Principal Software Engineer, Kernels
d-Matrix · Santa Clara
Infrastructureremoteprincipal$195k–$285k
At a glance
Highlights
- Help productize the software stack for next-gen AI compute engine.
- Build kernels bringing compute directly into memory for faster AI.
- Tackle AI infrastructure's hardest challenges from chips to rack-scale systems.
What they offer
- Equity
- Bonus
Ideal candidate
Veteran hardware-aware software leader energized by hard co-design problems and tight execution windows. Self-motivated owner who thrives collaborating across compiler, ML, and hardware teams.
Requirements
- MS in computer engineering, math, physics or related with 12+ years industry experience or PhD with 7+ years
- Strong grasp of computer architecture, data structures, system software and ML fundamentals
- Proficient in C/C++ and Python development in Linux environments
- Experience implementing algorithms for specialized hardware such as FPGAs, DSPs, GPUs and AI accelerators using CUDA
- Experience implementing ML operators such as GEMMs, convolutions, BLAS, SIMD ops for softmax, layer norm, pooling
- Experience with development for embedded SIMD vector processors such as Tensilica
- Experience with ML frameworks such as TensorFlow and/or PyTorch
- Experience with ML compilers such as MLIR, LLVM, TVM, Glow
- Prior startup, small team or incubation experience preferred
- Work experience at a cloud provider or AI compute/subsystem company preferred
- remote policy
- Hybrid, 3-5 days onsite per week
Lab profile
About d-Matrix
Pioneers memory-centric compute-in-memory chips and software for faster, energy-efficient AI inference.
What they build
- Digital in-memory compute inference chips
- Rack-scale AI inference systems
- AI inference software stack
Known for
- Compute-in-memory architecture
- Energy-efficient inference design
- Full-stack chips-to-systems expertise
Key facts
- Founded in 2019
- Headquartered in Santa Clara, California
- Backed by Temasek, Playground Global, and M12
Posted · Last seen