News Intel CEO Gelsinger proposes a fab tour for Elon Musk — could be an attempt to win orders from Tesla, other Musk companies

Status
Not open for further replies.

Admin

Administrator
Staff member
I think that is a very clever move by Gelsinger.
Instead of having to outbid other fabs for orders, he can have potential customers compete for fab allotment.
However, this is assuming the node is competitive.

Hopefully IFS has something that sells better than Samsung's SF4X or SF3E.
 
Years ago, I recall Zuckerberg being given a fab tour. It sticks in my memory because, at one point, he did something dumb like remove his face mask in the clean room, and that ruined like a whole batch of wafers.
The wafer are almost always contained in Pods (Self contain isolated carriers), so highly unlikely it did anything but cause contamination control folks some consternation.

In a mask room, or other highly sensitive area that would be a big no no, but your general 300mm fab environment sees a lot worse day to day from technicians tear gloves and what not as they work to keep the machines running.
 
  • Like
Reactions: coolitic
Years ago, I recall Zuckerberg being given a fab tour. It sticks in my memory because, at one point, he did something dumb like remove his face mask in the clean room, and that ruined like a whole batch of wafers.

He did indeed. Several tools had to be wiped down to get them operational again. Obama just strolled in wearing his normal street clothes.

The worst, however, I can remember is about 20 years ago. One of the vendor field service engineers ended up getting his hand stuck in a 1st gen copper electroplate machine. The robot had pinched his hand and held it in place about half an inch above a bath of fairly high-concentrate sulfuric acid. The other guy working on the machine with him was fairly new, and the only thing he knew to do was what was taught to him in training - hit the big, red EMO button. The EMO basically powers the machine completely down in event of an emergency.

So now this poor guy has his hand smashed by a wafer handling robot so close to an acid bath that if he rests his fingers they're going into a vat of boiling acid. And the machine is now unable to be powered up to manually move the robot with the teach pendent. He's stuck.

The fire department was called in. They ran into the cleanroom in just their normal firefighting gear and used a sawz-all to chop the robot arm in half and release the FSE. It took almost 2 weeks to decontaminate the facility after that one.

Mr. FSE was largely unharmed and returned to work within a couple weeks.
 
  • Like
Reactions: bit_user
So now this poor guy has his hand smashed by a wafer handling robot so close to an acid bath that if he rests his fingers they're going into a vat of boiling acid. And the machine is now unable to be powered up to manually move the robot with the teach pendent. He's stuck.

...

Mr. FSE was largely unharmed and returned to work within a couple weeks.
Thanks for the amazing anecdote, but I wonder how his hand wasn't utterly cooked by having it stuck what sounds like just a couple cm above "boiling acid" for like an hour.
 
Tom's Hardware complains that posts are political ...
The forum policies are separate from the site's editorial policies. For further details, see:


Fab is labour intensive
See:

Building Fab overcapacity is ludicrous
According to Sam Altman, it won't be nearly enough!

You raise some valid points about the long-term prospects for US-designed semiconductors (regardless of where they're made, IMO). I think those points have nothing to do with trade sanctions on China. They were on a path to become technologically independent and competitive vs. US semiconductors, either way. The trade sanctions were a speed bump, in some respects, and an accelerant in others. Ignoring the problem, for so long, hasn't helped...

BTW, I think you're misreading their progress on "narrow gates". The 5 nm multi-patterning node was already "baked in the cake", by the time the sanctions went into effect. What they sanctions should do is stop them from achieving yet smaller nodes or single-pattern 5 nm.
 
Last edited:
Status
Not open for further replies.