weekly-papers-2026-08-08

Host CPU utilization over time for a staged agentic workflow, showing long low-utilization stretches punctuated by saturation spikes

Architectural Implications of Agentic AI Workflows

Weekly Paper Notes — one of the top picks from the 2026-08-08 CS paper digest. Area: Distributed Computing / Computer Architecture. Authors: Jirong Yang, Peizhe Liu, Jovan Stojkovic (UT Austin); Chaojie Zhang (Microsoft Azure) arXiv: 2608.04458 · PDF TL;DR Datacenter servers have been optimized for two workload shapes: CPU-centric services (web serving, key-value stores, analytics) and monolithic LLM inference, where a GPU does dense tensor math and the host merely feeds it....

Two CSP processes meeting at a rendezvous, with an alternative command below

Communicating Sequential Processes (Hoare, 1978)

Weekly Paper Notes — the Seminal Paper of the Week for the 2026-08-08 CS paper digest. Area: Programming Languages / Concurrency (rotating away from a long run of distributed-systems picks). Author: C. A. R. Hoare (Queen’s University, Belfast) Published: Communications of the ACM, Vol. 21, No. 8, August 1978, pp. 666–677 DOI: 10.1145/359576.359585 Why the paper still matters If you have written ch <- v in Go, spawned an Erlang process and sent it a message, wired up an actor in Akka, or reasoned about a Rust mpsc channel, you have been programming in a dialect of a language sketched in eleven pages of CACM in 1978....

Collective completion times for AllReduce, AllGather and AlltoAll on Clos and torus as node count scales

On Topology's Role in ML Training Performance

Weekly Paper Notes — one of the top picks from the 2026-08-08 CS paper digest. Area: Systems / Networking. Authors: Sarah McClure (UC Berkeley), Tegan Wilson (Northeastern), Brad Karp (UCL / Google), Michael Mitzenmacher (Harvard), Sylvia Ratnasamy (UC Berkeley), Scott Shenker (UC Berkeley / ICSI), Minlan Yu (Harvard) arXiv: 2608.01707 · PDF TL;DR Every large ML training system sits on one of two interconnect families: the fat-tree Clos that GPUs inherited from datacenter networking, or the torus that TPUs inherited from HPC....