Simple - Gds Lef And Def Editor. Ok, so it now supports a few more formats.
If you're running on Windows, the answer is most likely yes, unless you've got a very old graphics card. On Linux the situation is a bit more complex but again if you've got a fairly recent graphics card and have installed the vendor's drivers AND set Linux to use 3D graphics acceleration, then you should be OK. The big advantage of OpenGL is hardware accelerated graphics and support for transparency of layers.
Set the environment variable GLADE_USE_OPENGL to NO. On Windows, right click on My Computer and then select the 'Properties' item. In the dialog choose the 'Advanced' tab, then press the 'Environment Variables' button. Click 'New' on the User Variables section and enter the variable name and value as above. On Linux, set in your .bashrc
In Version 4.0.6 and later it now can. Be aware that there are some limitations - it can only handle stipple patterns that are 32x32, 16x16, 8x8 or 4x4 bits; other sizes are rounded up to the next power of two. It may also fail to read hand-edited display.drf and techfiles, so make sure you've dumped the techfile from a Cadence technology library.
In Version 4.0.42 and later it now can. Be aware that there are some limitations - it can only handle stipple patterns that are 32x32, 16x16, 8x8 or 4x4 bits; other sizes are rounded up to the next power of two. It may also fail to read hand-edited display.drf and techfiles, so make sure you've dumped the techfile from a Laker technology library.
The GDS2 spec is sometimes interpreted as allowing non-orthogonal arrays (arrays where the row spacing and/or column spacings contain both X and Y components). Cadence, who 'own' the GDS2 spec, do not allow non-orthogonal arrays in their translators. So to maintain interoperability, Glade does not either.
As of version 4.1.2 Glade now supports magnification. With some trickery this comes at no memory or runtime overhead. However, it is almost never really needed and other tools may not support it, so beware.
Most people don't require it, Cadence Virtuoso does not support it. In theory it can be supported at no overhead but will require a lot of changes to the UI and careful testing. So if you really, really need it let us know and it could be added to the roadmap.
Well, maybe we don't know about it. Mail firstname.lastname@example.org and tell us what you need. Nothing is promised, but your request may be (a) interesting or (b) simple to implement. So feel free to suggest.
Please email email@example.com and describe the problem as accurately as you can. Again we cannot promise to fix all bugs but so far all the reproducible ones that have been reported have been fixed. A small test case, if possible, greatly increases the chance of reproducing the problem.
You need to run extraction on a layout view before running LVS in order to identify devices (MOS, resistor etc) and generate connectivity info.
No, sorry, not directly; the formats are proprietary. You may be able to translate rules into Glade's format, but there is no guarantee that all rules are supported. If there is a type of rule you need that is missing, let us know and it may be possible to implement it.
No, sorry, Skill is a Cadence proprietary language based on Lisp which was popular in the 1980's. Glade uses Python which is not proprietary and there are plenty of tools to help write and debug Python code. Cadence PCells can only be evaluated by Virtuoso, so they cannot be used outside the Cadence environment.
Glade tries to import the Python wrappers for the gui and database from a file called ui.py, which is located in the distribution directory. Python import uses a search path, set in the environment variable PYTHONPATH, to look for this file (and python PCells, if you use them). See the platform installation instructions on how to set an environment variable.
Check if you are using the autorouter (the 'route' option if you display the Create Wire dialog / hit F3 key). The autorouter obeys the schematic snap grid (Display Options->Snap Grid); if the pins of components in your schematic are not on the snap grid then it is unable to route. For the example library, a snap grid of 0.0625 (inch) is recommended.
The netlister uses the NLPDeviceFormat property on symbol instances when netlisting a schematic to format the entries in the netlist for each instance. If this is incorrect or missing, the netlist will have the instance name but the nets will be missing. See the section on symbol creation for the syntax and usage of the NLPDeviceFormat property.
Copyright © Peardrop Design 2023.