seems like a misunderstanding. It wasn't released until dec 20 1999. So there is no reasonable expectation of why you could find anyone having one in 1999 unless they got it as a Christmas present or something. Seems insane to me. But there is plenty of evidence that Anand Lal Shimpi had them in his hands and tested them pre launch. Which is fairly normal for hardware review.

I don't see why it matters.

According to this page, it was pushed from end of 1999 to 1st quarter of 2000:
https://www.theregister.co.uk/1999/11/17/pent … rought_forward/

Not even listed in the Intel quick processor reference:
https://www.intel.com/pressroom/kits/quickrefyr.htm

All the reviews I am seeing are March 2000 or later.

Warlord wrote:

I don't see why it matters.

It matters for "period-correct" builds. I'm in a similar boat as I'm currently trying to figure out the date when my motherboard hit the market. It's an Intel Advanced/ATX aka Thor. It's significant for me, because it's the original ATX board:

- Intel introduced the ATX standard in July 1995;
- Intel started showing reference boards to system builders in August 1995;
- there is no mention of the board on Intel's site (seems like Intel only put up some info retroactively around 1999). Perhaps, Intel figured that since people couldn't buy the board in retail, they should get support from their system builders;
- a search through newsgroups shows that people first got their hands on Thor in early 1996 at best with systems from Gateway and others.
- the board I got was originally bought in December 1997;

Whether or not you should obsess so much over dates is up to debate, but yeah, "period-correctness" is a real rabbit hole.

Warlord wrote:

seems like a misunderstanding. It wasn't released until dec 20 1999. So there is no reasonable expectation of why you could find anyone having one in 1999 unless they got it as a Christmas present or something. Seems insane to me. But there is plenty of evidence that Anand Lal Shimpi had them in his hands and tested them pre launch. Which is fairly normal for hardware review.

I don't see why it matters.

I remember that in the Quake3 timedemo thead someone showed off a receipt for a Geforce 256 DDR which they bought right before christmas 1999, even though there are no reviews for one before Y2K.
This made the card eligible for a period-correct 1999 build. Same reason why evidence for this particular CPUs release date is needed.

I always thought "period-correctness" means more "mainstream during that period" than "elite latest model but could be bought if you're a millionaire"

Munx wrote:

Warlord wrote:

seems like a misunderstanding. It wasn't released until dec 20 1999. So there is no reasonable expectation of why you could find anyone having one in 1999 unless they got it as a Christmas present or something. Seems insane to me. But there is plenty of evidence that Anand Lal Shimpi had them in his hands and tested them pre launch. Which is fairly normal for hardware review.

I don't see why it matters.

I remember that in the Quake3 timedemo thead someone showed off a receipt for a Geforce 256 DDR which they bought right before christmas 1999, even though there are no reviews for one before Y2K.
This made the card eligible for a period-correct 1999 build. Same reason why evidence for this particular CPUs release date is needed.

1999 is the same time period as 2000. I am almost 40 years old I lived then.

If it was available for sale new at that point in time, it would technically be "period correct".

Maybe we need some more categories:

1. Low to just below Mainstream
2. Mainstream
3. Mainstream to High End
4. Pipe dream for almost everybody

Now everybody needs 4 or more builds for every period.

jheronimus wrote:

It matters for "period-correct" builds.

What has the release date of something to do with "period-correct" builds? When you bought a PC back then, did it only contain parts that were all released in the same month of the same year? Of course they were not. A period-correct build is made of parts that were common and available back then. That's it.

Period correct 1999/2000 is pretty stupid by the way. 75% of the retro computers that you can find fall into 1999-2000 category or at very least the OS, Motherboard, CPU, etc.

Unless your computer is 1999 than you cannot run 98se because that wouldn't be correct since 98se is 1999. 🤣

One reaction to summarize the circle-argumentation of this thread..:

piatd wrote:

I'm looking for evidence that the PIII 800 (either E or EB versions) was available for purchase and use by an individual in 1999, either as a component or part of a system/bundle.

The EB is coppermine and according to Intel the EB (133mhz version):
Processor Base Frequency 1.00 GHz
Cache 256 KB L2 Cache
Bus Speed 133 MHz
Sockets Supported PPGA370, SECC2
Launch Date: Q1'00.
Which means it could have been as late as March before they released it to market or even slightly later. Looking for an OEM build with it would probably not have happened until summer or fall of 2000. HP released the Vectra VL400 with P3-1Ghz, 133bus after Nov 2000 from what little I can gather.. Here are some links descriping the HP
https://www.hp.com/ecomcat/hpcatalog/specs/P4151T.htm
https://www.cnet.com/products/hp-vectra-vl400 … or-none-series/

Some anecdotal evidence: I've been watching listings on eBay for a 99 800EB bound for my super-serious New Year's Eve 1999 $5000 build. Every one I've seen so far has had a year 2000 datecode. They may have been coming out the factory in December, but not in volume.

As a consequence I've adjusted down to a 733 Slot 1, which does seem to have been available by the end of the year on the basis of datecodes as far back as Week 41 1999 (lines up with Wikipedia's listed release date).

cyclone3d wrote:

Maybe we need some more categories: […]
Show full quote

Maybe we need some more categories:

1. Low to just below Mainstream
2. Mainstream
3. Mainstream to High End
4. Pipe dream for almost everybody

Yes, if you want to be immensely retentive about period-accuracy like me, then the 800 is Level 4 and the 733 is Level 3. 🤣

Big Pink wrote:

Some anecdotal evidence: I've been watching listings on eBay for a 99 800EB bound for my super-serious New Year's Eve 1999 $5000 […]
Show full quote

Some anecdotal evidence: I've been watching listings on eBay for a 99 800EB bound for my super-serious New Year's Eve 1999 $5000 build. Every one I've seen so far has had a year 2000 datecode. They may have been coming out the factory in December, but not in volume.

As a consequence I've adjusted down to a 733 Slot 1, which does seem to have been available by the end of the year on the basis of datecodes as far back as Week 41 1999 (lines up with Wikipedia's listed release date).

cyclone3d wrote:

Maybe we need some more categories: […]
Show full quote

Maybe we need some more categories:

1. Low to just below Mainstream
2. Mainstream
3. Mainstream to High End
4. Pipe dream for almost everybody

Yes, if you want to be immensely retentive about period-accuracy like me, then the 800 is Level 4 and the 733 is Level 3. 🤣

In my book, given the official dates, I consider my PIII-800 build to be the closest to the edge of the new millennium, CPU-wise. Of course they couldn't deliver a specimen from a batch before 2000 to everyone that wanted one...

cyclone3d wrote:

According to this page, it was pushed from end of 1999 to 1st quarter of 2000: https://www.theregister.co.uk/1999/11/17/pent … r […]
Show full quote

According to this page, it was pushed from end of 1999 to 1st quarter of 2000:
https://www.theregister.co.uk/1999/11/17/pent … rought_forward/

Not even listed in the Intel quick processor reference:
https://www.intel.com/pressroom/kits/quickrefyr.htm

All the reviews I am seeing are March 2000 or later.

Interesting links, thank you. Looking further at The Register, I also found this alleged letter from Intel to its partners, dated January 18, 2000:

"On December 20, 1999 Intel launched the Pentium III processors at 800MHz and 750MHz for desktop PCs. Initial shipments of the 750 MHz processor into the distribution channel (box and tray) are expected to begin in early January '00. Shipments of the 800MHz processor into the distribution channel (box and tray) are expected to begin later in Q1'00."

If true, this is the epitome of a paper launch. I'm assuming that the PIII 800 was a year 2000 part, until evidence shows otherwise. Reviewers possessing chips in 1999 is not evidence (they could be engineering samples, for all we know), nor is a wikipedia article with no pertinent citations.

piatd wrote:

Interesting links, thank you. Looking further at The Register, I also found this alleged letter from Intel to its partners, dat […]
Show full quote

cyclone3d wrote:

According to this page, it was pushed from end of 1999 to 1st quarter of 2000: https://www.theregister.co.uk/1999/11/17/pent … r […]
Show full quote

According to this page, it was pushed from end of 1999 to 1st quarter of 2000:
https://www.theregister.co.uk/1999/11/17/pent … rought_forward/

Not even listed in the Intel quick processor reference:
https://www.intel.com/pressroom/kits/quickrefyr.htm

All the reviews I am seeing are March 2000 or later.

Interesting links, thank you. Looking further at The Register, I also found this alleged letter from Intel to its partners, dated January 18, 2000:

"On December 20, 1999 Intel launched the Pentium III processors at 800MHz and 750MHz for desktop PCs. Initial shipments of the 750 MHz processor into the distribution channel (box and tray) are expected to begin in early January '00. Shipments of the 800MHz processor into the distribution channel (box and tray) are expected to begin later in Q1'00." If true, this is the epitome of a paper launch.

I'm writing off the PIII 800 as a year 2000 chip, until evidence shows otherwise. Reviewers possessing chips in 1999 evidence (they could be engineering samples, for all we know), nor is a wikipedia article with no pertinent citations.

I agree, I was solely going by the "official date of release to the public", if that makes any sense... In terms of "when ppl finally got a hold of these", it remains (sadly) a very early 2000 chip.

Why I care:

I'm researching parts that will give me optimal performance in games during a given calendar year. For the purposes of my project, it is important that individual consumers in a retail market could obtain a given part during this year (i.e. "when ppl finally got a hold of these," as jaZz_KCS said before). I can readily admit that these criteria are arbitrary and may not be helpful to other people.

I can expand on why I've chosen them, if anyone is interested.

Back then regular consumers on average didn't have the fastest computers of a calendar year. Anyways calculating inflation costs a PIII 800 in march 2000 would of costed around 1,100 US dollars way out of the reach for most normal people.

I can tell you from experience of having lived during this time I was 19-20 years old, Id say something like PIII 600 is high end for the target you want for 1999

But yes it is extremely arbitrary and unnecessary because normally speaking higher FPS, higher details settings = optimal gaming experience.

konc wrote:

I always thought "period-correctness" means more "mainstream during that period" than "elite latest model but could be bought if you're a millionaire"

I feel the opposite of that

period correct to me simply means that the components existed at that moment in time

maxtherabbit wrote:

konc wrote:

I always thought "period-correctness" means more "mainstream during that period" than "elite latest model but could be bought if you're a millionaire"

I feel the opposite of that

period correct to me simply means that the components existed at that moment in time

I could build a computer with parts that were of the same year that crysis came out and it would run like shit. It would be period correct though. 🤣