Stephen Nellis
@stephennellis.bsky.social
1.4K followers 38 following 21 posts
Reporter at Reuters covering semiconductors, hardware and other technology topics from Silicon Valley.
Posts Media Videos Starter Packs
stephennellis.bsky.social
Try a full reboot. If that doesn’t work, throw your laptop in the nearest body of water and leave it for 24 hours, then send to IT.
stephennellis.bsky.social
It’s all about how it’s managed by IT. A dedicated enough IT department can ruin an Apple computer, though usually not to the same degree. (I also had the Great Reuters Trackpad Freakout of 2024 yesterday..)
stephennellis.bsky.social
But the company is hopeful that it has found an approach that it will work. And that's why Google did not give a headline number of logical qubits, which they said would land between 10-20 with the current chip. But Google said it's more focused on finding an approach that will scale.
stephennellis.bsky.social
The key breakthrough was finding an array arrangement of qubits such that Google can continuously scale up the number of qubits and drive error rates down. Google say sit still has several orders of magnitude reduction to go before its next milestone.
stephennellis.bsky.social
I’m at the IEDM conference in San Francisco today and tomorrow - if you want to talk process tech, come find me!
stephennellis.bsky.social
Wow, that sucks! I should probably be more paranoid!
stephennellis.bsky.social
Oh no, that is awful! How did it happen? At a gig?
stephennellis.bsky.social
My biggest ever regret was selling the first really nice guitar I ever owned - an American Standard Strat in Lake Placid Blue - to help fund a Les Paul Classic during a heavy metal phase.
Reposted by Stephen Nellis
stephennellis.bsky.social
We have yet to see the last of the major grants, so we also do not yet know who else will get a haircut to adjust for this DoD contract. TSMC did not. Will Samsung? Will smaller firms get shaved? (end)
stephennellis.bsky.social
So Intel had been in line for $8.5 billion in grants. Instead, it ended up w/ a $7.86 billion grant, and a $3 billion DoD contract. We know very little about that contract (it's classified) so we don't know how comparable it is to the CHIPS Act money, or what Intel's costs will be to fulfill it.
stephennellis.bsky.social
...so a danger in having specialized "defense only" factories is they'll either go out of date very quickly compared to commercial fabs, and/or be incredibly expensive. I'm not here to settle that debate, but I will point out that it was controversial to see a defense program funded by CHIPS Act.
stephennellis.bsky.social
...or specialized facilities that only do defense work. As most folks in the chip industry will tell you, the demand from the defense sector is just not that large compared to the broader industry, and if you'll recall, the chip industry is all about scale....
stephennellis.bsky.social
This was framed in some quarters as a stumble for Intel, but I've been wondering whether it was instead a masterstroke of political engineering. There's been a long-running debate about whether it's better to build a defense supply chain from commercial fabs...
stephennellis.bsky.social
tap tap....this on? Hi, Stephen here, covering semiconductors and technology for Reuters.