How uncertainty and noise impact timing?

oSetup Uncertainty = Jitter + skew + Margin

  • Hold Uncertainty = Skew + Margin – Jitter will not impact as hold is checked on same edge and no clock period involved in hold check.
  • Post CTS we have actual skew hence we reduce uncertainty by that amount.
  • Higher the uncertainty, it will reduce that much margin for timing fix, for setup, it reduces required time and hold, it increases required time.
  • Crosstalk delay can impact setup and hold. If Aggressor and victim switching in same direction, makes signal faster causing hold violation. And if switching in opposite direction, adds delay causing setup violation.
  • Crosstalk noise doesn’t directly impact timing but if glitch is above noise margin, it can be wrongly interpreted as valid switching causing functionality failure
Share:
‹ If MCP is 5 cycles, where is setup checked?
Home
How to fix max_cap violations? ›