PII vs PIII

[Gregory L. Hansen]

It looks like the machine I have (an HP Kayak XAS, you may have seen me
mention it already) can be upgraded from a PII 400MHz to dual PIII 600MHz.
But the PIII's are substantially more expensive than the PII's, around $70
each compared to $10 each.

I guess I'm not that excited about a 50% increase in clock rate, by
itself. But is there a great advantage just in going from a PII to a PIII
with a comparable clock speed? Would I get substantially more computing
for the money?

 

[Christopher Browne]

Martha Stewart called it a Good Thing (e-mail address removed) (Gregory L. Hansen)wrote:

It looks like the machine I have (an HP Kayak XAS, you may have seen
me mention it already) can be upgraded from a PII 400MHz to dual
PIII 600MHz. But the PIII's are substantially more expensive than
the PII's, around $70 each compared to $10 each.

I guess I'm not that excited about a 50% increase in clock rate, by
itself. But is there a great advantage just in going from a PII to
a PIII with a comparable clock speed? Would I get substantially
more computing for the money?

The bigger win there is liable to be in moving to faster and more
copious amounts of memory that you can get cheaply.

Jumping to even newer generations has the merit that you can get 512MB
sticks of memory very cheaply. I would expect _that_ to, by and
large, be the dominant reason to prefer fancy new architectures.
--
select 'cbbrowne' || '@' || 'acm.org';
http://www.ntlug.org/~cbbrowne/lsf.html
"La Cicciolina [...] Electing her was an interesting contrast to the
situation in the UK: In Italy they elect a representative from the sex
industry. In the UK, they elect their clients." -- Peter Gutmann

 

[Steve Wolfe]

It looks like the machine I have (an HP Kayak XAS, you may have seen me

mention it already) can be upgraded from a PII 400MHz to dual PIII 600MHz.
But the PIII's are substantially more expensive than the PII's, around $70
each compared to $10 each.

I guess I'm not that excited about a 50% increase in clock rate, by
itself. But is there a great advantage just in going from a PII to a PIII
with a comparable clock speed? Would I get substantially more computing
for the money?

There *can* be an advantage, but whether it's important to you is
debatable. Early P3's had the same cache architecture as the P2's -
namely, external cache running at half of the CPU's clock speed. Starting
around the 600 MHz mark, you were able to get P3's with on-die cache
running at full CPU speed.

So, if cache latency is a deal-breaker for your application, the extra
money would be worth it. If not, then it's a lot more of a
personal-preference type of thing.

As a side note, I'm surprised that a P3/600 would be $70, seeing that
you can buy something like an Athlon 2400+ for less than that. I looked
over on ebay, and found the P3's running $20 to $30 each.

steve

 

[Gregory L. Hansen]

There *can* be an advantage, but whether it's important to you is
debatable. Early P3's had the same cache architecture as the P2's -
namely, external cache running at half of the CPU's clock speed. Starting
around the 600 MHz mark, you were able to get P3's with on-die cache
running at full CPU speed.

So, if cache latency is a deal-breaker for your application, the extra
money would be worth it. If not, then it's a lot more of a
personal-preference type of thing.

As a side note, I'm surprised that a P3/600 would be $70, seeing that
you can buy something like an Athlon 2400+ for less than that. I looked
over on ebay, and found the P3's running $20 to $30 each.

There's a wide range of prices, so I tried to generalize, and I was
looking specifically at PIII 600MHz 512 cache. I was thinking especially
of a matched pair I saw for something over $100, although they had 256K
cache.

I suppose I should make sure I can get the system up and running in the
first place, before I start worrying about upgrading it. I'm still
waiting for some cables and adapters to arrive so I can plug it into my
monitor, so I still can't see what I'm doing and it's just sitting there
with a clean hard drive right now.

 

[Mike Walsh]

When Intel went to the on chip L2 cache it was not to increase performance, it was to save money. They cut the size of the cache from 512 Kb to 256 Kb. My old Pentium 2 L2 cache tested at half of the L1 cache speed, which it is supposed to be. With my Pentium 3 (1.26 Ghz with 512 Kb L2 cache) the L2 cache tests at 60% of L1 cache speed, not 100% as Intel would have you believe.

 

[Trevor Hemsley]

On Fri, 10 Oct 2003 17:34:03 UTC in comp.os.linux.hardware,

There's a wide range of prices, so I tried to generalize, and I was
looking specifically at PIII 600MHz 512 cache. I was thinking especially
of a matched pair I saw for something over $100, although they had 256K
cache.

The fact that the top speed for this board is 600MHz leads me to think
that it will only work with the "Katmai" flavour of P-III not with its
successor the "coppermine". The Coppermine was the one where they
halved the amount of L2 cache on the basis that they made it full
speed at the same time. These chips require a lower core voltage than
the Katmai ones and some motherboards don't have voltage regulators
that are capable of supplying the correct one. So, if it says top
speed is 600MHz they probably mean that you're limited to the older,
512KB cache, P-III's.

 

[Steve Wolfe]

The fact that the top speed for this board is 600MHz leads me to think

that it will only work with the "Katmai" flavour of P-III not with its
successor the "coppermine". The Coppermine was the one where they
halved the amount of L2 cache on the basis that they made it full
speed at the same time. These chips require a lower core voltage than
the Katmai ones and some motherboards don't have voltage regulators
that are capable of supplying the correct one. So, if it says top
speed is 600MHz they probably mean that you're limited to the older,
512KB cache, P-III's.

It still might pay to check if it will support them or not - there are a
number of boards which were originally designed for P2's which also had
the flexibility to support the P3 Coppermines - my home machine, with a
650 MHz P3, is running on such a motherboard.

steve

 

[Steve Wolfe]

When Intel went to the on chip L2 cache it was not to increase
performance, it was to save money. They cut the size of the cache from 512
Kb to 256 Kb. My old Pentium 2 L2 cache tested at half of the L1 cache
speed, which it is supposed to be. With my Pentium 3 (1.26 Ghz with 512 Kb
L2 cache) the L2 cache tests at 60% of L1 cache speed, not 100% as Intel
would have you believe.

The *frequency* at which the cache works in the on-die chips is actually
100% of clock speed. However, as you point out, that doesn't work out to
100% performance, as there are a lot of factors involved. However, in
apps where cache latency is a factor, the chips with on-die cache do come
out ahead.

Perhaps the largest limitation to the P3 line is it's limitted FSB, at
133 MHz. I've used a good number of dual-p3 servers, and found that once
the CPU gets up to about 866, you've got pretty much all the performance
you're going to get - a faster CPU doesn't do anything for you. Once you
start talking about the "-S" chips (the later P3's with 512k on-die cache,
and clock speeds of 1.13 to 1.4 GHz), you get an initial performance boost
from the doubling of the cache, but little to no returns from increasing
the CPU speed along the 1.13- to 1.4-GHz line.

steve

 

[Gregory L. Hansen]

The fact that the top speed for this board is 600MHz leads me to think
that it will only work with the "Katmai" flavour of P-III not with its
successor the "coppermine". The Coppermine was the one where they
halved the amount of L2 cache on the basis that they made it full
speed at the same time. These chips require a lower core voltage than
the Katmai ones and some motherboards don't have voltage regulators
that are capable of supplying the correct one. So, if it says top
speed is 600MHz they probably mean that you're limited to the older,
512KB cache, P-III's.

I didn't realize there was such a difference between a PIII 600MHz 512K
cache and a PIII 600MHz 256K cache. Are all PIIIs with 512K Katmais, and
all PIIIs with 256K coppermines?

When I asked about the VRM, I was told the part number 0950-2837 was for
any PII/PIII up to 600MHz, and they specifically said it's not for
"coppermine". I've asked if the machine would support a faster PIII if a
different VRM were installed, but haven't gotten an answer yet, and I'm
beginning to wonder if I will. I think the motherboard has the 440BX
chipset, if that makes a difference, but I know it also matters which
motherboard the chipset is sitting on. And maybe a BIOS upgrade, which HP
may or may not have, and which I've never done.

The more I learn about this, the more it all gets complicated by little
bits of information like what you've just said above.

 

[TCS]

It looks like the machine I have (an HP Kayak XAS, you may have seen me
mention it already) can be upgraded from a PII 400MHz to dual PIII 600MHz.
But the PIII's are substantially more expensive than the PII's, around $70
each compared to $10 each.

I guess I'm not that excited about a 50% increase in clock rate, by
itself. But is there a great advantage just in going from a PII to a PIII
with a comparable clock speed? Would I get substantially more computing
for the money?

For $140 you can get a new case that'll accept standard motherboards
and a motherboard'n'cpu on ebay that'll be twice as fast as a pair of
P3-600s.

 

[Lane Lewis]

It looks like the machine I have (an HP Kayak XAS, you may have seen me
mention it already) can be upgraded from a PII 400MHz to dual PIII 600MHz.
But the PIII's are substantially more expensive than the PII's, around $70
each compared to $10 each.

I guess I'm not that excited about a 50% increase in clock rate, by
itself. But is there a great advantage just in going from a PII to a PIII
with a comparable clock speed? Would I get substantially more computing
for the money?

No advantage at all with today's software, save your money towards a P4
or Athlon system. One other alternative if you have a lot of sdram memory
already is to go for a 1.4 gig celeron and a motherboard that takes Sdram.
Should be able to put a quality system together for under 150 that way.

Lane

 

[Steve Wolfe]

No advantage at all with today's software, save your money towards a
P4

or Athlon system. One other alternative if you have a lot of sdram memory
already is to go for a 1.4 gig celeron and a motherboard that takes Sdram.
Should be able to put a quality system together for under 150 that way.

But once you've used a dual-processer system, it's hard to go back. ; )

I have a dual Pentium 133 that is still *very* useable as a desktop, be
it under Linux or NT. A Pentium 233 would not be as usable. Remember
that CPU power is not the only componant in usability. In fact, in modern
processers, CPU power becomes a much smaller factor in the usability. In
fact, given the choice for a desktop system where I wasn't going to play
3D games, I'm much prefer a dual P3/600 over a 1.4 celeron.

You can take a single-CPU system and flood the processer with
interrupts, and while the processer does very little actual work, the
machine will slow to a crawl. The same is not true of an SMP system,
where one processer can be handling interrupts, and the other going on
about it's job of handling GUI, kernel, and other code. In a
single-processer machine with a gigabit network connection and no
interrupt coallescing, it's possible to bring out that sort of condition
by simply sending the machine lots of network data, without bringing any
sort of processing into the mix!

steve

 

[Skylar Thompson]

It looks like the machine I have (an HP Kayak XAS, you may have seen me
mention it already) can be upgraded from a PII 400MHz to dual PIII 600MHz.
But the PIII's are substantially more expensive than the PII's, around $70
each compared to $10 each.

I guess I'm not that excited about a 50% increase in clock rate, by
itself. But is there a great advantage just in going from a PII to a PIII
with a comparable clock speed? Would I get substantially more computing
for the money?

The P-III has more multimedia instructions, so it's a win if you're doing
multimedia applications (graphics, movies, etc.).

 

[kony]

On Fri, 10 Oct 2003 15:22:34 +0000 (UTC),

It looks like the machine I have (an HP Kayak XAS, you may have seen me
mention it already) can be upgraded from a PII 400MHz to dual PIII 600MHz.
But the PIII's are substantially more expensive than the PII's, around $70
each compared to $10 each.

I guess I'm not that excited about a 50% increase in clock rate, by
itself. But is there a great advantage just in going from a PII to a PIII
with a comparable clock speed? Would I get substantially more computing
for the money?

If it will accept a standard motherboard, you'd get better, more cost
effect results upgrading the motherboard, running a single Tualatin,
Athlon, or P4 CPU.


Dave

 

[Lane Lewis]

But once you've used a dual-processer system, it's hard to go back. ; )
nonsense deleted snip

steve

nonsense

So your trying to tell me that dual 600 system is better than a single 1.4
gig system when every benchmark in the world shows a huge gap between the
two. When every game plays twice as fast on the celeron, when photoshop,
excel, word, music software etc perform at least 40 to 60 percent better
maybe even more. Do you have anything at all that would back up what you
just said.

Lane

 

[~misfit~]

I didn't realize there was such a difference between a PIII 600MHz 512K
cache and a PIII 600MHz 256K cache. Are all PIIIs with 512K Katmais, and
all PIIIs with 256K coppermines?

Yep. Default core voltage for Katmai is 2.0v and for early coppermine 1.7v.

When I asked about the VRM, I was told the part number 0950-2837 was for
any PII/PIII up to 600MHz, and they specifically said it's not for
"coppermine". I've asked if the machine would support a faster PIII if a
different VRM were installed, but haven't gotten an answer yet, and I'm
beginning to wonder if I will. I think the motherboard has the 440BX
chipset, if that makes a difference, but I know it also matters which
motherboard the chipset is sitting on. And maybe a BIOS upgrade, which HP

may or may not have, and which I've never done.

Some BX boards handle coppermines fine. I have three BXs here running
coppermine CPUs. Some boards require a BIOS upgrade to enable the lower
vcore/different instuction set/microcode, some boards just can't handle
coppermines.

The more I learn about this, the more it all gets complicated by little
bits of information like what you've just said above.

It's all good fun though. I have several clone machines that started life as
PII350s that I got cheaply at auction and now are running various CPUs. The
best of them is a machine I am running as an internet gateway/file server on
our home LAN. It has a coppermine celeron 600 running at 927MHz with just a
slight core voltage increase running in a slightly modified 'slocket' with
standard heatsink. It benchmarks better than a PIII850. A bit af a waste of
a CPU really, the only difference I notice between it and the PII350 it
originally had is SETI work units are done in less than 10 hours now when
they took 18+ hours with the PII. My main machine is an Athlon XP 2200+
(1800MHz) that does work units in 4 hours.

As I said, it's all good fun.

 

[Trevor Hemsley]

On Fri, 10 Oct 2003 18:31:51 UTC in comp.os.linux.hardware,

Are all PIIIs with 512K Katmais, and
all PIIIs with 256K coppermines?

I think all P-II and P-III processors prior to coppermine had 512KB,
half speed cache.

beginning to wonder if I will. I think the motherboard has the 440BX
chipset, if that makes a difference

One of the better Intel chipsets.

 

[Gregory L. Hansen]

nonsense

So your trying to tell me that dual 600 system is better than a single 1.4
gig system when every benchmark in the world shows a huge gap between the
two. When every game plays twice as fast on the celeron, when photoshop,
excel, word, music software etc perform at least 40 to 60 percent better
maybe even more. Do you have anything at all that would back up what you
just said.

You must have missed the part where he said he wasn't going to play 3D
games with it.

I understand what he's saying about it. Try doing something in NT while
another program is loading, for instance. The mouse pointer jumps around,
things get choppy. A dual processor machine might not be faster, but you
don't get that sort of business because there's a second processor to
handle the GUI while the other handles interrupts. And I actually
wouldn't be surprised if the dual 600MHz *is* faster than the single 1.4
GHz if your benchmark involves doing more than one thing at a time.

 

[Lane Lewis]

You must have missed the part where he said he wasn't going to play 3D
games with it.

I understand what he's saying about it. Try doing something in NT while
another program is loading, for instance. The mouse pointer jumps around,
things get choppy. A dual processor machine might not be faster, but you
don't get that sort of business because there's a second processor to
handle the GUI while the other handles interrupts. And I actually
wouldn't be surprised if the dual 600MHz *is* faster than the single 1.4
GHz if your benchmark involves doing more than one thing at a time.

And more nonsense
Again do you have anything that shows that they are even remotely close
in benchmarks even while multitasking. Dual machines make good servers and
ok single app workstations but they are terrible desktop machines. The vast
majority of software is designed for single processors and are virtually
worthless on a dual machine and it's getting worse every day as software
developers code for the P4 single CPU. To sit there and wait for a program
that can't take advantage of the extra CPU doesn't make any sense with
today's prices. One thing that should be mentioned is the overhead that a
dual system has compared to single CPU which practically negates any
advantage.

Lane

 

[Steve Wolfe]

So your trying to tell me that dual 600 system is better than a single 1.4

gig system when every benchmark in the world shows a huge gap between the
two.

You're missing the point. I'm not talking about benchmarks. I'm talking
about usability.

Have YOU used a dual-CPU desktop? I didn't think so. Now go back home
and play with your toys.

steve