kens

21 ms·

391.
by kens 2y ago
There's also the bizarre story of Greg Packer, a highway maintenance worker who has been quoted hundreds of times as a "man in the street" in the New York Times and other papers. He apparently just likes being first in line f
392.
by kens 2y ago
The article talks about how people's immune systems can get synced to the first flu they encounter. Coincidentally, the latest issue of Science has a paper finding that genetics is a more important factor than first exposure in determi
393.
by kens 2y ago
The weird thing is that I traced out the circuitry and the bottom bit of the adder is discarded, not the top bit where overflow would happen. (Note that you won't get overflow for this addition because the partial remainder is in range
394.
by kens 2y ago
Dumping the microcode into a bitstream can be done in an automated way if you have clear, high-resolution die photos. There are programs to generate ROM bitsreams from photos. Part of the problem is removing all the layers of metal to expo
395.
by kens 2y ago
Yes, I wanted to use meaningful floating point examples instead of random numbers. You get a gold star for noticing :-)
396.
by kens 2y ago
Returning 0 for undefined table entries is the obvious thing to do. Setting these entries to 2 is a bit of a conceptual leap, even though it would have prevented the FDIV error and it makes the PLA simpler. So I can't fault Intel for t
397.
by kens 2y ago
Yes, the book "Inside Intel" makes the same argument about spreadsheets (p364). My opinion is that Intel's analysis is mostly objective, while IBM's analysis is kind of a scam.
398.
by kens 2y ago
No, IBM's estimate is for a single user. IBM figures that a typical spreadsheet user does 5000 divides per second when recalculating and does 15 minutes of recalculating a day. IBM also figures that the numbers people use are 90 times
399.
by kens 2y ago
There's an FSIN trig inaccuracy, but I don't know of other division issues: https://randomascii.wordpress.com/2014/10/09/intel-underesti...
400.
by kens 2y ago
Author here if anyone has Pentium questions :-) My Mastodon thread about the bug was on HN a few weeks ago, so this might seem familiar, but now I've finished a detailed blog post. The previous HN post has a bunch of comments: https:&
401.
by kens 2y ago
What book was it that you found helpful?
402.
by kens 2y ago
Many years ago, long before "The Bell Curve", I spent way, way too much time reading Murray's book "Losing Ground" (1984) and going to the library to read the papers he referenced and reading the papers those papers
403.
by kens 2y ago
> Has anyone here even read more than the title? Since it costs $35.95 to read the article, probably not. Seriously, paywalling of scientific research is obviously wrong.
404.
by kens 2y ago
Coincidentally, I was a developer on Maple for a summer, working on definite integration and other things.
405.
by kens 2y ago
It was on HN a few months ago: https://news.ycombinator.com/item?id=41418301
406.
by kens 2y ago
Intel used microcode starting with the 8086. However, patchable microcode wasn't introduced until the Pentium Pro. The original purpose was for testing, being able to run special test microcode routines. But after the Pentium, Intel re
407.
by kens 2y ago
For anyone who prefers Bluesky, my thread is also at https://bsky.app/profile/did:plc:svh6dgjnpkdl4dhxahj4xvkv/po... I'm still working on a blog post, which will go into much more detail.
408.
by kens 2y ago
That's an interesting question. The Navajo rug that Intel commissioned shows the Pentium P54C. This is the higher-frequency version of the original Pentium (P5). The FDIV bug had been fixed by this point, so the rug does not have the F
409.
by kens 2y ago
This is the September 2011 edition to be, uh, quantitative.
410.
by kens 2y ago
The IBM System/360 moved to card readers that operated column-by-column, e.g. the 2501 card reader or the 2560 Multi-Function Card Machine. As you said, reading row by row makes a whole lot of sense when you have electromechanical equi
411.
by kens 2y ago
My conclusion is that LETTER is a categorization subroutine that categorizes characters into one of 14 types. Type 2 is minus sign, 4 is alphabetic character, 5 is period, 6 is right paren, 9 and 13 are blanks, 10 is left paren. Types 1, 3,
412.
by kens 2y ago
> Notice cards are (were) 84 characters in width Maybe I'm misunderstanding what you mean by a card, but the 7090 used 80-column punch cards of which 72 columns could be read into the computer. (Which is why Fortran used 72-characte
413.
by kens 2y ago
Regarding terrorism, some may be too young to remember the Matsumoto sarin attack in 1994 and the Tokyo subway sarin attack in 1995. These attacks by the Aum Shinrikyo cult killed 8 and 13 people respectively, surprisingly low numbers given
414.
by kens 2y ago
Do you have any interesting stories of chip development during that era? What sort of EDA tools did you use?
415.
by kens 2y ago
> There’s this individual’s name which involves a clock sound: Nǃxau ǂToma I was extremely puzzled until I realized you meant a click sound, not a clock sound. Adding to my confusion, the vintage IBM 1401 computer uses ǂ as a record mark
416.
by kens 2y ago
They etch holes in the oxide for the vias and then use CVD (chemical vapor deposition) to put a tungsten film over the surface, filling the holes. Then they polish off the excess tungsten with CMP (chemical mechanical planarization). I thin
417.
by kens 2y ago
I spent a long time with various chemical processes to remove the layers of the chip before discovering that 1500-grit sandpaper works shockingly well. Then I took the pictures with my metallurgical microscope and stitched the images togeth
418.
by kens 2y ago
Yes, input pins are kind of an exception; the source or drain providing the voltage is external.
419.
by kens 2y ago
The diodes break down nondestructively under the voltage, getting rid of the charge regardless of polarity. During operation of the chip, the diodes are reverse biased, so they have no effect.
420.
by kens 2y ago
That Pentium is the 80502, so it's almost the same as the one in my article except that it is built with 600 nm technology instead of 800 nm and it has 200,000 more transistors. It's easy to knock the lid off the package with a ch

More ›