Our Portfolio Companies are Looking for YOU!

Applied Researcher

NeuroBlade

Ramat Gan, IL
  • Job Type: Full-Time
  • Function: Data Science
  • Industry: Semiconductors
  • Post Date: 01/12/2021
  • Website: neuroblade.ai
  • Company Address: Tel Aviv-Yafo

About NeuroBlade

Fast general purpose AI execution with no memory bottlenecks and no accuracy compromises.

Job Description

NeuroBlade is looking for a hands-on, highly capable algorithms & applied research engineer. We are looking for candidates that have experience in feature definition, performance modeling, and the ability to take complex research problem all the way to a proof of concept implementation on hardware

NeuroBlade set out on a mission to redefine computer architecture for memory intensive tasks. We build high performance solutions for the rapidly growing AI & Analytics market while lowering total cost of ownership. NeuroBlade’s unique hardware solution paired with a complete end-to-end SW stack, enables businesses to take the next leap forward by increasing the efficiency and affordability of their data-centers. 

What you’ll be doing:

Research, model and optimize data analytics algorithms on to Neuroblade’s HW
Work together with the system and architecture group to develop new features
Research and define as well as provide simulation level results, to optimized execution and data flows
Develop tools and simulations to model performance

Required:

Software development experience of 5+ years with proved track record
Algorithm development experience of 5+ years with proved track record
Analytical Database optimization experience of 5+ years with proved track record
Deep knowledge in one of the following fields: x86 ISA software optimization, Algorithm optimization for custom ASIC, Parallel Computing

Advantage:

Msc. Or Phd. Degree in CS
Research experience in one of the following fields: Optimization, Analytic Databases, Parallel Computing.
Experience in hardware optimized programming and parallel programming
Experience with storage systems