Chaitanya Kulkarni

With nearly two decades of experience in storage systems architecture, I bring deep expertise spanning the complete evolution of storage technologies—from traditional HDDs, SCSI/iSCSI, and SATA through the NVMe revolution, including NVMeOF, ZNS, and Live Migration. My hands-on experience encompasses the entire storage stack: systems software development, SSD controller design and implementation, storage virtualization, distributed storage-networking architectures, and high-performance clustered file systems.
As an NVMe Upstream Kernel Co-Maintainer, my most significant contribution to the NVMe ecosystem has been spearheading the industry-wide standardization of NVMe Live Migration. I led the development of the NVMe Live Migration Research Paper (August 2022) through multi-vendor collaboration, presenting this work at SNIA SDC 2022 and subsequently driving the specification changes through both the upstream Linux Kernel community and NVMe TWG. This two-year effort was recognized with the FMS 2024 Most Innovative Memory Technology Award and resulted in successful integration into the NVMe specification.
As an integral member of NVIDIA's Storage Next initiative, I currently focus on extending NVMe capabilities to support the high-concurrency, fine-grained I/O requirements of GPU-initiated storage workloads. My work addresses how AI workloads are fundamentally transforming data center storage requirements, developing AI-centric architectures that integrate NVMe with accelerated computing platforms including SSDs, GPUs, DPUs, and IPUs. Through active engagement with storage fabric protocols and contributions to cross-industry technical working groups, I work to advance storage technology standardization and build the broad-based consensus essential for the future of storage-centric AI workloads.