N

Neurophos

Optical metasurface AI accelerators

Principal Fabric/Network Architect

Full-Time
$250k - $300k/yr+ Stock options
Expert
Bachelor's, Master's, PhD
Austin, TX, USA+2 moreMore locations: San Mateo, CA, USA | Sunnyvale, CA, USA
In Person
Electrical EngineeringTelecommunications Engineering

About the job

Requirements
  • A Bachelor of Science, Master of Science, or PhD in Computer Science, Electrical Engineering, or a related field.
  • At least 10 years of experience in datacenter networking architecture, including RDMA transports deployed at scale.
  • Deep working knowledge of the Linux host networking stack, including kernel networking, VFIO, DPDK or libfabric, and IOMMU/IOVA behavior.
  • Fluency in PCIe, including root complex versus endpoint roles, bridges, peer-to-peer transactions, ACS, and Gen5/Gen6 signaling and protocol.
  • Familiarity with merchant NIC and switch silicon.
  • Ability to write architecture specifications that digital design and software teams can implement against.
  • Excellent communication skills and comfort leading cross-functional technical discussions.
Responsibilities
  • Own the end-to-end scale-out data path architecture spanning the host network stack, EIC PCIe bridge, NIC attachment, and DMA into HBM.
  • Define transport and offload architecture, including RoCEv2 and Ultra Ethernet Transport.
  • Specify peer-to-peer DMA topology and access control services policy across the PCIe hierarchy.
  • Define queue and completion semantics exposed to the runtime, in partnership with the software team.
  • Drive NIC vendor selection and own technical engagement with networking silicon partners.
  • Model and evaluate collective communication performance for distributed inference workloads.
  • Own the architectural case and roadmap for a future Neurophos DPU and for scale-up fabric in a later phase.
  • Represent Neurophos in relevant standards bodies and the data center networking community.
  • Establish networking architecture practices and mentor engineers across Architecture and Software.
Desired Qualifications
  • Participation in the Ultra Ethernet Consortium or a comparable standards body.
  • DPU or SmartNIC architecture experience.
  • Congestion control design and evaluation.
  • Experience bringing a novel accelerator into a production datacenter network.
  • Familiarity with collective communication libraries such as NCCL, RCCL, or equivalent, and distributed inference serving.
  • Understanding of machine learning inference workload characteristics and their network demands.
  • Experience with hyperscaler qualification and deployment processes.

About the company

Neurophos designs AI accelerator chips built with optical metasurfaces and silicon photonics to deliver high-density, energy-efficient AI computation for data centers. Its processors are designed to shrink optical components by about 8000x, increasing compute power per dollar spent and reducing both capital and operating expenditures. The product works by using optical processing techniques that handle AI inference workloads more efficiently than traditional GPUs or electronic AI accelerators, enabling faster throughput with lower power use. Neurophos differentiates itself by combining optical metasurfaces with silicon photonics to achieve higher density and energy efficiency, targeting data centers and enterprises that need scalable, cost-sensitive AI inference. The company aims to reshape the AI landscape by offering faster, denser, and more energy-efficient AI computation while lowering the total cost of ownership for data centers.

Company Size

51-200

Company Stage

Series A

Total Funding

$117.2M

Headquarters

Austin, Texas

Founded

2020

Get referred to Neurophos

See people who can refer or advise you

Simplify Jobs

Simplify's Take

What believers are saying

  • Neurophos raised $110 million in January 2026, led by Gates Frontier and M12.
  • By September 10, 2026, Neurophos said it quadrupled headcount and opened Silicon Valley.
  • Austin expansion and a San Francisco engineering site signal early customer demand and hiring momentum.

What critics are saying

  • Tulkas T100 remains in testing; mass production slips to after 2028.
  • Nvidia Vera Rubin and AMD keep tightening inference efficiency before Neurophos ships.
  • If 2028 silicon misses power claims, customers walk and the company becomes a science project.

What makes Neurophos unique

  • Neurophos uses metasurface photonic computing, shrinking optics 10,000x for chip-scale inference.
  • Its Tulkas T100 targets 470 peta-ops at 1-2 kilowatts, crushing GPU power density.
  • The company combines photonic hardware with a software stack for datacenter-ready OPU modules.

Help us improve and share your feedback! Did you find this helpful?

Benefits

  • Health Insurance
  • Unlimited Paid Time Off
  • 401(k) Company Match
  • Stock Options
  • Health Savings Account/Flexible Spending Account
  • Dental Insurance
  • Vision Insurance
  • Life Insurance

Growth & Insights and Company News

Headcount

6 month growth

8%

1 year growth

1%

2 year growth

6%
Archyde
Oct 10th, 2026
Neurophos raises $110M led by Gates Frontier for optical computing chips

Neurophos has raised $110M in a funding round led by Gates Frontier to build optical computing technology. The Austin, Texas, company uses silicon photonics circuits and metamaterials to run computations directly in the optical domain. M12 (Microsoft's venture fund), Carbon Direct Capital, Aramco Ventures, Bosch Ventures, Tectonic Ventures, and Space Capital also took part. The funding let the company quadruple its workforce in 2026. The approach targets the energy cost of data centres, where converting light to electrical signals and back wastes power. A hyperscale data centre uses 100MW to 1GW, about 50% of it for compute. "Moore's Law is slowing, but AI can't afford to wait," said chief executive officer and co-founder Dr. Patrick Bowen. Yole Group expects the first optical computing units to deploy in 2028, with the market reaching about $2.7 billion by 2034.

Third News
Sep 10th, 2026
Neurophos opens Silicon Valley office and expands team after $110M Series A round

Neurophos, a photonic AI chip developer, has opened a new Silicon Valley office and expanded its team following a $110 million Series A funding round announced earlier this year. The company appointed Dr Nathan Kundtz as chief technology officer and Dr Michael Shebanow as senior vice president of engineering. Kevin Keller joined as general counsel, whilst Preston Woo's role expanded to include chief commercialisation officer. Nigel Toon, former CEO of Graphcore, has joined Neurophos's board of directors. The company is expanding operations in Austin and Sunnyvale, with plans to launch a low-rate initial production facility in Austin. Neurophos is developing energy-efficient photonic chips for AI inference applications.

Eurekascoop
Jan 27th, 2026
Neurophos secures funding round to advance photonic AI chip

Neurophos secures funding round to advance photonic AI chip. Neurophos is a semiconductor startup based in Austin, that manufactures high powered, energy efficient photonic chips for AI inference. The company was established by Dr Patrick Bowen and Dr Andrew Traverso, who have gathered a professional team of experts with deep experience from major tech and hardware companies such NVDIA, Apple and Meta. Neurophos just secured a funding round from major investors including M12(Microsoft) to advance its photonic AI chip development. The company is addressing a major bottleneck in the AI industry. Traditional silicon based GPUs are struggling to support the demands of modern AI algorithms. To solve this Neurophos has developed a novel Optical Processing Unit (OPU). Unlike traditional chips, this chip uses light-based(photonic) computing and metamaterials to perform calculations. The company has been able to achieve massive miniaturization, they have managed to shrink optical components by 10000x, allowing them to fit over a million processing elements onto a single chip. The company has claimed that the chips have huge performance gains, their technology can provide up to 100x the performance and energy efficiency of traditional chips. Neurophos also claims that their chips are more sustainable, reducing the amount of electricity required for high level AI inference, hence reducing the carbon footprint of data centers. Neurophos also chips also boasts seamless integration, the chips are designed to be drop in replacements for existing GPUs, allowing data centers to easily upgrade their infrastructure. The new capital will be used to finalize the company's first integrated compute system, expand its software stack and open a new engineering facility in San Francisco. Investors and industry experts view this as a "physics-level" breakthrough that would be able to support modern AI models, as traditional Moore's law begins to plateau.

FOUNDERSMEDIA
Jan 25th, 2026
Aramco & Gates Frontier Lead USD 110M Round in US-Based Neurophos

Aramco & Gates Frontier lead USD 110M round in us-based Neurophos. * US-based Neurophos raised USD 110M in a Series A led by Gates Frontier, with participation from M12 and Aramco Ventures. * Founded in 2020, the startup is developing photonic AI inference chips aimed at cutting power consumption and easing data center energy constraints. * The funding will support launching its first photonic compute system and expanding operations from Austin to a new San Francisco site. Neurophos. US-based Neurophos, a photonic AI chip technology startup, raised USD 110 million in a Series A funding round. The round was led by Gates Frontier, with participation from M12 (Microsoft's Venture Fund) and Aramco Ventures. It also saw participation from Carbon Direct Capital, Bosch Ventures, Tectonic Ventures, Space Capital, and others. Founded in 2020 by Dr. Patrick Bowen and Dr. Andrew Traverso, Neurophos is an AI chip startup. It is focusing on developing high-performance, energy-efficient photonic AI inference chips. Its technology aims to enable significant reductions in power consumption. With this, it seeks to support the next generation of AI infrastructure without exponential increases in energy or physical resources. Moreover, it aims to facilitate making AI more accessible and cost-effective across industries. "Neurophos is addressing the only problem that really matters for the future of AI: the limits imposed by silicon. Their optical architecture provides the foundation for the next generation of machine intelligence," said Chris Alliegro, Managing Partner at MetaVC Partners. AI chip. Looking forward, with the new funding, it aims to accelerate the delivery of its first integrated photonic compute system. This includes datacenter-ready OPU modules, a full software stack, and early-access developer hardware. It also seeks to support its expansion from its Austin, Texas headquarters and open a San Francisco engineering site to meet early customer demand. "This physics-level shift means both efficiency and raw speed improve as we scale up, breaking free from the power walls that constrain traditional GPUs," said Dr. Patrick Bowen, CEO and Co-Founder of Neurophos.

FinSMEs
Jan 22nd, 2026
Neurophos Raises $110M in Series A Funding

Neurophos raises $110M in Series A funding. Neurophos, an Austin, TX-based company which specializes in photonic AI chip technology, has raised $110M in Series A funding. The round was led by Gates Frontier, with participation from M12 (Microsoft's Venture Fund), Carbon Direct Capital, Aramco Ventures, Bosch Ventures, Tectonic Ventures, Space Capital, DNX Ventures, Geometry, Alumni Ventures, MetaVC Partners, Morgan Creek Capital, Silicon Catalyst Ventures, Mana Ventures, Gaingels, and others. The company intends to use the funds to accelerate delivery of its first integrated photonic compute system, including datacenter-ready OPU modules, a full software stack, and early-access developer hardware. Led by CEO Dr. Patrick Bowen, Neurophos is a semiconductor company developing photonic AI inference chips. Its technology enables reductions in power consumption, supporting the AI infrastructure without the need for exponential increases in energy or physical resources. The company is expanding its Austin headquarters and opening a San Francisco engineering site to meet early customer demand.