Thread regarding Intel Corp. layoffs

You couldn't make it up

http://m.nasdaq.com/article/rumor-intel-corporations-coffee-lake-to-reuse-kaby-lake-architecture-cm711624

by
| 1519 views | | 9 replies (last November 21, 2016) | Reply
Post ID: @OP+KsQnOSe

9 replies (most recent on top)

Read @KtIDjc1-bpt comment on this thread

https://www.thelayoff.com/t/KtIDjc1

Who has a lifeboat and who believe in miracle on the titanic?

by
| | Reply
Post ID: @1yym+KsQnOSe

By many metrics 10nm from TSMC is matched to intel 14nm.

But TSMC 7nm is coming next year and already yielding SRAM and ARM cores!

Intel on the other hand just relased Kaby and Coffee is coming. First time in decades that intel has stayed on one process generation instead of moving to a new generation. Years ago moving from generation to generation faster was how they crushed all competion from AMD, Motorola, IBM, and everyone else!

These days they, Intel, have announced both 10 and 7nm are pushing out. First time intel is slowing! While TSMC and Samsung are accelerating their roadmap, sound like who is winning I ask you 😉

Intel stock will implode next year, short it or play options. TSMC silicon is on allocation even though their fabs are far bigger than intel. Google " Mediatek 10nm"

by
| | Reply
Post ID: @1ynu+KsQnOSe

@1pet, depends on whom you are asking and what metric you plan to use; according to Intel TMG senior leader, Yes, Intel 14nm has equivalent performance as TSMC 10nm. But industry wide typically compare two company/product based on same process node, like TSMC 16nm vs Intel 14nm; and TSMC 10nm vs Intel 10nm. which they can compare on different metrics.

by
| | Reply
Post ID: @1pfs+KsQnOSe

Question from a non-engineer, former Intel employee: is it true that Intel's latest 14nm process ("14nm+" I think they call it) provides equivalent performance to TSMC's 10nm process? Or does TSMC's 10nm outperform Intel's 14nm? I am trying to decide what to do with my Intel shares.

Thanks

by
| | Reply
Post ID: @1pet+KsQnOSe

6-core Kabylake? And almost 2 years needed to get that out to market? Laughable, Intel is a joke punch-line these days.

by
| | Reply
Post ID: @idz+KsQnOSe

don't worry @pjl everyone will get a prize just for showing up.

by
| | Reply
Post ID: @ylg+KsQnOSe

Intel will essentially have used the same basic Skylake processor and graphics architecture that first debuted in the fall of 2015 for three full product generations" -

What is wrong millennials??, no one holding your hand while you pee? You get what you deserve. Discriminate against the hard working OWM and replace them with lazy a-- millennials and poof...your company is junk bonds.

by
| | Reply
Post ID: @pjl+KsQnOSe

What's wrong with 10nm is PTD, there's your toxic culture right there. All the external abuse and internal corruption has finally caught up to them. Dedication and loyalty was about the only thing they had left and now that's gone too.

It's karma. Bad Karma. BKed again.

by
| | Reply
Post ID: @ixj+KsQnOSe

According to recently retired GM of TMG Holt nothing offers more value than scaling from Moore's Law. A third and fourth generation design on a mode means the Titanic has truely hit the iceberg and is sinking.

When you go to 10nm you get double the transistor density which enables chip size to get smaller to compensate for the increased cost of the process. 2x density of 10nm over 14nm allows you to do more cores and more GPUs and support more transistors while the overall chip size stays the same or gets smaller.

Scaling to 10nm allows you to improve transistor increasing performance, reduce power, and reducing interconnect distance all for better PPAC.

The original "Lakes" for sure baked in the planned /promised density and improved performance of 10nm to enable performance, power, area and cost, PPAC.

It isn't very hard to improve process performance at any node but you don't get area, leakage, capacitance or cost benefit. You add more cores or more transistors will make area and power go up far more than the small performance you get with the + process. Conclusion, the rumors of new architecture on 14nm is a wet dream told to senior executives, but you can't hit PPAC on 14 for a 10nm optimized design. The truth is coming!

This has three obvious implications

1) Coffee will be small increase In performance compared to what Canon might have been

2). Cost will be a huge issue as I predict the process window to deliver performance will be narrow. Little improvement, lots of smoke and mirrors and shortage of top parts on a huge die on 14nm.

3). Something must be wrong with 10nm to do a fourth generation on 14nm. This is almost exactly the fumble of AMD and the PPC. When process isn't there you are SOL. When you have process lead you hide a lot of things as intel had in the past , but all the rocks are getting exposed and the Titanic has struck them!

Moore's law is having your cake and eating it too. But if you don't scale than you eat the cake get fat and have a heart attack and die as someone else who can scale gets the cake and eats it too!

More ACTs are all but certain!

by
| | Reply
Post ID: @eqp+KsQnOSe

Post a reply

: