About the job
Cerebras Systems is at the forefront of AI technology, having developed the world's largest AI chip, which is 56 times larger than traditional GPUs. Our innovative wafer-scale architecture delivers the computational power equivalent to dozens of GPUs on a single chip while maintaining the programming simplicity of a single device. This unique approach enables Cerebras to provide unparalleled training and inference speeds, allowing machine learning practitioners to seamlessly run large-scale ML applications without the complexities of managing numerous GPUs or TPUs.
Cerebras proudly serves a diverse clientele, including leading model labs, global enterprises, and pioneering AI-native startups. Notably, OpenAI has recently formed a multi-year partnership with Cerebras to harness 750 megawatts of scale, revolutionizing key workloads with ultra high-speed inference.
Our groundbreaking wafer-scale architecture ensures that Cerebras Inference stands as the world's fastest solution for Generative AI inference, achieving speeds over ten times faster than GPU-based hyperscale cloud inference services. This remarkable increase in speed is transforming the user experience of AI applications, enabling real-time iterations and enhancing intelligence through additional agentic computation.
About The Role
Cerebras is expanding its Machine Learning team to spearhead a new initiative that aligns with our existing teams. We are seeking a Principal Investigator to collaborate with our ML leaders in shaping this new effort while building the team and enhancing our capabilities. This new team will work in concert with our current ML divisions: Field ML, which directly engages with customers, Applied ML, which develops new ML capabilities and applications, and Core ML, which adapts ML algorithms to leverage the unique features of Cerebras hardware. The new team may undertake similar or complementary responsibilities.
The new team will focus on areas such as:
- Post-training and reinforcement learning: Enhancing model deployment quality through advanced training, tuning, and reinforcement learning techniques, concentrating on specific downstream tasks;
- Dataset curation and optimization: Implementing strategies to gather and select high-quality data, facilitating quicker and higher-quality model training and tuning;
- LLM Pretraining: Engaging in...
