Talk:Tape-out
![]() | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
|
Myth
[edit]A plausibly looking myth ... preserved here for interest
- this was not a myth -- see below MFago 00:40, 19 September 2007 (UTC)
The process of creating the optical mask for a semiconductor device was originally done by laying pieces of tape representing the various parts of the chip on large tables. These layouts, or masks, would then be photographically reduced and used to selectively expose various layers of a chip to light in the manufacturing process. Long after the time of handmade masks, the creation of the datafiles for a chip's manufacture is still referred to as a "tapeout".
Indeed, the first masks were cut out manually. Even after the emergence industrial tooling for mask manufacture they were "edited" by cutting and pasting. I still have a chromium mask with some holes in it covered by a non-transparent tape. Mikkalai 19:04, 8 Oct 2004 (UTC)
- I did a couple of printed circuit board layouts with tape in 1989. —Tamfang (talk) 13:40, 11 September 2015 (UTC)
taping out a chip
[edit]I agree with this history; one would first do a sketch and a schematic design, and then proceed to do a "tape out" with colored transparent adhesive tape onto clear plastic sheets. In this sense, the tapeout was the layout phase, while now it refers to the end generation of the tape (or more recently, transmitted datafile), after layout had been complete.
Steve P. —The preceding unsigned comment was added by 71.141.253.203 (talk) 07:45, 20 March 2007 (UTC).
taping out: disputed/unreferenced
[edit]- I'm not sure about IC's, but circuit boards used to be "taped out" by laying down adhesive tape onto clear plastic sheets. I've done it (sad but true).
So if this was also true for chips, isn't the article false?!
MFago 23:47, 18 September 2007 (UTC)
- So I added a disputed tag. I suppose this is really related to the "unreferenced" tag. I'll try to find a reference either way.
MFago 23:53, 18 September 2007 (UTC)
- Stephen: was the taped layout photographed and then reduced for manufacture? That's what the procedure was for boards anyhow. If so, that's not all that different from today's usage.
MFago 23:55, 18 September 2007 (UTC)
- So I found a reference. I've left the other verbage in there as well as they are both correct -- but the adhesive tape certainly came first -- assuming that the comments above and the reference that I found are correct. I know that it is correct for circuit boards.
MFago 00:33, 19 September 2007 (UTC)
- Mikkalai (the editor who added reference #1): Our two references are not contradictory. MFago 00:42, 19 September 2007 (UTC)
tapeout's toll on engineers
[edit]Tapeouts are some of the most stressful deadlines possible since so many people are involved and a single mask set costs over a million dollars. Because of this, divorce rates are much higher during tapeouts. —Preceding unsigned comment added by Quanticles (talk • contribs) 14:03, 9 October 2008 (UTC)
computer tape?
[edit]This page no longer says what was disputed - maybe this: I always thought "tape out" was the act of plotting the CIF file or whatever from the data tapes (as was done decades ago) to generate mask artwork for photographic reduction. While the adhesive tape explanation makes possibly more sense, I've designed a lot of PCBs in my time and never ever heard the corresponding process (as described here and in the references) referred to as "tape-out". Maybe I'm too young, but unlikely because I remember using "tape and donuts"! In fact the process (including design which was integral with artwork creation) was known as "layout" (as it is now for the CAD design stage), presumably because of "laying out" the tape (but it could just as easily be from working out the "layout" of the design as in printing etc). Maybe bigger jobs (which would have included ICs) separated this into layout and tapeout. Maybe it's a country thing (I'm in NZ), but AFAIK "layout" is an internationally used term and "tapeout" only really applies to IC design? The reference(s) seem to be written by someone with similar experience to me, so might be just an opinion too. I guess it doesn't really matter, but I did come here to learn the origin of the term, and I'm leaving with more questions than answers. I could have just read a novel (but being an engineer I'm unlikely to ever do that!). --Adx (talk) 15:15, 13 February 2011 (UTC)
RIT
[edit]At IBM a tape-out is called a RIT (release interface tape), RIT-A for the non-metallic structure and RIT-B for the metal layers. Even though it is internal use it can be found on the web quite a bit, in papers, patents and lectures---and IBM represents a significant share of the industry. Not sure if that makes it relevant enough to be mentioned. --EnOreg (talk) 15:28, 6 February 2014 (UTC)
A Commons file used on this page has been nominated for deletion
[edit]The following Wikimedia Commons file used on this page has been nominated for deletion:
Participate in the deletion discussion at the nomination page. —Community Tech bot (talk) 15:52, 14 June 2019 (UTC)
Conflicting etymologies
[edit]The page seems to have two conflicting etymologies: a) tape-out comes from magnetic tapes, and b) tape-out comes from black tape used for PCB layout. The sources support the first etymology and there are no sources for the second. I'm changing the page to reflect this. I did a bunch of searching and for a PCB layout, "tape-up" seems to be the common term, not "tape-out", so they appear unrelated. KenShirriff (talk) 20:25, 15 October 2023 (UTC)
- Well, you didnt exercise due diligence, colleague. - Altenmann >talk 05:11, 16 October 2023 (UTC)
- There is no need for personal attacks. I looked at the added references. The description of Rubylith in "Essential Guide to Semiconductors" is completely wrong. You don't cut strips of Rubylith tape and tape them to large transparent sheets hanging on the wall. Rubylith consists of a red layer on a clear backing. You cut and remove pieces of the top layer that you don't want. Moreover, you don't do this hanging on the wall, which would be very inconvenient. Instead, it is done on a light table (https://www.computerhistory.org/revolution/digital-logic/12/287/1614). You don't end up with a criss-crossing pattern of taped-on stripes (I have examined a Rubylith mask). I recommend removing this reference.
- For the second reference, "Algorithms for VLSI Physical Design Automation", I was unable to find anything on page 30 to support the claim. Can you provide more details? The Robert Widlar link is irrelevant since it shows a mask but doesn't comment on the use of tape-out.
- After a lot of searching, I have been unable to find a use of tape-out for integrated circuits prior to the early 1980s. At that point, reference manuals for design software refer to tape-out utilities to produce a tape with the layout. The Genesil system literally had a "TAPEOUT" command. References: https://archive.org/details/bitsavers_zymosZymosl1985_31287023/page/n677/mode/2up?q=%22tape-out%22 https://archive.org/details/amethodologyforp1094534925/amethodologyforp1094534925.pdf
- It is clear that the term tape-out arose in the 1980s when IC design moved to automated systems and the layout was written to a magnetic tape. Rubylith didn't use tape, so it is irrelevant. Although printed circuit boards were produced with black tape (and I've done that myself), this was called "tape up" not "tape out" and is also irrelevant.
- (I am also unconvinced that paper tape was ever used for tapeout, even though one source claims that.)
- KenShirriff (talk) 16:07, 16 October 2023 (UTC)
- I tend to agree with you, see next section. I will remove the dubious descriptions, but it would be good to find sources that describe this as kind of "urban legend".- Altenmann >talk 17:39, 16 October 2023 (UTC)
- re:
There is no need for personal attacks
- this was not an attack on a person, this was a statement of fact as I saw it: there are a number of easily findable sources that give this description, but you wrote "there are no sources for the second". Hence my conclusion. BTW, "no due diligence" is a frequent accusation if AfDs, so please take it easy on criticim. For example, quite often other editors comment on my poor English, and unless this is done in a rude way, I do not react as a "snowflake". - Altenmann >talk 17:46, 16 October 2023 (UTC)
- Well, you didnt exercise due diligence, colleague. - Altenmann >talk 05:11, 16 October 2023 (UTC)
Rubylith
[edit]There are other descriptions about the use of rubylith, rather credible: in these stories the rubylith was not "taped", but instead on the large rybylith sheet the non-conductor areas were mad transparent by scraping off the light-blockinmg red layer of rubylith reaving the tranparent bottom layer. So the etymology of "tape out" must be scrutinized further. [1]- Altenmann >talk 05:49, 16 October 2023 (UTC)
References
- ^ The History of the Semiconductor Photomask: From Rubylith to EUV. (And in fact this in written in our "photomask" article.
Start a discussion about improving the Tape-out page
Talk pages are where people discuss how to make content on Wikipedia the best that it can be. You can use this page to start a discussion with others about how to improve the "Tape-out" page.