80 citations
- Massachusetts Institute of TechnologyUS2 papers
- Princeton UniversityUS2 papers
- Birla Institute of Technology and Science, PilaniIN1 paper
- Carnegie Mellon UniversityUS1 paper
- Cerebral Diagnostics (Canada)CA1 paper
- Columbia UniversityUS1 paper
- Duke UniversityUS1 paper
- ETH ZurichCH1 paper
- European Organization for Nuclear ResearchCH1 paper
- Fermi National Accelerator LaboratoryUS1 paper
- Google DeepMind (United Kingdom)GB1 paper
- Google (United States)US1 paper
4 papers · 1 filter
Beyond Exascale: Dataflow Domain Translation on a Cerebras Cluster
Tomas Oppelstrup, Nicholas Giamblanco, Delyan Z. Kalchev +5
Simulation of physical systems is essential across scientific and engineering domains. Commonly used domain decomposition methods are unable to simultaneously deliver both high sim…
Near-Optimal Wafer-Scale Reduce
Piotr Luczynski, Lukas Gianinazzi, Patrick Iff +3
Efficient Reduce and AllReduce communication collectives are a critical cornerstone of high-performance computing (HPC) applications. We present the first systematic investigation…
Wafer-Scale Fast Fourier Transforms
Marcelo Orenes-Vera, Ilya Sharapov, Robert Schreiber +3
We have implemented fast Fourier transforms for one, two, and three-dimensional arrays on the Cerebras CS-2, a system whose memory and processing elements reside on a single silico…
Dynamic Control Flow in Large-Scale Machine Learning
Yuan Yu, Martín Abadi, Paul Barham +12
Many recent machine learning models rely on fine-grained dynamic control flow for training and inference. In particular, models based on recurrent neural networks and on reinforcem…