# AI Accelerator Design Verification Engineer - Full Time — RosaicLabs

- Location: Santa Clara, CA
- Posted pay: $125k–$250k/yr
- Posted: Oct 5, 2026
- Status: open (checked Nov 20, 2026)
- Apply (employer's site): https://jobs.lever.co/rosaiclabs/2d6be587-79ac-413e-be7f-ef54efac96e3
- Employer record: https://visajobs.io/company/rosaiclabs.md

## Sponsorship history
- Tier: Good history
- Occupation (SOC): Electronics Engineers, Except Computer (17-2072)
- Employer's certified LCAs for this occupation in CA, FY2024–FY2026: 5
- …in the whole US: 7
- With this exact title: 0
- Median DOL prevailing wage, level I / II: $124k / $127k

## Description
Join our hardware team as AI Accelerator Design verification engineer verifying the state of the art AI chip, with hands-on ownership of verification environments from test plan through coverage closure and SI bring-up.
What you'll do
Develop verification plans for AI accelerator blocks: tensor processing, fabric / network-on-chip, and SoC flows
Write and debug SystemVerilog/UVM testbenches, checkers, stimulus, scoreboards, and reference models for your assigned block or subsystem
Build emulation-friendly testbenches: synthesizable, transactor-based environments that run the same tests in simulation and on emulators or FPGA prototypes
Use agentic AI tools across the verification workflow: generating test plans from specifications, writing and reviewing testbench code, triaging regression failures, root-causing bugs, and analyzing coverage. Review and validate everything the agents produce.
Verify numerical correctness of compute datapaths (FP*) against bit-accurate golden models
Run real ML kernels and random programs on the RTL, co-simulated against an ISA/architectural reference model
Define, implement, and close functional and code coverage goals, analyzing holes and driving them to closure
Debug failures across RTL, testbench, and reference models, and work with RTL designers to root-cause and resolve bugs
Apply formal verification, assertions (SVA), and emulation where appropriate to improve verification quality and efficiency
Run and maintain regressions across simulation, emulation, and FPGA platforms, and track bug trends and verification progress metrics
Partner with RTL design, performance modeling, CAD/methodology, and software teams to align on verification strategy, improve tools and flows, and drive shared quality, schedule, and tape-out goals
Participate in microarchitecture, RTL, and verification reviews, and write technical documentation
What we're looking for
Bachelor's or Master's degree in Electrical Engineering, Computer Engineering, Computer Science or a related field.
Strong foundation in AI accelerator and computer architecture and digital design, with solid understanding of following:
Tensor cores and SIMT / SIMD / vector processing
Floating-point / computer arithmetic, including reduced-precision formats
On-chip memories, scratchpads, and DMA engines
Cache hierarchy, memory systems, and HBM/DRAM interfaces
Network-on-chip and on-chip interconnects
Dataflow, scheduling, and synchronization across many compute cores
Working knowledge of modern AI accelerator flows and ISAs, including familiarity with ISA simulators or reference models
Proficiency in System Verilog, C/C++, assembly
Familiarity with emulation-friendly verification (synthesizable testbenches, transactors) is a plus
Hands-on experience using agentic AI tools across the engineering workflow: delegating multi-step tasks to agents, writing clear instructions and context, and checking their results critically
Scripting experience for automation, test generation, regression management, or data analysis, such as Python, Tcl, or Rust
Exposure to ML frameworks or workloads (PyTorch, TensorFlow, or ONNX) is a plus

> History, not a promise: past H-1B filings do not mean an employer will sponsor you or a given role. Data: U.S. DOL OFLC LCA disclosure (FY2024–FY2026 for role counts, certified LCAs) and the USCIS H-1B Employer Data Hub. Not legal advice. Methodology: https://visajobs.io/methodology
