| *** tpb <[email protected]> has joined #symbiflow | 00:00 | |
| tpb | <lkcl> marzoul, you can use a proxy/bouncer such as bnc4you which is entirely free to use for Libera, and, also, Libera themselves have a matrix bridge set up. you can see a number of people with [m] on the end | 11:52 |
|---|---|---|
| tpb | <lkcl> which means they're being bounced/relayed from a matrix account. both solutions will mean you can come back later and receive IRC conversations stored-and-forwarded for you | 11:53 |
| tpb | <lkcl> marzoul, tnt, it's very common in ASICs to have testing / production failures: for example in the Foundry someone banged their coffee cup too hard, caused a ripple, and the laser went skewy and created an incorrect transistor | 11:55 |
| tpb | <lkcl> (ok that's a silly example, but a minor earthquake or some power fluctuations isn't) | 11:56 |
| tpb | <lkcl> rather than chuck the entire design out, what they do is: test the ASIC, then either cut off bits that fail, or just mark it as a "lower resource or lower performance part" | 11:56 |
| tpb | <lkcl> so yes, although the 35t might actually be the exact same die as the 50t, if you were to try actually *using* the "unused" 15k resources it is highly likely that you'd encounter an "inexplicable failure" because of a faulty transistor or faulty track | 11:59 |
| tpb | <lkcl> from a commercial perspective it should be obvious why this is standard practice right across the ASIC industry :) | 11:59 |
| tpb | <nimh> Hi - I still have some questions regarding the fasm_params in the pb_type.xml | 13:12 |
| tpb | <nimh> are they to map to the entries in the prjxray-db? | 13:13 |
| tpb | <nimh> e.g. for the dsp I see files like prjxray-db/artix7/segbits_dsp_r.db | 13:14 |
| tpb | <nimh> it has entries like DSP_R.DSP48.DSP_0.ZAREG_2_ACASCREG_1 26_139 | 13:14 |
| tpb | <nimh> It's still not clear to me how I map the synthesis attributes to this kind of thing | 13:15 |
| tpb | <nimh> or am I barking up the wrong tree? | 13:16 |
| tpb | <nimh> is this what I should be reading? https://symbiflow.readthedocs.io/projects/prjxray/en/latest/dev_database/common/segbits.html | 13:24 |
| tpb | <tpb> Title: segbits files — Project X-Ray 0.0-3534-g5349556b documentation (at symbiflow.readthedocs.io) | 13:24 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!