What TSMC traditionally does is keep trailing edge fabs online that are fully depreciated and use those to produce chips that don’t need to be leading edge. It wouldn’t make sense to create a new fab for trailing edge chips.
Car manufacturers aren’t going to all of the sudden start using 2mm expensive chips for their cars.
Even for TVs, the BOM for the “smarts” need to be under $10.
"chose" is doing a lot of the heavy lifting here. Suppose you ran a Mastodon server and it turned out some people were using it to share revenge porn unbeknownst to you. Suppose further that they did it in a way that didn't make it easily detectable by you (eg. they did it in DMs/group chats). Sure, you can dump out the database and pore over everything just to be sure, but it's not like you're going to notice it day to day. If a few months later the revenge porn ring got busted should you be charged with "intentionally eavesdropping" on revenge porn or whatever? After all, to some extent, you "chose" to run the Mastodon server.
- You do not want the government to know which websites you visit. This rules out any kind of redirect / forwarding via a government website or app.
- You do not want websites to correlate their requests, as that would allow for cross-website tracking. Request data from website A should be completely useless to website B. This rules out most regular certificate schemes.
- You do not want a website to correlate multiple data requests, as that would allow websites to create some kind of supercookie. Requests should be completely independent, and two requests from the same user should be indistinguishable from requests from two different users.
- You do not want to lose privacy when the government and the website work together. The request should still be anonymous when the two collaborate, or else there can be no reasonable assumption of privacy. This rules out most clever pass-a-one-time-code schemes.
- You want the request to be unique and time-bound. It should not be possible to replay a response, either to the same website or a different one.
- You do not want to send more data than strictly necessary. If a website needs to know if you are 18 or older, it should only receive a boolean flag.
Getting some of those properties is easy. Getting all of them at the same time? Nearly impossible. And the worst part is that I almost certainly forgot a handful of requirements!
Edit: And as Doctorow points out there are a host of other issues that arise from actually deploying a working system.
Court documents says that they blocked access as soon as they were aware of it. They also "built out its systems to detect and filter out “potentially health-related terms.”". Are you expecting more, like some sort of KYC/audit regime before you could get any API key? Isn't that the exact sort of stuff people were railing against, because indie/OSS developers were being hassled by the play store to undergo expensive audits to get access to sensitive permissions?
1) How might Nvidia etc respond? They've made one-off SKUs for crypto, they could certainly respond quickly with a part that matched on memory but had much better software (meaning, more compatible with tools and better performance. AMD doesn't have the software, but their hardware is find and they could similarly up on-board memory. So Intel would really have to compete on price.
2) Ok, now we've found some 2nd or 3rd place success in a business built on logic fabbed at TSMC and DRAM from Samsung or Micron. If this is the future, why have fabs or any of the associated R&D?
I don't know what the right answers are but maintaining Intel at anything resembling its current size seems like a pretty tough puzzle.
Indeed. And his first action was to diss their own AI efforts. Because AI is just some niche area that they can ignore.
Just as Battlemage GPUs were getting decent reviews and sold above the MSRP.
To paraphrase, Intel has to go of the notion that for Intel to win AMD and TSMC have to lose. The strategy that follows from that might involve some painful choices.