Ray is a powerful distributed computing framework. However, as data sets grow and computation requirements become more complex, managing memory usage across multiple computing nodes becomes increasingly challenging. Issues that slow down performance include the data copying between the computing nodes, data spilling out of memory into storage, and the data skew among computing nodes. We'll introduce Gismo, a multi-node shared memory object store based on Compute Express Link (CXL) technology to address this challenge. With the help of Gismo, Ray does not need to serialize and transfer the extra copies of the objects across network. Data spill and skew issues will be minimized because each host has direct memory access to the whole object store. This talk will demonstrate how Gismo is integrated with Ray and showcase how it improves overall performance and reduces memory overhead for Ray users.
Charles Fan is co-founder and CEO of MemVerge. MemVerge delivers software-defined memory service to applications in a multi-cloud environment. Before MemVerge, Charles led VMware’s storage business unit and the big data group and developed industry-leading products including Virtual SAN. Before VMware, Charles was a co-founder and CTO of Rainfinity and was the inventor of its file virtualization and high availability technology. Rainfinity was acquired by EMC. Charles received his Ph.D. from Caltech doing research in distributed computing and got his undergrad degree from Cooper Union.
Come connect with the global community of thinkers and disruptors who are building and deploying the next generation of AI and ML applications.