Skip to main content
Broadcom

# AI/ML Silicon Accelerator Memory Architect

Broadcom USA-CA San Jose Innovation Drive, United States Full-time 1 day ago
Digital Design & SoC Architecture

$200,000 - $320,000 USD yearly

Originally posted 3 November 2026 by the employer.

About the role

You will define the end-to-end memory and data transport architecture for accelerators, spanning controller design, on-die networks, DMA engines, and compute scratchpads.

What you'll do

  • Define multi-generation memory architecture and strategy across HBM, LPDDR, GDDR, DDR, on-die SRAM hierarchies, and scratchpads.
  • Derive capacity, bandwidth, and latency requirements from AI workloads including model size, KV-cache growth, prefill versus decode, and MoE sparsity.
  • Architect memory controllers and cache/scratchpad microarchitecture covering scheduling, address mapping, and QoS across compute, DMA, and host traffic.
  • Define data transport paths from memory to compute blocks, covering on-die interconnect topology, flow control, deadlock avoidance, QoS, and pJ/bit energy budgets.
  • Specify DMA, gather/scatter, layout-transform, and multicast engines that support data movement.
  • Develop analytical and simulation-based models to evaluate architectural trade-offs and drive silicon realization.

What you'll need

  • MSEE/MSCE or PhD in EE, CE, CS, or related field, or equivalent demonstrated architectural impact.
  • 18+ years in ASIC/SoC architecture focused on memory systems and data movement for HPC or AI.
  • Deep knowledge of memory subsystems including HBM, DDR/LPDDR/GDDR, controller architecture, cache coherency, memory consistency models, and tiered/pooled memory.
  • Demonstrated ownership of data movement architecture including on-die interconnect, DMA/copy engines, buffer hierarchies, and dataflow to compute.
  • Understanding of the latency, area, and power costs of protocols such as NVLink, UALink, RDMA, ARM CHI, UCIe, and PCIe.
  • Strong performance modeling and system-level simulation background with command of latency, throughput, utilization, and scalability.

Skills: HBM, DMA, Interconnect, Memory Subsystems, SoC Architecture, AI Accelerator

More semiconductor jobs: Broadcom jobs · SoC design jobs

Apply now
USA-CA San Jose Innovation Drive, United States
On-site
Full-time
$200,000 - $320,000 USD yearly
1 day ago

Share this job