Post
Topic
Board Securities
Re: [LABCOIN] IPO [BTCT.CO] - Details/FAQ and Discussion (ASIC dev/sales/mining)
by
iCEBREAKER
on 10/09/2013, 16:38:10 UTC
You should wait and see what HashFast's next chip will do before offering erroneous advice.

They likely won't have another chip design for years.

They might have revisions of their current chip.  They might get slightly better performance. But it won't be worth it to do a 20nm chip or radically change the design, or something like that.

And anyway, I don't see why there's any reason to think IceDrill will have any "dibs" on chips that no one else will have access too. HashFast didn't even offer them a free MPP the way they did with the first Babyjet customers. So they actually got a worse deal then people who bought babyjets and will likely get extra compensating chips.

Ultimately, IceDrill is just one of HashFast's customers.

IceDrill is HashFast's largest "customer."  They have first dibs on bulk orders because they help with the NRE and whatever, making them more of a partner.

Quote
http://www.coindesk.com/hashfast-tapes-out-400-ghsec-28-nm-mining-chip/

The firm is already well into development for its second-generation chip, which uses a FinFet design. FinFet uses 3D structures, in which the microscopic transistors on the chip rise above the planar substrate on the chip. This gives them more volume than a traditional transistor gate, which enables them to conduct electricity more effectively, reducing the necessary voltage and increasing switching performance. The result is a faster, lower-power chip.

“It’s the most aggressive design available in our time schedules,” said Barber, who believes he’ll be taping out this chip in the fourth quarter. He wouldn’t say what process node the design used, but CoinDesk notes that last year, TMSC announced plans for a 16 nm FinFet capability in early 2014.

Good thing you chickened out of backing up your previous ignorance WRT HashFast/IceDrill's next chip!

"They likely won't have another chip design for years?"  LOL, shows what you know.  

"Won't be worth it to do a 20nm chip or radically change the design, or something like that?"  LOL, tell us another joke you fool.