GPU Physical Design & Power

Section goal

Large-array floorplanning, clock/power architecture, thermal control, and timing closure for high-throughput GPU silicon.

How to study this section

1. Start with each topic hub and sketch the SIMT-to-silicon mechanism.

2. Use reports and debug pages to prove bottlenecks with evidence.

3. Practice worked examples and interview drills for design-review fluency.

4. Close with checklist and silicon impact before signoff claims.

Topics

1. sm-array-floorplanning/ - SM Array Floorplanning

2. clock-power-domains-gpu/ - Clock & Power Domains in GPU

3. thermal-and-dvfs/ - Thermal Management & DVFS

4. gpu-timing-closure/ - GPU Timing Closure

Related topics

• GPU Design course hub

GPU deep dive

GPU PPA closure must co-optimize floorplan locality, IR stability, thermal headroom, and timing margin.

Concept diagram

GPU PD VIEW

HBM edges + SM clusters + cache rings + power/clock grid

Metric graph

CLOSURE PRESSURE

timing risk        ███████
thermal risk       █████
IR transients      ████

Reports and artifacts

• SM-array congestion map

• thermal hotspot report

• IR drop during burst load

• timing closure dashboard

Mini case study

A floorplan iteration improved routing but worsened hotspot density, forcing DVFS throttling in sustained workloads.

Debug branches

• Map critical paths to floorplan and thermal zones

• Run burst-current IR checks, not only static IR

• Tie DVFS behavior back to physical hotspot evidence

Senior review question

Ask: which metric and benchmark pairing proves this topic is truly closed in production context?