design will also allow high clocks.
For whatever reason, Moto seems to have had trouble getting the clocks up on
the 68K family. The '060 will barely reach 66Mhz by the end of the year. By
any definition it's a "medium" speed chip. It will help people run exisiting
binaries faster, or make a simple "plug-in", but it won't compete in raw speed
with any other modern CPU.
—— Andrew
Thanks for the in-depth reply…
Who do I make the tuition check out to? <g>
Chris – Fast and Low on my '020!
…Is it significant that my spell checker doesn't know "Intel"?
What you say is true, that is, a fast clock speed doesn't necessarily mean a
fast system. To use (another) car analogy, a big engine doesn't necessarily
make a fast car. BUT, sooner or later, if you put a big enough engine in, a
car is going to be fast. It may handle like crap, but it will be fast.
Witness the latest muscle pickup wars.
I've found with PCs in particular that two factors make more "perceived"
difference than the processor, however, and that's disk and display
performance. A slow VGA card or older IDE drive will make a 66MHz 486 machine
seem "slow" compared with a 33MHz 486 with snappy VGA card and either a newer
IDE drive or SCSI.
A lot of people get rid of an old PC when in fact by putting in an accelerated
VGA card and a new hard disk (or SCSI card and drive) makes a BIG difference.
The processor does help display and disk, but the underlying additional
hardware is key. The processor is good obviously for computational things.
Reflowing text in a high end wordprocessor or desktop publishing program,
recalculating a spreadsheet, sorting and searching in a database (although
again the disk plays a big part here), these are the tasks where a high speed
processor helps. I've seen Word6 for Windows on a 66MHz 486 and a Pentium, and
there's a noticable difference.
The PCI video card I have in my Pentium machine at work is absolutely
incredible. I thought the S3 adaptor I had in my last machine was amazing, but
this is much faster. Display updates appear to be instantaneous.
Good display and disk technology are the very things that have made the Amiga
so great. The display was WAY ahead of PCs in raw performance a while ago, and
having SCSI controllers with good DMA support is great.
Anyway, food for thought…
I love it when I post something simple and get a books worth of education
in the replies!
Great stuff… thanks!
Chris – Fast and Low on my '020!
…Is it significant that my spell checker doesn't know "Intel"?
Paul:
Are you saying that Intel made a greater leap from '486 to Pentium
tham Motorola made from '040 to '060? I hadn't considered that. All the
preious moves were on a see-saw basis. All comparisons on an equal clock
speed of course.
regards, Lu on AutoPilot downunder
> Are you saying that Intel made a greater leap from '486 to Pentium
> tham Motorola made from '040 to '060? I hadn't considered that.
> All the preious moves were on a see-saw basis. All comparisons on
> an equal clock speed of course.
Comparing clock rate is a tempting path to confusion. While clocks speed can
provide information about the cost or complexity of a chip, it means very
little across different families. It was almost a coincidince that the 80×86
and 68K chips were similar speeds at similar clocks.
For example, some MIPS chips at 150Mhz are about as fast as some PowerPC chips
at 80Mhz. They use different strategy. MIPS uses superpipelining, while the
PowerPC aims at superscalar. Two equally good ways to reach a speed goal.
There is also the false illusion of equality using clocks. This is because
Intel became very aggressive with the clocks starting with the 486 so the chips
became very quick. So while equal at same clock, many people were running
66Mhz 486's! Very depressing.
The 68K chips were always more elegant, had fewer bugs (ask me about the first
386's!), and operated cleaner, but Moto seemed to have trouble delivering them
starting with the '040 and really had trouble with higher clocks. I suspect
that with CISC getting so complex, Intels deeper pockets helped make the
difference.
Now with PowerPC the shoe is on the other foot. Moto/IBM/Apple is coming out
with PowerPC chips several times faster than Intel. They also look to be
getting more agressive with the clocks than Pentium.
Word on the street is that the Intels P6 may be "late", and server oriented.
The "Microprocessor Report" recently went so far as to speculate that with 604,
PowerPC will just keep pulling ahead of Intel. They even think the 604 may be
faster than the P6, and that 620 will be a whole new level of performance. It
will be fun to be aboard for the ride.
—— Andrew
Thanks for the info – I'm getting lost in the maze, the media out here
said that PowerPC was the Moto name for the 68050 in line with Intel
calling theirs the Pentium. Then in this forum it was said that Moto
jumped from the 68040 to the 68060 so I was assuming this was aka PowerPC.
From what you say, it's a different series. C= apparently then have
never considered it.
regards, Lu on AutoPilot downunder
>From what you say, it's a different series. C= apparently then have
never considered it.
Although C= wasn't planning on using the PowerPC chip they did have plans for
their own RISC chip that they were working on with … I forget now, but does
it matter?
Vree in VA
Paul –
It's been my impression that the Motorola series have been released
somewhere about halfway between Intel generations (and vice versa). And
that the Motorola release has generally been more competent (faster,
whetever) than the last previous Intel release. So I think it went:
8086 80286 80386 80486 Pentium
68000 68020 68030 68040 68060
. . . . JimDoc, on AutoPilot
Jim-
Yah, that's about right, but until the 486 Moto was equal or faster in
clocks. Now, the 060 is late and its clock is slow. 🙁
Paul
> It's been my impression that the Motorola series have been released
> somewhere about halfway between Intel generations (and vice versa).
> And that the Motorola release has generally been more competent
> (faster, whetever) than the last previous Intel release. So I think
> it went:
> 8086 80286 80386 80486 Pentium
> 68000 68020 68030 68040 68060
Vaugely true. Each family had the best that technology had to offer at that
date, with 68K being "cleaner". The problem is that at the 80486/68040
junction, Intel started to really get aggressive with the clocks. Add in the
fact that the '040 was late and Intel ended up pulling ahead with higher
clocks.
At the same clock the '040 was nicer than the '486, but the '040 has not even
come close to getting to 66Mhz, much less 100Mhz.
The '060 is not really ahead of the Pentium. It's cheaper (1/2 the price), but
is very close in overall speed at similar clocks. It's the '040 story all over
again! The '060 is _late_ and has a very slow clock. 50Mhz, that's not much
to brag about.
90Mhz P54C chips from Intel are out now, 100Mhz by the end of the year. '060
will barely get to 66Mhz by then. PowerPC601 should be near 135Mhz by years
end.
As usual, they have a killer chip a year late. 🙁
Let's hope that at least PowerPC makes a dent in the Intel empire. Just once,
I want a "clean" architecture to win.
—— Andrew
Andrew,
A clean architecture, right, when has that ever happened? My hopes, too.
Robb