Daniele De Sensi

The eBPF eXpress Data Path (XDP) enables high-performance packet processing by executing programmable logic early in the Linux networking stack. However, XDP does not support packet duplication, a key requirement for publish-subscribe systems and high-performance computing (HPC) applications relying on message broadcasting. To overcome this limitation, developers typically use the Traffic Control (TC) layer, where the bpf_clone_redirect() helper allows packet replication. TC operates after socket buffer (sk_buff) allocation, a major source of overhead, and requires the creation of additional buffers for each duplicated packet. In this work, we present XDP_CLONE, a system that performs packet duplication directly within the network driver, eliminating socket buffer allocation. This approach significantly reduces overhead and improves performance for eBPF network applications up to 2.3x.

In recent years, GPUs have become the preferred accelerators for HPC and ML applications due to their parallelism and high memory bandwidth. While GPUs boost computation, inter-GPU communication can create scalability bottlenecks, especially as the number of GPUs per node and cluster grows. Traditionally, the CPU managed multi-GPU communication, but advancements in GPU-centric communication now challenge this CPU dominance by reducing its involvement, granting GPUs more autonomy in communication tasks, and addressing mismatches in multi-GPU communication and computation. This article provides a landscape of GPU-centric communication, focusing on vendor mechanisms and user-level library supports. It aims to clarify the complexities and diverse options in this field, define the terminology, and categorize existing approaches within and across nodes. The article discusses vendor-provided mechanisms for communication and memory management in multi-GPU execution and reviews major communication libraries, their benefits, challenges, and performance insights. Then, it explores key research paradigms, future outlooks, and open research questions. By extensively describing GPU-centric communication techniques across the software and hardware stacks, we provide researchers, programmers, engineers, and library designers insights on how to exploit multi-GPU systems at their best.

Cloud computing and AI workloads are driving unprecedented demand for efficient communication within and across datacenters. However, the coexistence of intra- and inter-datacenter traffic within datacenters plus the disparity between the RTTs of intra- and inter-datacenter networks complicates congestion management and traffic routing. Particularly, faster congestion responses of intra-datacenter traffic causes rate unfairness when competing with slower inter-datacenter flows. Additionally, inter-datacenter messages suffer from slow loss recovery and, thus, require reliability. Existing solutions overlook these challenges and handle inter- and intra-datacenter congestion with separate control loops or at different granularities. We propose Uno, a unified system for both inter- and intra-DC environments that integrates a transport protocol for rapid congestion reaction and fair rate control with a load balancing scheme that combines erasure coding and adaptive routing. Our findings show that Uno significantly improves the completion times of both inter- and intra-DC flows compared to state-of-the-art methods such as Gemini.

Efficient Reduce and AllReduce communication collectives are a critical cornerstone of high-performance computing (HPC) applications. We present the first systematic investigation of Reduce and AllReduce on the Cerebras Wafer-Scale Engine (WSE). This architecture has been shown to achieve unprecedented performance both for machine learning workloads and other computational problems like FFT. We introduce a performance model to estimate the execution time of algorithms on the WSE and validate our predictions experimentally for a wide range of input sizes. In addition to existing implementations, we design and implement several new algorithms specifically tailored to the architecture. Moreover, we establish a lower bound for the runtime of a Reduce operation on the WSE. Based on our model, we automatically generate code that achieves near-optimal performance across the whole range of input sizes. Experiments demonstrate that our new Reduce and AllReduce algorithms outperform the current vendor solution by up to 3.27×. Additionally, our model predicts performance with less than 4% error. The proposed communication collectives increase the range of HPC applications that can benefit from the high throughput of the WSE. Our model-driven methodology demonstrates a disciplined approach that can lead the way to further algorithmic advancements on wafer-scale architectures.

The allreduce operation is an essential building block for many distributed applications, ranging from the training of deep learning models to scientific computing. In an allreduce operation, data from multiple hosts is aggregated together and then broadcasted to each host participating in the operation. Allreduce performance can be improved by a factor of two by aggregating the data directly in the network. Switches aggregate data coming from multiple ports before forwarding the partially aggregated result to the next hop. In all existing solutions, each switch needs to know the ports from which it will receive the data to aggregate. However, this forces packets to traverse a predefined set of switches, making these solutions prone to congestion. For this reason, we design Canary, the first congestion-aware in-network allreduce algorithm. Canary uses load balancing algorithms to forward packets on the least congested paths. Because switches do not know from which ports they will receive the data to aggregate, they use timeouts to aggregate the data in a best-effort way. We develop a P4 Canary prototype and evaluate it on a Tofino switch. We then validate Canary through simulations on large networks, showing performance improvements up to 40% compared to the state-of-the-art.

Cloud computing represents an appealing opportunity for cost-effective deployment of HPC workloads on the best-fitting hardware. However, although cloud and on-premise HPC systems offer similar computational resources, their network architecture and performance may differ significantly. For example, these systems use fundamentally different network transport and routing protocols, which may introduce network noise that can eventually limit the application scaling. This work analyzes network performance, scalability, and cost of running HPC workloads on cloud systems. First, we consider latency, bandwidth, and collective communication patterns in detailed small-scale measurements, and then we simulate network performance at a larger scale. We validate our approach on four popular cloud providers and three on-premise HPC systems, showing that network (and also OS) noise can significantly impact performance and cost both at small and large scale. The full paper of this abstract can be found at https://doi.org/10.1145/3570609.