CPU Clock Distribution and Custom Clock Design

Full-time  *
* · US / Canada / India*
* Nuvacore is building ground-up high performance, low-power CPU for next-generation compute workloads. As a member of the CPU clock design team, you will architect and deliver the clock generation and distribution networks that set the frequency, skew, jitter, and power ceiling for the product.

THE ROLE
As a Custom Clock Design and Distribution (CCD) Engineer, you will participate in the following :

Architect and drive the end-to-end clock generation and distribution methodology using H-tree, mesh, spine, and multi-source clock-tree synthesis for high-frequency CPU cores, co-optimizing skew, insertion delay, jitter, duty-cycle distortion, and clock power.

Design and characterize custom clock circuits at transistor level clock buffers and repeaters, dividers, deskew and duty-cycle correction, clock-gating and clock-mux cells meeting PPA targets across PVT corners.

Own end-to-end clock-network closure: SPICE-driven skew, latency, and jitter analysis, EM/IR and signal-integrity signoff on the clock grid, and STA correlation (OCV/AOCV/POCV) across Functional and DFT modes.

Partner with PLL/FLL, timing, power, CAD, physical design, and DFT teams aligning clock source models, jitter budgets, and power-aware clocking and drive Tools, Flows, and Methodology (TFM), including clock-analysis automation and team-wide best practices.

MINIMUM QUALIFICATIONS

Hold a BS or MS degree with 12+ years of experience (for Lead roles) or 5+ years of experience (for Individual Contributor roles) in clock design, clock distribution, and/or custom circuit design.

Hands-on transistor-level design experience with clock dividers, deskew and duty-cycle correction, clock-gating and mux cells at advanced CMOS nodes.

Proven experience constructing and analyzing low-skew, low-power clock distribution networks using H-tree, mesh, spine, and multi-source CTS for high-frequency CPUs.

Strong SPICE simulation and analysis skills, with solid grounding in device physics, RC delay, and signal integrity applied to clock-path design and verification.

Well-versed in STA and clock timing signoff including setup, hold, duty cycle closure across Functional and DFT (Design for Test) modes.

Demonstrated track record delivering clock networks and/or clocking IP through tape-out, with silicon-validated PPA — frequency, skew, jitter, and power — in high-volume products.

Strong communication and collaboration skills, comfortable working directly with PLL, CAD, physical design, and foundry partner teams.

REQUIREMENTS — PREFERRED

Proficiency with PLL/FLL specifications, clock source modeling, skew estimation, jitter budgeting, and jitter/duty-cycle measurement.

Experience with power-aware clocking — clock gating, DVFS, clock-skipping, and resonant or mesh-based clocking strategies for clock-power reduction.

Deep understanding of clock-network signoff — EM/IR on the clock grid, dynamic voltage-drop-aware timing, crosstalk and noise, and Monte Carlo variation studies.

Experience with testchip clock development and benchmarking, and partnering directly with foundries on PDK bring-up and clock-device co-optimization (DTCO) at leading-edge nodes.

Familiarity with the full clock flow — architecture, circuit design, physical implementation of the clock network, and timing/model generation — at block and full-chip level.

Proficient in Python, Perl, or TCL to build and improve clock analysis, characterization, and methodology automation.

Experience using Agentic AI tools for productivity and automation.

$140,000 - $250,000 a year