The Grand Tour of Affinity — First Stop: Locality and Memory CachesWe are exploring the topic of affinity in computing — In this first installment we look at the principle of locality and cache hierarchiesFeb 21Feb 21
Design insights for programmable, high performance datapathsWe discuss the seven core design insights that motivated our architecture for programmable datapaths with high performance.Jan 262Jan 262
Monotonic State Variables: The fast and furious of IPC communicationsWe talk about Monotonic State, MSVs, Variables and introduce Stateful MSVs — these are techniques for super fast IPC communications.Jan 193Jan 193
Byte Queue Limits: the unauthorized biographyByte Queue Limits is one of the tools in the fight against bufferbloat. It determines precise limits one device queues to minimize queuing.Jan 123Jan 123
Hardware FIFOs — Accelerator FIFOs editionWe present the concepts and a design for “accelerator FIFOs” — the world’s fastest form of IPC for low level hardware components and CPUs!Jan 6Jan 6
Hardware FIFOs — Device queues editionWe explore the wonder of device queues that are hardware FIFOs commonly used in devices like NICs to facilitate host-device communicationsDec 29, 20243Dec 29, 20243
S’more ‘bout FIFOsThe two FIFO conditions, full and empty, are dominant considerations in FIFO design — we identify the challenges and explore solutions.Dec 22, 20243Dec 22, 20243
Everything you need to know about FIFOsFIFOs are a core data structure in computing and even in real life in the form of lines. We’ll delve into FIFOs and how to make them fly!Dec 16, 20244Dec 16, 20244
Privacy in Internet Addressing (or lack thereof)We talk about the state and particulars about privacy in Internet addressingDec 8, 2024Dec 8, 2024