What does EM signoff check: average, RMS or peak current?

Short Answer

EM signoff compares the current density in every wire segment, and the current per cut in every via, against the foundry's limits, and it can do that for three kinds of current: average, RMS and peak. Average current drives the slow drift of metal atoms that eventually opens a line, RMS tracks Joule heating, and peak guards against short high-current pulses. Static analysis only knows the average, dynamic analysis can check all three, and perform emcheck (RH) runs AVG, RMS, PEAK or all, with all as the default.

Technical Explanation

• The mechanism and Black's equation are covered on the routing EM page. This page is about which current you compare against which limit, and where each check applies.
• Average (DC) EM is the power-grid check. PG current flows mostly in one direction, so the average is large and it is the average that moves metal over years. RedHawk static EM uses the true average current density for wires and current per cut or area for vias.
• RMS EM is about heating. Signal and clock currents flow both ways, so their average is small, but the RMS value still heats the wire, and heat shortens EM lifetime for that wire and for its neighbours.
• Peak EM limits the largest instantaneous current, which matters for short, strong pulses such as a big driver switching a heavy load.
• RedHawk does not run EM as part of static or dynamic simulation by default. Set the GSR keyword ENABLE_AUTO_EM 1 (RH) to check during post-simulation processing, then use perform emcheck -mode AVG|RMS|PEAK|all -net VDD (RH). Skipping -mode and -net checks all modes on all nets.
• Limits can differ per mode. Simple per-layer limits in the tech file apply to every mode; separate files through EM_TECH_DC (RH), EM_TECH_RMS (RH) and EM_TECH_PEAK (RH), or named EM_RULE_SET (RH) groups, give each mode its own rule. DC rules are also used for static EM.
• Reports: static EM writes <design_name>.em.worst under adsRpt/Static with segments over 100% of the limit, top 1000 by default. Dynamic EM writes .em.worst.peak, .avg and .rms files under adsRpt/Dynamic.
• In ICC2, analyze_rail -voltage_drop static -electromigration -nets {VDD VSS} (ICC2) runs PG EM through RedHawk Fusion using the limits in rail.tech_file (ICC2). The static map shows average values; a dynamic run can show average, peak or RMS. Save the block, because the EM error data is not written to the working directory.

Follow-up Question & Model Response

"Why do PG nets usually fail average EM while signal nets fail RMS or peak?"

Candidate Model Response: A power strap carries current in one direction all the time, so its average is close to its typical value, and that steady flow is what moves metal. A signal wire charges and discharges its load, so the current reverses every transition and the average nearly cancels. What remains is heating from the RMS value and short peaks when a strong driver switches. That is why PG EM is usually a static, average check and signal EM needs dynamic current with RMS and peak limits.