Towards Disaggregation-Native Data Streaming between Devices

TL;DR

This paper proposes a disaggregation-native data streaming approach that allows direct device-to-device data flows, reducing latency in datacenter workloads by bypassing CPU staging.

Contribution

It introduces a novel data streaming facility that enables direct device-to-device communication in disaggregated datacenter architectures.

Findings

Reduced processing latencies in datacenter workloads

Enabling direct device-to-device data flows improves efficiency

Supports flexible resource pooling in disaggregated systems

Abstract

Disaggregation is an ongoing trend to increase flexibility in datacenters. With interconnect technologies like CXL, pools of CPUs, accelerators, and memory can be connected via a datacenter fabric. Applications can then pick from those pools the resources necessary for their specific workload. However, this vision becomes less clear when we consider data movement. Workloads often require data to be streamed through chains of multiple devices, but typically, these data streams physically do not directly flow device-to-device, but are staged in memory by a CPU hosting device protocol logic. We show that augmenting devices with a disaggregation-native and device-independent data streaming facility can improve processing latencies by enabling data flows directly between arbitrary devices.