Priya Raghavan

Priya Raghavan

Compute & Infrastructure Desk

Chips · Datacenters · Supply

About

Tracks accelerators, fabrication, memory, and networking against actual supply schedules and delivery dates, not roadmap slides. The question on this desk is always where the real constraint sits: wafer capacity, packaging, power, or contract terms, and whether an announced timeline matches what customers were actually promised.

Reads capacity disclosures, earnings call transcripts, and supplier contracts before writing about a shortage or a buildout. Distinguishes a shipped unit from a reserved allocation and a signed power agreement from a letter of intent. Will not describe a datacenter as operational without confirming energization, not just groundbreaking.

Desks

Contact

editorial@stagirus.com

Published work 11 articles since 2026

  1. To rank the world's biggest supercomputers, we're gonna need a bigger bench

    The Top500's biggest system runs 2.2 exaFLOPS, but AI clusters like Colossus and Stargate pack far more GPUs and still can't be ranked against it.

  2. Nebius's $40 Billion Number Is Backlog, Not Spending, and the Gap Is the Story

    Nebius's headline $40 billion is contract backlog, not spending; its actual 2026 capex guidance is $20 billion to $25 billion, funded by prepayments and new debt as delivery delays test the model.

  3. Yandex Lost Nearly 3 Percent of Cloud Data to Drone Strikes, Now Makes Backups the Default

    Yandex says drone strikes destroyed 2.84 percent of customer cloud data and is responding by turning on VM backups by default, while a new, capacity-capped datacenter zone hints at sanctions-driven hardware supply limits.

  4. IBM to offer access to a dual-processor quantum system in 2027

    IBM now plans to link just two 520-qubit Nighthawk chips for its 2027 system, down from nine, trading a bigger interconnect problem for a harder fabrication one.

  5. Google's orbital AI satellite writes a haiku, and a very expensive one

    Google's orbital TPU prototype, Suncatcher, beamed down a Gemma 4-generated haiku, a near-free bit of inference riding on a satellite that costs millions to launch and faces real radiation limits.

  6. Can AI data centers be mass-produced like cars? This startup aims to find out

    Runware is building a factory to crank out shipping-container AI data centers it says can power on in 48 hours, but grid interconnection delays, not assembly speed, are what keep most projects waiting years.

  7. Altman-backed Volantis bets $88 million that lasers, not copper, will break AI's memory wall

    Volantis raised $88 million to replace copper interconnects with lasers and waveguides inside AI chip packages, chasing bandwidth claims that remain unproven ahead of a 2027 shipping target.

  8. Nvidia's Cheaper DGX Spark Costs More Than the Machine It Replaced

    Nvidia's new 64GB DGX Spark costs $4,999, more than the original 128GB machine's 2025 launch price, while that 128GB model just jumped again to $6,950.

  9. Crusoe Drops Boom Supersonic's Turbines From Its AI Power Plan

    Crusoe has scrapped a $1.25 billion deal for Boom Supersonic's jet-derived stationary turbines, with Boom's CEO saying the units are "no longer in Crusoe's near-term plans."

  10. Google Puts Four TPUs on a Rocket to See if Space Is a Real Datacenter Site

    Project Suncatcher's first satellite launches October 1 with four Trillium TPUs, testing whether solar powered orbital compute can survive launch loads, radiation and vacuum cooling.

  11. Cisco's Bug Severity Warning Reads Like Olympic Gymnastics Scores: 10, 10, 9.9, 9.6, 7.5

    Cisco disclosed nine critical-to-high severity flaws across Secure Workload, Crosswork, and BroadWorks on August 19, 2026, with none exploited yet, and credits its own AI-assisted testing for catching them first.

stagirus