DX4ODP75 & 33MHz Base Clock

[w***@hotmail.com]

Will an Intel DX4ODP75 work on a 33MHz base clock? I'm guessing that it
will run at 100MHz, but I'm not sure...

William the Curious

[w***@hotmail.com]

Post by w***@hotmail.com
Will an Intel DX4ODP75 work on a 33MHz base clock?

I guess so. It's sitting at 99.8MHz (!!) right now in Windows NT and
has been doing so for the past ~2 hours. It took about half an hour of
diddling around to unsquash all the pins on this CPU. I must be getting
quite good at it, though, because it fit easily into the LIF socket on
the first try.

Perhaps Intel was labeling parts however they saw fit--and not in line
with the actual specs--once again.

William

[David L. Beem]

Hi William,

Post by w***@hotmail.com
Will an Intel DX4ODP75 work on a 33MHz base clock?
I'm guessing that it will run at 100MHz, but I'm not sure...

Yes, it will run (or try to) at 100MHz. I would think that Intel
labeled DX4s to avoid confusion, but you might want to monitor heat
initially because they may have tested units & put the cooler running at
higher speeds. All-in-all there shouldn't be any damage to see if it works.
David
***@IBMMuseum.com

[w***@hotmail.com]

Hi!

...but you might want to monitor heat initially because they may
have tested units & put the cooler running at higher speeds.

Interestingly enough, this chip runs cooler in regular use and at idle
than any of the "proper" 100MHz Intel ODP CPUs I have.

Running distributed.net on it resulted in things warming up, but at no
point did the CPU become abnormally hot.

William

[IBMMuseum]

Hi William,

...Interestingly enough, this chip runs cooler in regular use and
at idle than any of the "proper" 100MHz Intel ODP CPUs I have...

S-spec?...
David
***@IBMMuseum.com

[w***@hotmail.com]

Hi!

Post by IBMMuseum
S-spec?...

Appears to be SZ923. The copyright on the chip reads both 1989 and 1993
and has a "modified" symbol as well.

Jan Steunebrink's CHKCPU for DOS 1.12 says the following:
(cleaned up somewhat for proper display without extended ASCII
characters)

CPU Vendor and Model: Intel -Unknown model-
Internal CPU speed : 99.8 MHz
Clock Multiplier : ---
CPU-ID Vendor string: GenuineIntel
CPU-ID Signature : 1480

Current CPU mode : Real
Internal (L1) cache : Enabled in Write-Through mode

The "1480" breaks down as follows:
1 = Overdrive type
4 = Family 4 (486)
8 = Model (indicates CPU model and 486 cache mode)
0 = Stepping or sub-model number

The CPU has run all night long under 100% load without issue.

William

[David L. Beem]

Hi William,

Post by w***@hotmail.com

Post by IBMMuseum
S-spec?...

Appears to be SZ923. The copyright on the chip reads both
1989 and 1993 and has a "modified" symbol as well...
...CPU-ID Signature : 1480

So another rare 1480h CPUID & another one for my chart (should show it
is the "1.0" version in print too). SZ925 (ODPR75) & SZ926 (ODPR100) are the
only others I have seen with 1480h. Somehow I think that the missing piece
of the set (ODP100) will be S-spec SZ924.
David
***@IBMMuseum.com

[William R. Walsh]

Hi!

Post by David L. Beem
So another rare 1480h CPUID & another one for my chart (should show it
is the "1.0" version in print too).

Rare? Any idea how or why, other than the 75MHz rating? (I can't ever
remember seeing the 75MHz 486 in great numbers even when it was new. People
certainly were selling them...but how many would have actually purchased one
instead of upclocking their motherboard and going that much further with the
100MHz part?)

At least it was worth the effort to straighten all the pins. This poor CPU's
pins were almost completely flattened. I thought sure a few would even break
off as I attempted to fix them.

I've forgotten the link to your chart of 486 CPU IDs.

Post by David L. Beem
Somehow I think that the missing piece of the set (ODP100) will be S-spec

SZ924.

Let me look around. I have a couple of "ODP" 486-100 CPUs and maybe one of
them is it. The 75MHz CPU we've been discussing was purchased along with a
100MHz one that is in another Model 85.

William

[Count Floyd]

On Thu, 26 Oct 2006 04:27:50 UTC, "William R. Walsh"

Post by William R. Walsh
Hi!

Post by David L. Beem
So another rare 1480h CPUID & another one for my chart (should show it
is the "1.0" version in print too).

Rare? Any idea how or why, other than the 75MHz rating? (I can't ever
remember seeing the 75MHz 486 in great numbers even when it was new. People
certainly were selling them...but how many would have actually purchased one
instead of upclocking their motherboard and going that much further with the
100MHz part?)

Post by David L. Beem
William

I remember having one, it was used on a PS/ValuePoint SX-25 and it was
advertised as a DX4-75, to be used with a 25mHz chip. The DX100 was
advertised to be used with a 33mHz chip, go figure. It really didn't
run that much faster however.

[Robert E. Watts]

Hi William !

( I know this is a little late, but....... )

Yes.

You might even try it at 40MHz. I know that the POD83MHz will run just fine
at 40MHz, resulting in a POD 100MHz. I have done that lots of times.

I have used a POD63 at 33MHz lots of times also. And 40MHz. Never a
problem.

( And then of course there is the famous AMD ADZ-133 running at 160MHz,
40MHz bus speed. )

I know that (years ago, or so the story goes) Intel sometimes sells chips
that were supposed to be a 33MHz CPU, but the yield was bad, so they simply
sold them at 25MHz.

But I would theorize that they sold upgrade CPU's to people who really
didn't have a clue about setting the bus speed, so they sold 486 75MHz,
100MHz, etc. chips that were really all the same.
--
boBWatts®©
EartH
Watts Carburetion Service
Whizzbang Computers
Official collector of: transfat asian plastic junk trinkets !

Post by w***@hotmail.com
Will an Intel DX4ODP75 work on a 33MHz base clock? I'm guessing that it
will run at 100MHz, but I'm not sure...
William the Curious

[William R. Walsh]

Hi!

Post by Robert E. Watts
You might even try it at 40MHz. I know that the POD83MHz will run just
fine at 40MHz, resulting in a POD 100MHz. I have done that lots of times.

If this were a clone, I'd consider such a thing. This CPU is in a 9585-0X6
and...well...I'm just not going to mess with the magical blue power inside.
I like my 9585 "X" machines, and the thought of killing one while doing
something that has the potential to damage hardware (the "goose with the
golden egg" principle) in search of a little more performance just isn't a
good thought.

Post by Robert E. Watts
But I would theorize that they sold upgrade CPU's to people who really
didn't have a clue about setting the bus speed, so they sold 486 75MHz,
100MHz, etc. chips that were really all the same.

I think you've got something there. Why go to the trouble of engineering a
75 (or, conversely, a 100) MHz part when one could fill the role of both.
And you get to charge more for the "faster" one.

William

[Robert E. Watts]

Hi William !

Post by William R. Walsh
Hi!
If this were a clone, I'd consider such a thing. This CPU is in a 9585-0X6
and...well...I'm just not going to mess with the magical blue power inside.
I like my 9585 "X" machines, and the thought of killing one while doing
something that has the potential to damage hardware (the "goose with the
golden egg" principle) in search of a little more performance just isn't a
good thought.

I agree. As soon as I posted that message, I realized that I was posting to
the IBM PS/2 NG, and it was probably not an option. You just can't go
a'changin' bus speed as easily as a clone for one thing, and the risk of
damaging stuff isn't worth it.

bob