We are seeking a talented Code Generation and Optimization Expert to join our bleeding-edge CodeGen team in Israel. In this high-visibility, hands-on role, you will play a pivotal part in building next-generation runtime compiler.

Responsibilities
Design and maintain the distributed and heterogeneous executable that is generated by our compiler stack.
Using the MLIR framework, transform high level compiler outputs into hardware-specific binary images, memory mappings, execution parameters, resource allocation, region grouping, and cross-domain coordination for distributed systems.
Design and maintain the API layer (libRT.a) connecting compiler-generated code with runtime services, ensuring seamless integration across different execution domains.
Collaborate closely with hardware, architecture, verification and other compiler teams to align software with hardware requirements and behavior.

Our team is looking for a skilled professional to join the AI software group. In this role, you will focus on our AI kernel compilers, ensuring our hardware delivers peak optimization for sophisticated and custom AI kernels. You will work extensively with various kernel infrastructures, including Triton and CUDA, while collaborating closely with our AI graph compiler teams.



Responsibilities
Develop our AI kernel compilers and related components in the SW stack.
Drive the performance optimization of AI models running on our unique architecture.

We are seeking a talented engineer to work on our AI software stack-specifically, the low-level optimization of AI models and kernels for our unique hardware architecture.

Responsibilities
Develop our optimized AI libraries and related components in the SW stack.
Drive the performance optimization of AI models running on our unique architecture.

we are looking for talented networking software engineers to develop an innovative new communication stack. Were starting from the ground up, rethinking existing protocols and network accelerators while designing both hardware and software for a new low-latency compute platform. We care about results and insist on doing things right. Our engineers are not siloed into narrow areas; our ideas are cross-fertilized among teams. Our technical backgrounds are diverse and we don't have preconceived approaches or cookbooks, so come and invent! The hardware teams are just a few steps away and really like hearing our ideas. This is a unique opportunity to develop creative solutions from concept to implementation.

Responsibilities
Take part in developing a networking stack from the ground up:

Analyze customer applications and explore creative ways to accelerate them in hardware or software
Find the best ways to process packets on hardware platform
Participate in architecture of future NIC generations
Develop user-space libraries and benchmarks for low-latency communication
Work closely with the hardware teams to offload key features into the next generation of products