rmay635703 wrote on Today, 01:06:

Anyway, I have never seen your narrative but I have always seen the narrative here (and variations on it) having lived it, right or wrong the accepted narrative was that the sx25 was around to kill the 386 market.

Yes, that was the common belief in the press before 486sx release, but it subsided the moment chip hit the market. Press had to walk back their grave predictions 😀 That linked June 1991 Byte issue is all about 486SX 😀 Editors note:

"The first systems we've seen that are built around the i486SX perform about as you'd expect, but they cost
much more than alternative systems that deliver equivalent performance"

"In this month's cover story, you'll see how a Club American Technologies system built around AMD's
new 40-MHz 386 clone chip easily matches and/or exceeds the CPU performance of i486SX-based systems
from ALR and AST. But the Club system costs one third to one-half less than the ALR and AST systems. "

And the page 42 "486SX vs Am386-40 Lab analysis reveals a surprise winner" cover story is already titled "The 486SX Falls Short" giving away the surprise 😀

JANUARY 22, 1992 MICROPROCESSOR REPORT issue (specialized paid newsletter for industry insiders https://en.wikipedia.org/wiki/Microprocessor_Report) I linked earlier :

"Price War Coming?
...Since Intel hopes to draw system makers and buyers away from the 386 and toward the 486SX, very aggressive pricing on that device is likely"

its 1992, half a year after release, and market experts still think "aggressive pricing ... is likely" somewhere in the future 😀. But they already see thru intel BS
"The transition from the 386 to the 486 will not be nearly as rapid or as complete as the 286-to-386"
"Despite Intel’s 486 push, the 386SX and 386DX will remain the volume leaders in 1992, and Intel will have to share this market with its competitors"

Semiconductor procurement worldwide, 1992-1993, that one is an even more hardcore business/accountant insiders newsletter
https://archive.computerhistory.org/resources … 7-05-01-acc.pdf :

"The concomitant introduction of low-power 80486/80486SX
devices is a sign of Intel's goal to migrate users now to the higher-priced 80486 areiia
and away from the lower-priced 80386/80386SX market. According to the SP quarterly
price survey, Intel's 16-MHz 80386SX and 20-MHz 80386SX devices (1,000-5,000 unit
volume) should sell during the fourth quarter of 1991 for $56.35 and $84.90, respectively. The
25-MHz 80386 device should sell for $153.50 during the fourth quarter (same volume). By
contrast, the new Intel devices are priced higher. Prices for the low-power versions of
the following ICs (1,000 units) are as follows: 16-MHz 80486SX—$235; 20-MHz 80486SX—
$266; 25-MHz 80486—$471." [/details]

Your screenshot is mid 1992 https://www.nytimes.com/1992/05/14/business/c … -than-half.html
at that point 386 were 2x cheaper already. Q1 1993 prices are on page 8 of "Semiconductor procurement worldwide" 1991-1993 by Dataquest https://archive.computerhistory.org/resources … 7-05-01-acc.pdf

11993
2Product			  	Q1 		Q2 		Q3
3386SX-16 PQFP     	39 		36		33
4386SX-25 PQFP     	42 		37		33
5386SL-25 PQFP     	66 		61		57
6AM386DX-40 PQFP   	51 		49		48
7386DX-25 PQFP     	61 		57		52
880486SX-25 PQFP   	99 		89		85
980486DX-33        	317		298		285
1080486DX-50        	469		447		442
1180486DX2-50       	428		412		395
12Pentixun-60 CPGA  	NA 		878		843

rmay635703 wrote on Today, 01:06:

Commentary similar to this one can be found everywhere mid/late 92 and early 93

By journalists still on hopium 😀 Sure 486SX was $90 in 1993, but 386DX was $50. Then you had to add motherboard cost that had similar ratio. 386 held low end up to late 1993 with retail prices (Byte 8 1993 ads):
386DX40 +mobo ~$200
486SX25 +mobo $350
486DX33 +mobo $490

rmay635703 wrote on Today, 01:06:

On the original topic Intel dx4-75’s were extremely popular far beyond what they should have been, Intel seemed to prefer selling them over the 100 for some reason.

75 was a specialized low power Laptop part, Intel couldnt get dx4 100 power down, not to mention bigger silicon Pentiums. Same way Pentium 3s had a long Laptop life during Pentium 4 reign.

Obviously the real reason it existed was to give DX2-50 buyers a chance to double down and make a second mistake... well on desktop at least.

I was still a bit green at the time these were around, but even then I would have laughed at choosing one for a desktop over an AMD DX2-80 for probably a third the price. New motherboards around that time should have been perfectly happy about 40 or 50. OEMs gonna OEM though.

Laptops are the only real use case. IIRC there wasn't a 3.3V DX2 until after... and that might have been a DX4 die underneath.

I will just leave it with Intel cuts price in half articles are from March 92, with the sx at $119

As far as I know there was a massive and real price cut around that time just a short time after your quotes from January.

Your narrative about system level pricing is more adue as you say about OEMs absorbing price drops and placing in pockets.

The response to $119 sx chips was a lower 386 price plus 386 motherboards became smaller and more integrated than 486 class thus maintaining the rest of the price difference in the board and other components.

While Intel had the cpu price war beige box makers were further outsourcing components east while multiple OEMs were still trying to build or assemble onshore placing a rift above the China price.

It’s worth knowing the average consumer in the us was almost unaware of the dx40 and would go pay OEM $$$$ for any machine whether it was the 386sx or 486sx.

My friend base at that time 1992
Pb 486sx25
Me Tandy 1000rlx - $399 full system 286-10
PCJR - 640k
Compaq 286slte (of some kind was maybe a year old and tiny)
386dx33 multimedia- assembled in 1989 by the one technical industry dad adding upgrades every once and a while
PB 386sx25

One beige box for our group

BitWrangler wrote on Today, 03:21:

Obviously the real reason it existed was to give DX2-50 buyers a chance to double down and make a second mistake... well on desktop at least.

I was still a bit green at the time these were around, but even then I would have laughed at choosing one for a desktop over an AMD DX2-80 for probably a third the price. New motherboards around that time should have been perfectly happy about 40 or 50. OEMs gonna OEM though.

Laptops are the only real use case. IIRC there wasn't a 3.3V DX2 until after... and that might have been a DX4 die underneath.

The local tech college had piles of old gen Omnitech 486 boxes with ISA Only Micronics motherboards that it appears were upgraded during their lives to dx2-50’s
My guess is with 0kb to 128kb of cache and trident isa video the upgrade wasn’t as satisfying as they wanted from the DX25 (likely) or sx25 (less likely due to build dates)

I only know because I bought tons of the machines at auction years later.

The systems had an socketed oscillator and were rated to 50mhz FSB so one would have thought they could have upgraded differently.

When I bought them it was a mix of 486dx2-66 WD vlb 256k machines mixed with the older isa dx2-50’s
Was a massive rift in performance there.

Ah well, was interesting times.