AI Systems Performance Engineer - New Graduate
Core
Bring up and optimize state-of-the-art foundation models on SambaNova's reconfigurable dataflow platform to achieve high throughput and low latency.
Role type
New graduate AI Systems Performance Engineer
Builds
High-performance AI inference systems on SambaNova hardware
Domain
Generative AI / High-performance computing
Deliverable
production ML models
Required skills
Python, C++, algorithms, data structures, computer architecture, operating systems, parallel computing, deep learning, PyTorch/TensorFlow/JAX
Preferred skills
LLMs, transformer architectures, model inference, KV cache, quantization, distributed execution, CUDA/Triton/OpenCL, vLLM/DeepSpeed/Megatron/TensorRT, memory hierarchy, profiling
Technologies
SambaNova Suite, SN40L chip, PyTorch, TensorFlow, JAX, CUDA, Triton, OpenCL, vLLM, DeepSpeed, Megatron, TensorRT
Responsibilities
Bring up cutting-edge foundation models on the SambaNova platform; Analyze and profile model execution to identify performance bottlenecks; Optimize AI workloads for throughput, latency, memory efficiency, and scalability; Collaborate with ML, compiler, runtime, and hardware engineers; Explore and integrate new techniques in model architecture, quantization, scheduling, and caching; Develop tools, benchmarks, and performance analysis methodologies; Investigate new model architectures and translate research advances into efficient implementations
Seniority
New graduate, hands-on IC