Talk:GDSII

Alternative formats
What are the alternatives to GDSII? I know that IBM accepts IC layout files in a format called GL1 or GL/1. Are there others? --HelgeStenstrom 09:21, 29 August 2007 (UTC)

One alternative is the Caltech Intermediate Form - CIF. The main drawback is the file size, a CIF file is larger because it is a text file, compared with the binary data of GDSII. --70.178.221.231 (talk) 06:42, 7 May 2008 (UTC)

Format ownership and external references
Is it really true that Cadence owns the GDSII format? If so, there should be an external reference to them, because they has the most accurate information on the file format. --HelgeStenstrom 09:24, 29 August 2007 (UTC)

Since Cadence bought Calma, they now "own" GDSII. However, they have never made any documents public, and therefore the best data available is found in papers and documents published before the Cadence purchase. For example: "Calma Corporation, GDS II Stream Format, July 1984". Strubin 15:48, 29 August 2007 (UTC)

Here is a link to a Cadence publication with the GDSII/Stream Format Design Data Translator's Reference. An other link to a nice GDSII Format Descripion is GDSII format. To mention two little gds-tools (klayout and gds2pov) is totally arbitrary. Other tools are much more popular. And POV-Ray has nothing to do with GDSII. --Jurgenthies (talk) 21:59, 17 December 2007 (UTC)

3D view
I thought that GDSII contained only 2-D data. If that's the case, the 3-D image in the article is deceptive. It should be replaced, or a clarification should be added. --HelgeStenstrom 09:26, 29 August 2007 (UTC)

I agree. GDSII has layers, but the layer numbers do not relate to "height" on the chip. Therefore, it is not possible to construct the pictured 3D view from GDS data without additional information. Strubin 15:41, 29 August 2007 (UTC)

Removed biased OASIS info
I removed part of the OASIS text format, since it was biased.

"It has been claimed that the OASIS format is between 10 to 50 times more efficient than GDSII." This is IMHO a highly biased statement. We have written a GDS-II file optimizer that uses smart combinations of AREF an SREF, to search for regularity (not necessarily orthogonal), together with grid searches. When these files were than gzipped as well, the compressed files also reached factors up to roughly 30 times for complex (analogue) chip layouts. It should perform much better on digital circuits, with eg memory banks, which have lots more regularity. The reference should compare like with like. An optimized OASIS generator should be compared with an optimized GDS-II file.

"Unlike GDSII, OASIS allows to store geometries as shape arrays." As said, GDS-II has the AREF and SREF command.

"and does not exhibit intrinsic limitations such as the 32 bit coordinate space." Based on a smallest dimension of 1nm, 32 bit allows you to draw a chip of 16 square meters. So where is the limitation in practice? The highest compression ratios are also only possible on 16-bit or 24-bit integers of OASIS. —Preceding unsigned comment added by 24.132.19.10 (talk) 22:22, 15 November 2007 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on GDSII. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20120119053116/http://www.wxnanotech.com/ to http://www.wxnanotech.com/

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 12:49, 9 October 2017 (UTC)