ITHome reports that China’s latest Tianzhou-8 spacecraft was packaged with a space-based system powered by Loongson processors - reportedly fabbed using an unnamed “advanced” process node
Aren’t semiconductors for space use on older nodes because the priority is radiation hardening and reliability?
Pretty sure for outer space application they have to understand that the chip need to be shielded and withstand the radiation. Single bit flip if not handled correctly could result in chain reaction.
As long as they succeed in doing so. It is good for competition. But yeah without transparency/a publication, the claim can be iffy not to mention you can interpret “advanced” as coming from previous iteration.
I am not an expert on this, so feel free to correct me, but it is my understanding for high reliability scenarios you want to use older nodes were the quirks have been worked out.
First of all, I am not an expert (23, Mechatronics grad mostly focusing on Electronics and Hardware) all we can rely on are published documents that we can read.
As we delve into practical answers and from several references that I skim-read through, the correlation between older nodes and reliability exists however enhancing the fabrication process is possible.
This article provides a good write-up regarding the common knowledge of radiation and IC. Ultimately, we shouldn’t overthink what engineers are doing unless we also have a specific interest/being part of the project. Reading/Learning something is fun as long as you know the limit of what you can process at the time and at the same time have the limit of what you should read and should not. Again, I am not an expert; I only provide knowledge that is published and based on my limited understanding. Besides I found it interesting to dig for information from published documents :)
As much as skepticism, MIPS itself is old. Besides, good press (yet they don’t fully open the details) can boost their publicity (if you understand what I mean).
Aren’t semiconductors for space use on older nodes because the priority is radiation hardening and reliability?
Pretty sure for outer space application they have to understand that the chip need to be shielded and withstand the radiation. Single bit flip if not handled correctly could result in chain reaction.
As long as they succeed in doing so. It is good for competition. But yeah without transparency/a publication, the claim can be iffy not to mention you can interpret “advanced” as coming from previous iteration.
I am not an expert on this, so feel free to correct me, but it is my understanding for high reliability scenarios you want to use older nodes were the quirks have been worked out.
That’s why I somewhat sceptical about this.
First of all, I am not an expert (23, Mechatronics grad mostly focusing on Electronics and Hardware) all we can rely on are published documents that we can read.
As we delve into practical answers and from several references that I skim-read through, the correlation between older nodes and reliability exists however enhancing the fabrication process is possible.
This article provides a good write-up regarding the common knowledge of radiation and IC. Ultimately, we shouldn’t overthink what engineers are doing unless we also have a specific interest/being part of the project. Reading/Learning something is fun as long as you know the limit of what you can process at the time and at the same time have the limit of what you should read and should not. Again, I am not an expert; I only provide knowledge that is published and based on my limited understanding. Besides I found it interesting to dig for information from published documents :)
As much as skepticism, MIPS itself is old. Besides, good press (yet they don’t fully open the details) can boost their publicity (if you understand what I mean).
References:
Appendix Three: Space Radiation Effects on Integrated Circuits, NASA
Radiation effects in the electronics for CMS, Faccio, F., CERN
Common misconceptions about space-grade integrated circuits, Shunkov, V