Digital Design & SoC Architecture
AI/ML Hardware
Originally posted August 16, 2026 by the employer.
About the role
As an ASIC Architect, you will define the silicon and system microarchitecture for custom compute platforms targeting LLM and diffusion model inference. You will lead hardware and software co-design efforts to build custom mixed-signal compute tiles, memory hierarchies, and tile interconnects.
What you'll do
- Define the architecture and microarchitecture of AI accelerator compute blocks, including PE array design, datapath organization, sparsity exploitation, and reduced-precision computation.
- Translate transformer architectures, attention mechanisms, KV-cache management, and diffusion execution flows into unambiguous microarchitecture specifications for RTL engineers.
- Evaluate full-stack PPA trade-offs spanning algorithm behavior, memory hierarchy, on-chip interconnect, and physical design constraints.
- Partner with the compiler lead on ISA co-design, aligning the programming model with the underlying hardware microarchitecture.
- Direct pre-silicon validation and prototyping strategies, collaborating with FPGA design engineers to de-risk architectural decisions before tapeout.
- Build analytical and performance models to evaluate throughput-per-watt across generative AI workloads.
What you'll need
- Degree in Electrical Engineering, Computer Engineering, Computer Science, or equivalent experience.
- Experience in architecture or microarchitecture of high-performance digital systems, AI accelerators, or compute engines.
- Experience using cycle-accurate or analytical simulation models to drive and defend architectural decisions prior to RTL implementation.
- Experience writing microarchitecture specifications and collaborating directly with RTL engineers through design closure.
- Practical understanding of quantization, reduced-precision approaches for inference, and their corresponding hardware implementation implications.
- Proficiency in Python or C++ for performance modeling and analysis, alongside familiarity with SystemVerilog or equivalent RTL.
More semiconductor jobs: Normal Computing jobs · Hardware design jobs · Mountain View jobs
Market context
Tracked in our index since 17 Aug 2026. This is our observation date, not the employer posting date. Specialty: Hardware Systems Engineering (AI ML Hardware). 1242 active roles in this specialty at 122 employers in our index, as of 21 Nov 2026.
How we counted: postings read from employer career sites. How we count
Normal Computing
Mountain View, CA, United States
On-site
Full-time
3 months ago