azonenberg changed the topic of #homecmos to: Homebrew CMOS and MEMS foundry design | Wiki: http://homecmos.drawersteak.com/wiki/Main_Page | Repository: http://code.google.com/p/homecmos/ | Logs: http://en.qi-hardware.com/homecmos-logs/
soul-d has quit [Remote host closed the connection]
soul-d has joined #homecmos
Charlie has quit [Ping timeout: 240 seconds]
chris_99 has joined #homecmos
Charlie has joined #homecmos
<berndj> magic's drc complains when a substrate contact is less than 4 lambda from a diffusion contact, also that opposite diffusions are too close
<berndj> but when i put the substrate contact directly adjacent to the diffusion contact, drc is happy. what gives? why is "opposite diffusions too close" not a problem then?
<berndj> in both cases the intent is to strap the contacts together with m1. is it that magic's drc can't "see" that the contacts are strapped together, when there's a short m1 jumper connecting them?
<azonenberg> Probably
<berndj> whee, the dff simulates correctly!
<berndj> not a full spice sim, just irsim's switch model simulation
<azonenberg> :D
<azonenberg> On a somewhat related note
<azonenberg> i want to make a tool for automatic recognition of standard cells at some point
<azonenberg> In other words, given optical or secondary electron images of each layer
<berndj> that sounds hard :-/
<azonenberg> generate vectors
<azonenberg> (possibly after manually drawing the boundary of each cell)
<azonenberg> then extract transistors from the vector images
<azonenberg> and so a behavioral (not necessarily SPICE-level since process details are unknown) simulation with all possible inputs
<azonenberg> and maybe some heuristics to speed things up
<azonenberg> to determine "this is a NAND3"
<azonenberg> "this is a positive edge triggered dff"
<berndj> that last step is the one i don't even know how i don't know how to do it
<azonenberg> Yeah
<azonenberg> Basically, right now the first step in REing a standard cell chip is manually figuring out what each cell does
<azonenberg> i'd like to automate that
<Sync_> yeah that's a good idea ;)
<azonenberg> say "this is a cell", highlight it
<azonenberg> five seconds later "NAND2x1"
<berndj> would it be useful if you could at least find instances of common cells, regardless of their function?
<azonenberg> degate can find cells once you higlight one instance and give it a name
<berndj> "there's a cell here, and it's the same as the one here, here and here"
<azonenberg> But it can't tell what is a cell
<azonenberg> That would be a little harder but nice
<Sync_> oh degate can do that now
<Sync_> cool
<azonenberg> Degate also segfaults
<Sync_> yes it is good at that
<azonenberg> whenever i try to add >1 layer in the last version i tried
<azonenberg> i'm leaning toward writing my own tool
<azonenberg> its too unstable :p
<Sync_> yeah it sucks
<azonenberg> What I dream of is an IDA for hardware
<azonenberg> something that takes in either an FPGA bitstream or die photos
<azonenberg> does extractions and recognizes cells on the photos, reverses the bitstream
<azonenberg> now you have a cell-level netlist
<azonenberg> then do heuristics and isomorphism checking within neighborhoods to find higher level structures like adders and muxes
<azonenberg> and perhaps even known hard IP blocks
<berndj> hash it all and make a rainbow table!
<Sync_> oh actually, have y
<Sync_> gah
<Sync_> wrong channel
<azonenberg> berndj: lol
<azonenberg> seriousl, it would be awesome to take in die images and have it automatically find "functions" like ida does
<azonenberg> but they're standard cells
<azonenberg> then find higher level structures
<azonenberg> and "decompile" the result to generic RTL
<azonenberg> one module at first, you could then split stuff off as you saw fit
<Sync_> I wonder when chip manufacturers will start to implement measures to stop people FIB from the bottom of the die
<azonenberg> well 3D stuff is going to be a big PITA to reverse :p
<Sync_> yeah but that's not too popular :P
<Sync_> having annoying structures on top is
<Sync_> and I finally made some progress in metalization
<berndj> what's #D stuff going to do to power densities?
<berndj> aargh shift key! i mean 3D stuff
azonenberg1 has joined #homecmos
azonenberg has left #homecmos [#homecmos]
azonenberg1 is now known as azonenberg
chris_99 has quit [Quit: Leaving]
soul-d has quit [Remote host closed the connection]