everyone is pushing forward; if you don't, you're basically giving up and quitting the business!
-
-
Replying to @FioraAeterna @Pinboard
cost per node kinda varies. so far I *think* 28nm was the best cost per transistor; 14nm went up, but prices always drop over time
1 reply 1 retweet 0 likes -
Replying to @FioraAeterna @Pinboard
here's a chart from 2014ishpic.twitter.com/jCNpAM6UDX
2 replies 1 retweet 3 likes -
Replying to @FioraAeterna
thank you! Yeah, the cost I have in mind is the cost of going from research to a factory making chips with the new process
1 reply 0 retweets 0 likes -
Replying to @Pinboard
those went up mainly because multi-patterning means you need a lot more masks. I think they estimate ~80 masks per chip at 7nm
1 reply 0 retweets 0 likes -
Replying to @FioraAeterna
I’m curious if there will be a situation where it’s physically possible to do another node, but economically unappetizing
1 reply 0 retweets 0 likes -
Replying to @Pinboard
it kinda already happened to a degree with GPUs; desktop GPUs moved to 14nm very late in large part because their dies are so big
1 reply 0 retweets 3 likes -
Replying to @FioraAeterna @Pinboard
whereas mobile *really really needs* the low power, so it moved to 14nm as soon as possible, since it was more valuable there
2 replies 0 retweets 5 likes -
Replying to @FioraAeterna
Interesting. At some point I would like to work out a knowledge transfer with you where I ply you with beer and questions
1 reply 0 retweets 1 like -
sweet. Let me ping you next time I visit my beloved servers nearby. I have so many stupid questions!
-
-
-
Replying to @FioraAeterna
the look on your face when you find out I’m a bot
1 reply 1 retweet 7 likes - Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.