I found that flow incorrectly streams out GDS from DEF using open-source-silicon #openlane

Channels

mpw-6plus-silicon

pd-openlane-and-sky130

chipignite2206q_stanford_bringup

cadence-virtuoso

fossee-iitb-esim

dynamic-power-estimation

ieee-sscs-dc-23

gfmpw-0-silicon

silicon-validation

fossee-iitb-google-sky130

ieee-sscs-dc-22

mpw-one-silicon

cadence-innovus

verification-fe

foss-asic-tools

openpositarithmetic

openadiabaticlogic

hardware-beginners

design-services

private-shuttle

tapeout-pakistan

rad-lab-silicon

cadence-spectre

mpw-one-clean-short

coalition-for-digital-environmental-sustainability

vlsi_verilog_using_opensource_eda

vlsi_verilog_using_opensoure_eda

openlane_cloudrunner

digital-electronics-learners

verilog-learners

open-microelectronics-textbook

mwcas-2023-rtl2silicon

system-verilog-learners

vlsi-learners-group

chilechipmakers

silicon-validation-private

openhighqualityresonators

waveform-viewers

vendor-synopsys

silicon-photonics

opendataset-circuitnet-for-ml-tasks-in-vlsi-cad

generative-ai-silicon-challenge

community_denmark_dtu

shuttle-precheck

openlane-development

ieee-sscs-dc-21q3

sky130-pv-workshop

verification-be

Title

e

Eugene Stolbikov

08/10/2021, 8:01 AM

I found that flow incorrectly streams out GDS from DEF using KLayout. It implements incorrect DBU=0.005 defined in .lyt klayout technology file in PDK. Therefore, all sizes in GDS are 5 times smaller than in DEF. But sizes of merged STDcell topologies remains normal. In result we have the following GDS (spm design example). The DBU in .lyt must be 0.001!

m

Mitch Bailey

08/10/2021, 8:22 AM

Could it be related to this https://skywater-pdk.slack.com/archives/C01DH7SPX41/p1624562711029700

e

Eugene Stolbikov

08/10/2021, 8:25 AM

Exactly!

I'm not sure who is the author of *.lyt klayout technology file in sky130A PDK - pdk authors or openlane authors.

m

Matt Liberty

08/17/2021, 3:27 PM

This seems worth reporting as an OL issue