Collectors Guide for Vintage Intel Microchips... It's Done!

[pphillips12]

I'm really interesting in seeing what you think about the guide. Maybe you could help me figure out how Intel computes power dissipation!

Some data sheets list mW, some list mA. Sometimes the mW listed is
different depending on which data sheet you look at. One year it's one value and then the data sheet for exact same chip has a slightly different value the following year. Everytime I think I have figured out how they
are computing power dissipation, I find a data sheet where I can't make heads or tails at how they came up with what's listed on the data sheet.

Something else that baffles me is why does Intel sometimes list different power dissipations for the military versions of a chip? I thought the only difference between say an C1702A and a MC1702A was the testing procedure yet the data sheets list a different mW rating for the two chips. Is ithe power dissipation somehow related to the rated operating temp?

[gshv]

[pphillips12]

Thanks. Do you know exactly how the temp factors into the formula used to compute the power dissipation? So what you are saying is that the power dissipation is higher on the milspec chips because they are showing the power dissipation at the upper range of the 10% variation in supply voltage as opposed to the non-milspec chips 5%???

[gshv]

[CPUShack]

[gshv]

[CPUShack]

[CPUShack]

[pphillips12]

I take good care of my vintage processors but I've been running my P4 2.4C @ 3.25Ghz for going on two years and I'm running it slightly overvolted as well. I'm running 1 GB of Mushkin L2 PC 3500 way overvolted too. My ATI 9800XT's overclocked as well.

I usually build a new machine every few years so I really don't care
about shortening the lifespan of the components. They'll all be obsolete
in a few years anyway. So basically I just crank everything up until
it starts to fail prime95 and memtest and then back off a little and run
it like that day in and day out.

I haven't killed a processor yet but I'm sure my day is coming:)

[ahia123]

Hello George,

You mentioned Prime95, are you a member of the GIMPS group testing Mersenne numbers? I've been a member for a long time (been in and out of the top 100 list a number of times). I hope George Woltman will look into the Cell architecture and see if there are any possibility of speeding up the project (even though the performance is quite good at the moment).

Bjørn

[pphillips12]

Nope. Not a member. I just use it as a stability test when I'm overclocking. Besides, primes drive me crazy. Deep in my heart
I know there's got to be a way to predict them:)

They're almost as irritating as the traveling salesman problem
which will truely drive someone insane.

The TSP is like trying to climb to the highest part of a tree.
The problem is you can't see the top from the ground for the leaves. Which branch do you climb? The logical choice would be the one that
goes the highest before the leaves obscure your view. So you climb
that branch but when you emerge thru the leaves you see that branch ends but you can now other branches that go much higher.

In other words, the best choices can ultimately lead you down
the wrong path.

A word of warning about the TSP. One of my math professors many years ago told me to never try and solve the TSP. It will suck you
in and the next thing you know you'll have spent your entire life
looking for a solution;)

[pphillips12]

Preview Copies were mailed today.

Those of you getting a copy, make sure and read the readme file
before trying to use the guide. The readme file says:

You'll need to make sure you are using Adobe Reader 5.0 or later
AND you'll need to make sure that your Adobe preferences are
set to open cross-links in a new window.

Otherwise, every time you click a link, the current window will
be closed which will make it very difficult to nagivate using
the links.

Enjoy,

George

[CPUShack]

ahh sounds good.

could make that an auto run feature in the release?

[willsmith]

I just received my chip collectors CD yesterday and it's an excellent resource.

A couple of suggestions for the next edition:

it's hard to differentiate between the variants of the chips. E.g. for the C4004, there are 7 variants shown. For ease of reference, maybe the variants could be numbered? Since this is THE DEFINITIVE guide, make up the numbering yourself, George!

Also I think a better pricing estimate is needed. Perhaps a few people could volunteer to 'maintain' the prices for different chips. E.g I could supply details of 1103's (my personal interest) and many of the 4004's sold over the last 2 years. Unfortunately I did not accurately record the condition of each, just the price and date. Then they could periodically mail in prices in the format of George's choosing, with URLs to back up the data. Maybe we could even somehow build a public pricing database, with an 'ebay mirror', since ebay annoyingly removes auction details quickly.

Regarding the spread of data, I would personally suggest more info on the most valuable and widely traded / widely sought after chips (4004's, 4040's, etc).

One last thing George. Since we will increasingly rely on your excellent reference guide, what happens if something should happen to you? Maybe you can make a will and in it, assign the copyright to somebody else who you trust (or make it creative commons or something). Sorry to be morbid

Overall, an excellent buy, I'm very pleased to own a copy, thanks George for your major step in 'formalizing' our hobby. I hope the money you receive from the CD sales helps to compensate for the huge amount of time you've obviously spent compiling this data. I guess you'll be receiving lots of comments on the first edition - don't be afraid to bring out a 2nd edition in say the Autumn!

Thanks
Will Smith

[pphillips12]

I just received my chip collectors CD yesterday and it's an excellent resource.

A couple of suggestions for the next edition:

it's hard to differentiate between the variants of the chips. E.g. for the C4004, there are 7 variants shown. For ease of reference, maybe the variants could be numbered? Since this is THE DEFINITIVE guide, make up the numbering yourself, George!

>> I've thought about that but have not been able to come up with any >> numbering scheme I'm happy with. Currently you can ref. chips in
>> the guide as C4004, variant #3, etc.

Also I think a better pricing estimate is needed. Perhaps a few people could volunteer to 'maintain' the prices for different chips. E.g I could supply details of 1103's (my personal interest) and many of the 4004's sold over the last 2 years. Unfortunately I did not accurately record the condition of each, just the price and date. Then they could periodically mail in prices in the format of George's choosing, with URLs to back up the data. Maybe we could even somehow build a public pricing database, with an 'ebay mirror', since ebay annoyingly removes auction details quickly.

>> The values listed in the 1st edition should be taken with a grain of salt.
>> It's hard to pin down exact values when the same chip sells on EBAY
>> week for $20 and the next week for $100:) That said, the values are
>> not purely based on EBAY. I also take into account how many of
>> a particular chip is still in stock.

Regarding the spread of data, I would personally suggest more info on the most valuable and widely traded / widely sought after chips (4004's, 4040's, etc).

>> I'll be adding more historical tidbits in the 2006 edition. That said,
>> I'm trying to stay away from general history of Intel type stuff
>> (There's already books on that) and concentrate on the chips
>> themselves.

One last thing George. Since we will increasingly rely on your excellent reference guide, what happens if something should happen to you? Maybe you can make a will and in it, assign the copyright to somebody else who you trust (or make it creative commons or something). Sorry to be morbid

>> Strange you should mention that. The week before I put the guide
>> up for sale I came down with a stomach virus and the only
>> thing I could think about was that I was going to die and my book
>> would never be released. Since I felt like I was going to die at the
>> time, it seemed like a logical concern. I actually crawled out of bed,
>> burned several current copies and made my wife promise she would
>> get copies made and put it on EBAY so all the work would not be lost.
>> She assured me that I was not going to die and fortunately she was
>> right.

>> But you are right, I'll give this some thought.

Overall, an excellent buy, I'm very pleased to own a copy, thanks George for your major step in 'formalizing' our hobby. I hope the money you receive from the CD sales helps to compensate for the huge amount of time you've obviously spent compiling this data. I guess you'll be receiving lots of comments on the first edition - don't be afraid to bring out a 2nd edition in say the Autumn!

>> I've only sold about 30 copies to date so the 2005 guide may end up
>> being more rare than many of the chips inside it! I am already
>> working on the 2nd edition for 2006. I just found some extremely
>> rare original data sheets for the 1402/1403/1404 and 1405 shift
>> registers printed in 1970. I'm tightening up the introduction dates
>> on a number of chips, added new photos, adjusting values, etc.

Thanks
Will Smith