Senior Performance Engineer, Inference

Cerebras Systems builds the world's largest AI chip, 56 times larger than GPUs. Our novel wafer‑scale architecture provides the AI compute power of dozens of GPUs on a single chip, with the programming simplicity of a single device. This approach allows Cerebras to deliver industry‑leading training and inference speeds and empowers machine learning users to effortlessly run large‑scale ML applications, without the hassle of managing hundreds of GPUs or TPUs.

Cerebras' current customers include top model labs, global enterprises, and cutting‑edge AI‑native startups. OpenAI recently announced a multi‑year partnership with Cerebras, to deploy 750 megawatts of scale, transforming key workloads with ultra high‑speed inference.

We are hiring a Senior Performance Engineer to join our Product team. You are an expert on state‑of‑the‑art inference performance and will serve as our resident expert on how Cerebras stacks up against alternative inference providers on both price and performance. This role sits at the intersection of performance benchmarking from first principles and competitive intelligence. The role has two core pillars:

• Performance Benchmarking:
  You will build, run, and maintain reproducible benchmarks that measure Cerebras inference performance for real customer workloads. This includes metrics like tokens per second, time to first token, latency under concurrency, and total cost of ownership (TCO).
• Competitive Pricing Intelligence:
  You will build and maintain a living model of competitor pricing across the AI inference landscape, including cloud providers, custom silicon vendors, and inference API platforms. You will work directly with our Sales and Product teams to translate this intelligence into pricing recommendations for enterprise contracts, ensuring Cerebras offers a compelling value proposition for every customer.

This role requires deep, hands‑on fluency with open‑source inference stacks (vLLM, SGLang, Tensor

RT‑LLM), GPU kernel‑level optimization tool chains (CUDA, Triton), and an intuitive understanding of how transformer architecture decisions—attention mechanisms, model sizing, quantization, KV‑cache strategies—interact with the realities of GPU memory hierarchies and compute budgets.

• Design standardized benchmark suites for inference workloads (code generation, summarization, multi‑turn conversation, agentic tool use) that enable fair, reproducible comparisons.
• Stay current with GPU optimization communities (CUDA, Triton, Tensor
  
  RT) and evaluate how new kernel fusions, flash‑attention variants, and quantization techniques shift performance ceilings.
• Build and continuously update a competitive pricing model covering token‑based pricing, throughput‑based pricing, and enterprise contract structures across major inference providers.
• Monitor industry announcements, pricing changes, and new product launches. Synthesize findings into actionable briefs for the Sales and Product teams.
• Partner with Sales to build deal‑specific competitive analyses showing total cost of ownership and performance advantages for enterprise prospects.
• Collaborate with Product and Engineering to identify where competitors are closing gaps or where Cerebras has underappreciated advantages.
• Track third‑party benchmarking sources (Artificial Analysis, Inference
  
  X) and ensure Cerebras is well‑represented and accurately measured.

Required

• Deep practical experience with state‑of‑the‑art open‑source inference frameworks like vLLM, SGLang, or Tensor
  
  RT‑LLM.
• 5+ years of experience in ML systems, ML research engineering, or high‑performance computing.
• Strong understanding of LLM inference economics: tokens, throughput, latency, batch sizes, precision trade‑offs, and how these translate to customer cost.
• Strong understanding of transformer model architecture internals such as attention mechanisms (MHA, MQA, GQA, MLA, DSA, MHA), KV‑cache management, and how each affect memory and compute profiles.
• Self‑directed and resourceful.

Preferred

• Background in ML research (publications or significant open‑source contributions) with a systems or efficiency focus.
• Contributions to open‑source…