Cerebras is developing the world’s first wafer-scale system. Since it was revealed in 2019, the CS-1 has continued to get favorable coverage from news outlets ranging from the Wall Street Journal to Anandtech. Further, one of our customers, Argonne National Labs, recently started applying the CS-1 to Covid-19 research. It is quite rewarding to be part of a team that works on state-of-the technology that can be applied to such an important problem.
The CS-1 is built from the ground up to accelerate deep learning in the data center. It is a complete solution for artificial intelligence compute powered by the Cerebras Wafer Scale Engine. It is programmable with the Cerebras software platform and package in an innovative system that fits directly into existing infrastructure.
Our system team is experienced and innovative. They think far outside of the box, which inevitably leads to new mechanical, thermal, and electrical challenges: nobody in the world has direct experience with our system bringup. We are therefore looking for an extraordinary individual with an outstanding track record for debugging and solving problems. This ‘MacGyver’ must have solved several unexpected problems.
Our work environment is rapidly changing as we are growing fast from the initial startup to a company manufacturing and already selling complete systems. We cannot predict exactly what you will be working on when you join. Below are some possible areas that you will be working on, but it could change the day you join.
This position is in the hardware team that is on the critical path of getting Cerebras to volume manufacturing: all eyes from engineering to management are on you. You will be given good mentoring, but you are also expected to start contributing right away. When you prove yourself, it will be rewarding in itself to have contributed to launching the first wafer scale company. In addition, there are multiple opportunities from silicon and system design to test and manufacturing operations that you could choose from.
- Electrical: with oscilloscope and DMM, study failure in chip or board. Then pursue failure analysis with local lab with scanning electron microscopy (SEM) and/or focused ion beam (FIB). Make recommendations to design/mfg groups
- Electrical/Script: With Python, get coordinates of pads from chip/board (>100k), and develop probe test to validate units before assembly.
- Script: with Python, study and organize data to facilitate data correlation and root cause of signatures. Provide input/suggestions to our data science group
- Mechanical: Study physical manufacturing tolerances of components, organize data, analyze variation and recommend update to test specs
- Manufacturing: Take on a repeating marginality in our manufacturing. Study, root cause, do experiments, and recommend improvements on current or new systems
- Strong analytical, diagnostic and problem-solving skills
- Strong drive to self-educate
- Experience with DC probing, oscilloscope, logic analyzers and Python
- Relevant experience in Computer, Electrical or Mechanical Engineering
- Proven track record of independently carrying out lab projects
- Written and verbal communication skills
- Headquarters/Los Altos Office