Bspdn Quantify the IR-Droop Reduction It Delivers
# Backside Power Delivery: Step 10 — Quantify the IR-Droop Reduction BSPDN Delivers
## 1. Putting a Number on the Fix
Step 2 quantified frontside-only IR-droop as consuming a substantial double-digit percentage
of the operating voltage budget. This step quantifies what moving power delivery to the back
side actually buys back, using the same droop = current times resistance relationship as the
baseline.
## 2. Why Backside Metal Can Be So Much Lower Resistance
Backside power layers are built on what is, from a routing-congestion standpoint, a completely
free surface: no signal wiring competes for space there, so the metal can be made thick and
wide without the routing-track tradeoffs that constrain frontside power straps. Sheet
resistance for a comparably positioned backside layer can run an order of magnitude lower than
a frontside layer serving the same role, because thickness and width -- the two levers that
most directly lower resistance -- are no longer rationed against signal routing.
## 3. The Reported Range
Published and foundry-disclosed figures describe IR-droop improvements on the order of five to
ten times versus a comparable frontside-only design, with droop collapsing from a double-digit
percentage of the voltage budget down toward low single digits once backside delivery, nano-
TSVs, and buried rails are all in place together.
## 4. Where the Remaining Resistance Lives
Step 9 already identified the rail-to-via contact junction as the dominant remaining
resistance term even after backside delivery is adopted -- meaning the droop reduction
described here is not yet the theoretical limit of what the architecture could achieve, but
the real-world result given current contact and barrier-metal process maturity. Continued
improvement at that junction is the most direct lever for pushing droop even lower.