Greenland Ditches Starlink for French Satellite Service
dagens.com
Evaluating the Infinity Cache in AMD Strix Halo
chipsandcheese.com
MinIO (apparently) becomes source-only
github.com
Show HN: Cadence – A Guitar Theory App
cadenceguitar.com
The Gypsy Life of Robert Louis Stevenson
hudsonreview.com
Go Subtleties You May Not Know
harrisoncramer.me
rlsw – Raylib software OpenGL renderer in less than 5k LOC
github.com
Show HN: Modshim – A new alternative to monkey-patching in Python
github.com
Ask HN: Our AWS account got compromised after their outage
Knocker, a knock based access control system for your homelab
github.com
Neural audio codecs: how to get audio into LLMs
kyutai.org
Replacing a $3000/mo Heroku bill with a $55/mo server
disco.cloud
NASA chief suggests SpaceX may be booted from moon mission
cnn.com
Evaluating Argon2 Adoption and Effectiveness in Real-World Software
arxiv.org
The Hidden Engineering of Niagara Falls
practical.engineering
Power over Ethernet (PoE) basics and beyond
edn.com
Mathematicians have found a hidden 'reset button' for undoing rotation
newscientist.com
Getting DeepSeek-OCR working on an Nvidia Spark via brute force with Claude Code
simonwillison.net
Erowid - Documenting the Complex Relationship Between Humans and Psychoactives
erowid.org
Researchers complete first human trial on viability of enteral ventilation
newatlas.com
Obviously cache snooping is going to be device-dependent, but the way this shows things happening is that if you write an address that exists in another CPU's cache, that the invalidation doesn't happen until that CPU later tries to read the address. That requires two snoop cycles be dedicated, which is a waste. The secondary CPU is already listening to the snoop bus, it can/should (and in practice does on basically all devices, AFAIK) be able to recognize that write fly by and free up its cache line preemptively.