Teams at MulticoreWare

Develops and optimizes machine learning, computer vision, and numerical libraries for CPUs, GPUs, DSPs, FPGAs, and other accelerators. Responsibilities include algorithmic optimization, parallelization, vectorization, memory and cache optimization, benchmarking, profiling, performance analysis, customer collaboration, and coordination with software and hardware teams. Requires strong C/C++ and parallel programming experience, hardware architecture knowledge, and familiarity with FPGA development tools and AI/ML frameworks.

Lead the implementation, optimization, and delivery of machine learning, computer vision, and numeric libraries across heterogeneous hardware architectures. Own technical execution, quality, timelines, and customer outcomes while mentoring engineers in low-level optimization, parallel programming, vectorization, profiling, debugging, benchmarking, and performance tuning. Collaborate with customers and stakeholders to identify performance requirements and bottlenecks, make technical decisions, and deliver high-performance software solutions.

Develop and optimize deep learning models for efficient inference across CPUs, GPUs, accelerators, and edge devices. Implement quantization algorithms, model compression, and high-performance C++ kernels. Optimize transformer, LLM, MoE, and PEFT workloads for latency, throughput, and memory usage. Profile and benchmark models across hardware platforms, translate research into production solutions, and collaborate with machine learning, compiler, and hardware teams.