Monday, October 04
Bite Me Edition
- New research suggest that the Wuhan Bat Flu may have come from the Wuhan Bat Market and not from the Wuhan Bat Virus Factory across the road according to a paper published by the Wuhan Bat Virus Factory. (LA Times / Yahoo News)
A second paper from France that - the article kind of slides past this - has not actually been published anywhere concurs.
The article is 90% just the LA Times sniping at CNN, but that has a certain charm in itself.
- M1X Arm-based MacBook Pros may be arriving later this month. (WCCFTech)
The rumoured specs are inviting. Shame about the operating system, the price, the fact that it's all glued together into an unmaintainable brick, and that the company behind it basically just plain sucks.
- The chip shortage will continue until morale improves. (ZDNet)
If you're waiting for prices to come down then congratulations, you can keep right on doing that.
- Why carmakers can't just update to newer chips. (Jalopnik)
In short: Because cars are expected to work.
- Ransomware gangs are getting ripped off by rival gangs that are undercutting their prices. (ZDNet)
The free market at work.
- JMS is working on a reboot of Babylon 5. (Variety)
Look, it's his baby, and the original production was, shall we say, not untroubled. If he wants to try making it again, and he has funding, let's see how it comes out.
And I do like the nearly-requiring-a-fainting-couch assertion that power window chips (?) are on the decrepit 45- or even 90-nm nodes. Nobody tell them that there are plenty of chips still made on 130nm and larger. I've been looking to see what the most recent generation of Arm Cortex-M microcontrollers use and I haven't found any, but commonly-used ones like STM32F4s and the SAMD21 and '51 used in Arduinos and Adafruit boards and the like seem to be 130nm or even larger. (There's a Wikipedia article that purports to list all the active foundries in the world, and none of the Microchip foundries (that make the SAMDs) are smaller than 130nm, according to that article. (I use the Cortex-Ms as sort of a proxy for things like ECUs based on an article I read a few months ago, that I can't find any more, that gave an approximation of node, transistor counts, and clock speed, that were in the ballpark of those ARM controllers.)
Posted by: Rick C at Tuesday, October 05 2021 12:06 AM (Z0GF0)
Posted by: Rick C at Tuesday, October 05 2021 12:10 AM (Z0GF0)
Posted by: normal at Tuesday, October 05 2021 04:01 AM (LADmw)
Of course, given the relative lack of success of his foray into writing comics, perhaps this is for the best.
Posted by: cxt217 at Tuesday, October 05 2021 11:36 AM (MuaLM)
And, I find myself wanting to rave at the 'update the chips in the cars' folks. Do they get how complicated cars are? Do they understand what that means for design projects? How awful that makes part inventories and logistics at the volumes they work at? Do they understand data on service life, mechanical reliability, or quality control in manufacturing? Do they even understand the significance of the wiring harness, and how terrible they are?
It is the 21st century, and it is fine if we still use analog circuits for some things. It is fine if we still use vacuum tubes for some things. If we have a lot of old equipment, parts are still cheap, and we have a lot of people who can maintain it, we do not necessarily have to upgrade to something 'better'. We do not need for everything to have a network connection all the way to H3ll and back, so that the gremlins can work from the comfort of home.
Okay, that is a bit hypocritical coming from me. I do hang around some people who are into some digital and solid state stuff, and they have a point about it being cool.
Posted by: PatBuckman at Tuesday, October 05 2021 01:28 PM (r9O5h)
No. Or, rather, they don't realize the amount of testing and validation necessary to be sure that putative new designs work and are safe and reliable. The article, to its credit, touches on this.
As proof of my claim, I submit the Intel idiot's comment about selling them all the 16nm chips they want. Intel, you will recall, is the company that was stuck at 14 for like 5 years, and it took 3 generations on 10nm--laptop only, and the first generation sucked (and consisted of one single chip, which was only ever sold in cheap laptops in China), and the second wasn't very good--before they are finally, this fall, ready to release (hopefully) good desktop chips at that node. They have no excuse for not realizing how hard it is.
Posted by: Rick C at Tuesday, October 05 2021 02:10 PM (Z0GF0)
Posted by: Rick C at Tuesday, October 05 2021 02:16 PM (Z0GF0)
Posted by: Rick C at Tuesday, October 05 2021 02:17 PM (Z0GF0)
Posted by: Pixy Misa at Tuesday, October 05 2021 04:51 PM (PiXy!)
Posted by: StargazerA5 at Wednesday, October 06 2021 02:56 AM (uii8N)
58 queries taking 0.2564 seconds, 342 records returned.
Powered by Minx 1.1.6c-pink.