16 February 2024 in Proceedings of the ACM on Measurement and Analysis of Computing Systems
Processing-in-memory (PIM), where the compute is moved closer to the memory or the data, has been widely explored to accelerate emerging workloads. Recently, different PIM-based systems have been announced by memory vendors to minimize data movement and improve performance as well as energy efficiency. One critical component of PIM is the large amount of compute parallelism provided across many PIM "nodes'' or the compute units near the memory. In this work, we provide an extensive evaluation and analysis of real PIM systems based on UPMEM PIM. We show that while there are benefits of PIM, there are also scalability challenges and limitations as the number of PIM nodes increases. In particular, we show how collective communications that are commonly found in many kernels/workloads can be problematic for PIM systems. To evaluate the impact of collective communication in PIM architectures, we provide an in-depth analysis of two workloads on the UPMEM PIM system that utilize representative common collective communication patterns -- AllReduce and All-to-All communication. Specifically, we evaluate 1) embedding tables that are commonly used in recommendation systems that require AllReduce and 2) the Number Theoretic Transform (NTT) kernel which is a critical component of Fully Homomorphic Encryption (FHE) that requires All-to-All communication. We analyze the performance benefits of these workloads and show how they can be efficiently mapped to the PIM architecture through alternative data partitioning. However, since each PIM compute unit can only access its local memory, when communication is necessary between PIM nodes (or remote data is needed), communication between the compute units must be done through the host CPU, thereby severely hampering application performance. To increase the scalability (or applicability) of PIM to future workloads, we make the case for how future PIM architectures need efficient communication or interconnection networks between the PIM nodes that require both hardware and software support.
11 June 2024 in ACM SIGMETRICS Performance Evaluation Review
Processing-in-memory (PIM) has been widely explored in academia and industry to accelerate numerous workloads. By reducing the data movement and increasing parallelism, PIM offers great performance and energy efficiency. A large amount of cores or nodes present in PIM provide massive parallelism and compute throughput; however, this also proposes challenges and limitations for some workloads. In this work, we provide an extensive evaluation and analysis of a real PIM system from UPMEM. We specifically target emerging workloads featuring collective communication, demonstrating its role as the primary limitation within current PIM architecture.
28 August 2024 in IEEE Computer Architecture Letters
Deep learning recommendation model (DLRM) is an important class of deep learning networks that are commonly used in many applications. DRLM presents unique challenges, especially for scale-out training since it not only has compute and memory-intensive components but the communication between the multiple GPU is also on the critical path. In this work, we propose how cold data in DLRM embedding tables can be exploited to propose proactive embedding. In particular, proactive embedding allows embedding table accesses to be done in advance to reduce the impact of the memory access latency by overlapping the embedding access with communication. Our analysis of proactive embedding demonstrates that it can improve overall training performance by 46%.