NVIDIA's Multi-Instance GPU (MIG) technology enables partitioning GPU computing power and memory into separate hardware instances, providing complete isolation including compute resources, caches, and memory. However, prior work identifies that MIG does not extend to partitioning the last-level TLB (i.e., L3 TLB), which remains shared among all instances. To enhance TLB reach, NVIDIA GPUs reorganized the TLB structure with 16 sub-entries in each L3 TLB entry that have a one-to-one mapping to the address translations for 16 pages of size 64KB located within the same 1MB aligned range. Our comprehensive investigation of address translation efficiency in MIG identifies two main issues caused by L3 TLB sharing interference: (i) it results in performance degradation for co-running applications, and (ii) TLB sub-entries are not fully utilized before eviction. Based on this observation, we propose STAR to improve the utilization of TLB sub-entries through dynamic sharing of TLB entries across multiple base addresses. STAR evaluates TLB entries based on their sub-entry utilization to optimize address translation storage, dynamically adjusting between a shared and non-shared status to cater to current demand. We show that STAR improves overall performance by an average of 30.2% across various multi-tenant workloads.

翻译：NVIDIA的多实例GPU（MIG）技术能够将GPU计算能力和内存划分为独立的硬件实例，提供包括计算资源、缓存和内存在内的完全隔离。然而，先前的研究发现MIG并未扩展至最后一级TLB（即L3 TLB）的划分，该TLB仍由所有实例共享。为增强TLB覆盖范围，NVIDIA GPU重组了TLB结构，在每个L3 TLB条目中设置16个子条目，这些子条目与位于同一1MB对齐范围内、大小为64KB的16个页面的地址转换一一对应。我们对MIG中地址转换效率的全面研究发现，L3 TLB共享干扰导致两个主要问题：（i）降低并发应用的性能表现，且（ii）子条目在驱逐前未被充分利用。基于这一观察，我们提出STAR方法，通过跨多个基地址动态共享TLB条目来提高TLB子条目的利用率。STAR根据子条目利用率评估TLB条目以优化地址转换存储，在共享与非共享状态间动态调整以适应当前需求。实验表明，STAR在各种多租户工作负载下平均提升整体性能30.2%。