XTier improves memory speed by smartly managing DRAM and CXL tiers
xTier: Intelligent Tiering for CXL-Enabled Memory
Operating Systems
Summary
Servers use two types of memory: fast DRAM and slower CXL memory, and need to decide where to keep data for best speed. The authors created xTier, a system that lives inside the operating system kernel and learns which memory pages to put in DRAM or CXL by watching how programs use memory. xTier reacts quickly to changes in program behavior, moving fewer pages around while keeping systems fast. Tests showed it works well, especially when there is less fast DRAM available.
What this means in practice
- •For data center operators: Improve server memory performance and reduce data movement by dynamically placing memory pages across DRAM and CXL in real time.
- •For cloud infrastructure teams: Manage memory tiering efficiently for various workloads to maintain speed while using large amounts of slower but cheaper CXL memory.
Authors
Sriranga Ramaswamy, Yueqi Chen
Abstract
CXL-enabled memory expands server memory capacity, but introduces a page-placement problem: the operating system must decide which pages should reside in DRAM and which should reside on slower CXL memory. Existing systems make this tradeoff in one of two ways. Userspace controllers support flexible policies, but expose placement decisions to scheduler jitter and kernel-userspace crossing overhead. Kernel-space systems avoid this latency, but rely on fixed heuristics that must generalize across workloads. We present xTier, a kernel-resident learned memory-tiering system. xTier attaches eBPF programs to PEBS events and uses a compact quantized MLP to score sampled pages inside the kernel at microsecond-scale latency. Rather than reacting to every candidate, xTier converges to a low-churn placement for the current workload phase, reduces sampling cost after convergence, and returns to a higher sampling cadence when the workload shifts. We evaluate xTier on six memory-bound workloads at DRAM:CXL ratios from 1:5 to 1:25. The advantage grows as the DRAM budget tightens. At 1:15 and beyond, xTier is the fastest system in 14 of 18 configurations. Where it is not fastest, it trails the best baseline by 3.9% on average. It reaches this performance while moving 13% fewer pages in geometric mean, and 22% fewer at the tighter ratios. When a workload changes phase, xTier rebuilds its hot set in DRAM faster and more completely than any baseline.