Skip to content

Search jobs inside your AI agent. ClaudeChatGPTCursorGrok Bot

Connect

Neurophos

Develops high-performance photonic processors for AI inference acceleration.

H-1B sponsor history — This employer has had H-1B petitions approved by USCIS in the last five fiscal years, per the USCIS H-1B Employer Data Hub. The approval belongs to the employer and is not a commitment to sponsor this role.
ai
neurophos.com ↗
Jobs Hiring Closed Discussion

Company facts

Series A-stage
Founded
2020
Employees
38
Headquarters
United States
Type
Private
Funding
Series A · $110M · 2026
Subsidiaries
SiliconBee

Open positions

4 jobs
Neurophos
8 views 2 months ago

Staff Modeling Architect

Open 52d On-siteNorth AmericaFull-timeStaff H-1B sponsor history — This employer has had H-1B petitions approved by USCIS in the last five fiscal years, per the USCIS H-1B Employer Data Hub. The approval belongs to the employer and is not a commitment to sponsor this role.
United States

Staff modeling architect at a silicon-photonics AI chip startup: builds performance/energy models and bit-accurate functional models of an optical AI inference accelerator, co-designing tiling, ISA, and memory hierarchy with compiler and RTL teams. Core stack: modern C++, Python (NumPy/Pandas), simulator kernels (SystemC/gem5-style), Verilator/SystemVerilog.

AI C++ Event Driven Architecture FPGA GPU +13 skills
250K – 290K $
 Sign in to see your match
Neurophos
13 views 2 months ago

Senior Modeling Architect, Performance Benchmarking

Open 52d On-siteNorth AmericaFull-timeSenior H-1B sponsor history — This employer has had H-1B petitions approved by USCIS in the last five fiscal years, per the USCIS H-1B Employer Data Hub. The approval belongs to the employer and is not a commitment to sponsor this role.
United States

Owns the performance and energy benchmarking numbers for Neurophos' T100 optical AI inference accelerator, producing results at every fidelity level: roofline/limiter models, architecture performance models, RTL simulation, and measured competitor GPUs. Daily work centers on Python benchmark harnesses, GPU profiling (Nsight), and LLM inference stacks like vLLM and TensorRT-LLM, onsite in Austin or

AI AWS Azure Cloud CUDA +14 skills
210K – 250K $
 Sign in to see your match
Neurophos
8 views 2 months ago

Modeling Architect

Open 52d On-siteNorth AmericaFull-time H-1B sponsor history — This employer has had H-1B petitions approved by USCIS in the last five fiscal years, per the USCIS H-1B Employer Data Hub. The approval belongs to the employer and is not a commitment to sponsor this role.
United States

Hands-on modeling architect for Neurophos' T100 optical AI inference accelerator, building analytical performance models (roofline/limiter), Python/C++ functional and power models, and RTL simulations. Works on hardware/software co-design of workloads, memory hierarchy, ISA, and scheduling alongside compiler, runtime, and RTL teams.

AI C++ CUDA Event Driven Architecture GPU +8 skills
170K – 200K $
 Sign in to see your match
Neurophos
6 views 3 months ago

Principal Fabric/Network Architect

Open 75d On-siteNorth AmericaFull-timePrincipal H-1B sponsor history — This employer has had H-1B petitions approved by USCIS in the last five fiscal years, per the USCIS H-1B Employer Data Hub. The approval belongs to the employer and is not a commitment to sponsor this role.
United States

Principal Fabric/Network Architect designs the end-to-end data path for Neurophos’ optical AI accelerator, including RDMA transports, PCIe bridges, NICs, and HBM DMA, to scale distributed inference workloads.

AI Ethernet Linux Machine Learning Networking +2 skills
 Sign in to see your match