Location: Dubai - UAE
Team: SoC Physical Design Team
Type: Full-time – On-site
About Us
We are a tech company specializing in the design and development of cutting-edge, customized server hardware solutions optimized for artificial intelligence and machine learning applications. Our mission is to empower businesses and researchers to accelerate their AI initiatives by providing them with high-performance, scalable, and energy-efficient hardware infrastructure.
As a rapidly growing company at the forefront of AI hardware innovation, we are constantly seeking talented and motivated individuals to join our team. We offer a dynamic and challenging work environment with opportunities to make a significant impact on the future of AI.
Role Overview
We are seeking a Senior Physical Design Engineer to drive physical implementation of advanced AI SoC / ASIC accelerator designs. This role will own complex blocks and subsystems from RTL/netlist handoff through floorplanning, place-and-route, timing closure, physical verification, and signoff.
The ideal candidate has strong experience with advanced-node physical design, production tapeouts, PPA optimization, and cross-functional execution across RTL, verification, DFT, architecture, and signoff teams.
Responsibilities
- Own block-level or subsystem-level physical design from netlist to GDS.
- Drive floorplanning, placement, CTS, routing, optimization, ECOs, and signoff closure.
- Develop floorplans for compute clusters, memory subsystems, NoCs, clocking structures, and high-speed interface blocks.
- Close timing across multiple modes and corners, including setup, hold, skew, generated clocks, and SI‑aware timing.
- Optimize designs for performance, power, area, routability, and manufacturability.
- Support power planning, IR drop, EM, extraction, and reliability closure.
- Execute physical verification, including DRC, LVS, antenna, ERC, density, and metal fill.
- Collaborate with RTL, architecture, verification, DFT, and packaging, teams to resolve implementation issues.
- Support integration of complex IPs such as NoC, PCIe, Ethernet, UCIe, HBM/GDDR, PLLs, CPUs, GPUs, NPUs, or custom AI accelerators.
- Develop scripts and methodology improvements to improve implementation efficiency, reporting, and design quality.
Minimum Qualifications
- 7+ years of hands‑on experience in ASIC or SoC physical design.
- Experience taking complex blocks or subsystems through physical signoff.
- Strong knowledge of floorplanning, place-and-route, CTS, STA, ECO, and signoff flows.
- Experience with advanced-node, high-frequency, or high-density designs.
- Strong understanding of SDC, setup/hold timing, multi-mode multi-corner analysis, and clock-tree optimization.
- Experience with physical verification and signoff, including DRC, LVS, IR drop, EM, and extraction.
- Experience integrating complex IPs such as NoC, CPU, GPU/NPU, PCIe, Ethernet, UCIe, HBM/GDDR, or other high-speed interfaces.
- Familiarity with multiple clock domains, asynchronous interfaces, generated clocks, and PLL integration.
- Proficiency with industry-standard EDA tools such as Synopsys ICC2/Fusion Compiler, PrimeTime, StarRC, Cadence Innovus, Tempus, Voltus, RedHawk, Calibre, or equivalent.
- Strong scripting skills in Tcl, Python, Perl, or shell.
- Strong debugging, communication, and cross-functional collaboration skills.
Preferred Qualifications
- Experience with AI accelerators, GPUs, NPUs, or highly parallel compute architectures.
- Experience with HBM/GDDR, chiplets, UCIe, 2.5D packaging, or package-aware physical design.
- Experience with DFT-aware physical implementation, scan reordering, MBIST/LBIST, and test-mode timing.
- Prior tapeout experience on advanced-node SoCs.
What Success Really Means
- First silicon boots and runs real AI workloads, not just benchmarks.
- PPA targets are met, with performance per watt competitive with industry leaders.
- Timing, power, physical verification, and reliability signoff are completed with high quality and predictable execution.
- Compiler, firmware, and system teams can use the hardware efficiently without physical-design-driven workarounds.
- Critical silicon or customer issues can be debugged and root-caused quickly, with a target mean time to debug of less than 24 hours.
Join us in our mission to democratize AI compute — where your physical design expertise becomes the foundation of tomorrow's AI breakthroughs.