ML Hardware Accelerator Modeling, Lightmatter
$180-225k
Light-powered AI computing
Job no longer available
Light-powered AI computing
101-200 employees
Job no longer available
$180-225k
101-200 employees
Company mission
Lightmatter's mission is to build an engine that will power discoveries, drive progress, and redefine what computers and human beings are capable of.
Role
Who you are
- BS or higher in computer/software engineering, electrical engineering, or related field
- 5+ years of industry experience in mapping algorithms to hardware accelerators, hardware modeling, computer architecture, or related fields
- Proficient with C++ and Python
- Strong understanding of computer architecture
Desirable
- Familiarity with machine learning workloads
- Experience in mapping software algorithms to hardware accelerators
- Experience with modeling and analysis of machine learning accelerator architectures
- Proficient developing in a Linux-based environment
- Experience working with or modeling distributed hardware and workloads
What the job involves
- Design, development, and maintenance of functional and performance models of upcoming hardware
- Participate in the co-design process by working closely with hardware teams to iterate on early performance models, adding fidelity as designs progress
- Collaborate with machine learning teams to construct and evaluate the performance of machine learning workloads
- Build cycle approximate performance and bit-accurate functional models that further the co-design process and allow the software teams to build and optimize our next generation software stack
- Collaborate with Design Verification teams to validate hardware against the simulators
- Contribute to a system-level simulator that integrates the accelerator simulator into a larger system simulator
- Support the development and optimization of our software stack by diagnosing and fixing issues across the simulator and software stack
Otta's take
Xav Kearney
CTO of Otta
Neural networks and CPUs are vital to AI solutions, but manufacturers have largely failed to maintain innovation in recent years. Lightmatter solves this problem by offering a next-generation computing platform powered by photonics (light waves).
Lightmatter offers a forward-facing platform for AI-building that acts as an alternative to traditional solutions. It works by replacing wired CPUs with a combination of photonics and new algorithms. Lightmatter's solution is not only faster, but it is considerably more sustainable, and it has the potential to drive the next generation of AI development.
As of 2022, Lightmatter has raised $113 million across three funding rounds, which is testament to growing industry demand for faster, more effective AI-building solutions. Lightmatter currently has three products on the market - Envise, Passage, and Idiom - and is poised to take on the incumbent chiplet industry with a photonics-based alternative.
Insights
Company
Funding (last 2 of 5 rounds)
Dec 2023
$155m
SERIES C
Jun 2023
$154m
SERIES C
Total funding: $422m
Company benefits
- Commuter benefit
- Happy hour -Spend time together and build friendships
- Generous vacation policy and holidays
- 401(K)
- Opportunity to learn from your peers and teach them what you know across electronics, photonics, machine learning, and more
Company values
- WORK WIDE & PARALLEL: Working on hard problems, we have to go down many different paths quickly to find the right one.
- QUESTION EVERYTHING: People have opinions about what things work and what things don’t work. They’re not always right.
- LET FIRST PRINCIPLES GUIDE YOU: Stand behind statements that you know are true and build something amazing.
- DON'T BE DEFINED BY YOUR JOB TITLE: Many great ideas occur at the boundaries between disciplines.
- BE KIND: Help build the environment you want to work in.
- TAKE RESPONSIBILITY FOR YOUR WORK: Everyone’s contributions are critical to the company’s success.
Company HQ
Financial District, Boston, MA
Articles
Founders
Nicholas Harris
(CEO)Previously worked as an R&D Product Engineer at Micron Technology, and as a Postdoctoral Fellow at MIT.
Thomas Graham
(COO)Was an M&A Analyst at Morgan Stanley, and worked as an Agency Business Development Manager at Google.
Darius Bunandar
(Chief Scientist)Worked as a Research Collaborator at Caltech, and was a PhD Candidate at MIT.
Salary benchmarks
We don't have enough data yet to provide salary benchmarks for this role.
Submit your salary to help other candidates with crowdsourced salary estimates.
Share this job
View 9 more jobs at Lightmatter