OneSpin DV-Verify's surprise comeback in ABV is Best of 2019 #5a
ONESPIN'S SURPRISE COMEBACK: In the "Best of 2018", last year OneSpin was
amazingly weak. Users wrote 4,988 words about ABV tools in 2018; and only
a measily 59 words were about *anything* OneSpin. (Talk about weak! Ouch!)
But now, for "Best of 2019", OneSpin got ~3,000 users words on the OneSpin
vs. JasperGold wars covered here and in DAC'19 #5b -- plus ~1,400 more user
words OneSpin's other tools. (See DAC'19 5c.) Talk about a comeback!
---- ---- ---- ---- ---- ---- ----
WHERE ONESPIN WINS: The 3 areas where users gave OneSpin high marks were
assertion observation coverage, speed, and safety (fault injection).
1. Assertion Observation Coverage
"OneSpin's Quantify app uses the 'mutation' method. The mutation
method injects a forced logic state on each individual signal,
then evaluates if the assertion passes or fails. This gives you
100% coverage. ... JasperGold's Cone of Influence (COI) coverage
method is less comprehensive than mutation."
2. Speed
"OneSpin is better than JasperGold for complex properties. For
example, to prove a complex property on one design we verified,
OneSpin took 2 hours, JasperGold took 20 hours However, it
is difficult to generalize these results from a few designs."
3. Fault Injection
"With fault injection, to ensure your design won't let catastrophic
failures occur, Faults are deliberately injected. OneSpin has a
fault injection capability in their core tool. Cadence does not
include fault injection in their core tool -- it's an add-on."
---- ---- ---- ---- ---- ---- ----
WHERE JASPER WINS: From the user comments, JasperGold has this amazing debug
capabilty that's intuitive -- while OneSpin is seen as basic at best:
"OneSpin's user interface is a bit old-fashioned."
"It's debugging complex counter examples is painful within the GUI."
"OneSpin's debugger takes some getting used to."
"Their code debugging is "ugh"."
But beyond this GUI drawback, OneSpin has some killer capabilities.
