Discussion Intel current and future Lakes & Rapids thread

[jpiniero]

IEDM 2017: Intel details 22FFL, a relaxed 14nm process for foundry customers, targets mobile and RF apps

Trust me, there's no need to double post.

 

[Dayman1225]

Trust me, there's no need to double post.

Fair point, the article doesn't even mention 10nm so it's off topic too.

 

[IntelUser2000]

Cannonlake-Y coming, but in June of 2018.

 

[Yotsugi]

Cannonlake-Y coming, but in June of 2018.

And Cannonlake-U finally met the trashcan.
Oh well.

 

[IntelUser2000]

And Cannonlake-U finally met the trashcan.
Oh well.

Cannonlake-U was thought to have been canned months ago. The question was then about the Y.

 

[DrMrLordX]

Cannonlake-Y was supposed to be shipping right now. Good grief.

So are those the 2+2 chips or what?

 

[Yotsugi]

So are those the 2+2 chips or what?

Supposedly so.
What about gen10 though?
Will WHL-U be gen10?

 

[jpiniero]

Cannonlake-Y coming, but in June of 2018.

So Digitimes was right; since it'd take a month or two to get the chips into laptops and get them into the channel. Perhaps we now looking at maybe Q3 19 for Icelake U/Y and Q1 20 for everything else Icelake. I wonder if Intel is still releasing that 8 core CFL but it won't be released until Jan 19.

 

[IntelUser2000]

Cannonlake-Y was supposed to be shipping right now. Good grief.

So are those the 2+2 chips or what?

At 5.2W TDP, even with 10nm they can't go beyond 2 core without reducing clocks substantially. If its GT2 graphics, it'll use 40EUs, but again, I don't see the point of having that many resources at 5.2W. Unless 10nm is a surprise in that department and can take full advantage of full 40EUs. That's a stretch.

I see it unlikely we'll see Gen 10. They seem to be keeping things new all for Icelake. Perhaps we'll at least see Gen 10 media like Geminilake?


From Digitimes:

Some vendors are even considering skipping Cannon Lake to wait for the release of its successor, the Ice Lake CPUs, which according to Intel's roadmap, should be available shortly after the specific processors' rescheduled launch, the sources said.

This is what I mean with Whiskeylake. If Icelake is coming early Spring, manufacturers will skip Whiskeylake. Either that, or Icelake is more towards H2.

I wonder if Intel is still releasing that 8 core CFL but it won't be released until Jan 19.

This is why I think 8 core Coffeelake exists, despite their roadmaps to the contrary. Icelake will not be ready this year. For competitive reasons they need to get 8 core Coffeelake by end of 2018. The problem is they'll be squeezing two more cores on the same 14nm++ process.

Imaginary 3.5 year process lead is now zero.

 

[Yotsugi]

Q3 19 for Icelake U/Y and Q1 20 for everything else Icelake.

AMD will skin them alive for that.
It's suicide.

 

[IntelUser2000]

AMD will skin them alive for that.
It's suicide.

You are being dramatic.

In the days of Pentium III, the Athlon came out, and the downhill started for Intel. What no one realized was that their solution turned out to be none other than Netburst architecture. That, was their real mistake.

Will their Netburst of this generation turn out to be Skylake-based architectures and 10nm? Or will it start with Sapphire Rapid?

 

[Yotsugi]

You are being dramatic.

In the days of Pentium III, the Athlon came out, and the downhill started for Intel. What no one realized was that their solution turned out to be none other than Netburst architecture. That, was their real mistake.

Will their Netburst of this generation turn out to be Skylake-based architectures and 10nm? Or will it start with Sapphire Rapid?

Redacted, it's not about something being bad, it's about the thing being late.

No profanity in the tech forums.

AT Mod Usandthem

 

[IntelUser2000]

You seem to have an axe to grind based on your comments. Unlike on 1 on 1 conversations, use of profanity on posts suggest there's a significant amount of anger behind it.

 

[raghu78]

You are being dramatic.

In the days of Pentium III, the Athlon came out, and the downhill started for Intel. What no one realized was that their solution turned out to be none other than Netburst architecture. That, was their real mistake.

Will their Netburst of this generation turn out to be Skylake-based architectures and 10nm? Or will it start with Sapphire Rapid?

He might be dramatic but reality for Intel could be far worse. I think Intel has a few very tough years ahead. How they respond to the challenges will redefine themselves and their future as a IDM. I think AMD is going to take significant market share starting in 2018 in all segments - DT/NB/server . 2019 is when the screws will get turned on Intel as AMD has for the first time in a very long time access to a process node which is very competitive with Intel 10nm.

 

[Dayman1225]

Brian Krzanich abandons ship? Sell $54M worth of stock and options

ruh roh

 

[formulav8]

About time if he is leaving. He looks to be as bad as Hector Ruinz.

 

[Dayman1225]

About time if he is leaving. He looks to be as bad as Hector Ruinz.

I prefer Hector the Sector Rector, and for BK, Blunder King.

 

[raghu78]

About time if he is leaving. He looks to be as bad as Hector Ruinz.

imo BK is worse than Hector Ruiz. He will go down in history as the CEO who presided over the downfall of Intel as an IDM and a semiconductor manufacturing leader. Intel is going to go through a few very tough years before there is any hope of a turnaround.

 

[raghu78]

Brian Krzanich abandons ship? Sell $54M worth of stock and options

ruh roh

Got to cash those shares before it becomes a lot less valuable. I think by Q3 2018 it will be obvious which direction their stock is headed. Down, down and down.

 

[Dayman1225]

TigerLake has been added to Intel's website

 

[DrMrLordX]

In the days of Pentium III, the Athlon came out, and the downhill started for Intel. What no one realized was that their solution turned out to be none other than Netburst architecture. That, was their real mistake.

Will their Netburst of this generation turn out to be Skylake-based architectures and 10nm? Or will it start with Sapphire Rapid?

With the Pentium III, Intel couldn't figure out how to up clockspeeds. Took em awhile (Tualatin, Banias, Dothan, Yonah, Conroe, etc) to get back to it. I don't know that we'll see another Pentium III/Netburst moment. Besides, Sapphire Rapids may never make it to the desktop.

About time if he is leaving. He looks to be as bad as Hector Ruinz.

Maybe he'll try to spin off the fabs and take them along for the ride!

 

[IntelUser2000]

Besides, Sapphire Rapids may never make it to the desktop.

There was a German(not sure) site I gave a link to in one of these threads that showed Sapphire Rapid coming for client, just later than for Server.

With the Pentium III, Intel couldn't figure out how to up clockspeeds. Took em awhile (Tualatin, Banias, Dothan, Yonah, Conroe, etc) to get back to it.

It's not exactly the same of course. But that's how a repeat in history goes. Different actors, different backgrounds, but the struggles are similar. Technologies, actors change, but core human behavior stays the same.

With the problem they had back then, it is interesting to notice how they turned being unable to increase clocks by going all-in on a chip that was purely for clock speeds, which ended up in failing the goal anyway. The engineers, and managers despite their smarts and qualifications failed to see mere 5 years ahead.

 

[PingSpike]

Got to cash those shares before it becomes a lot less valuable. I think by Q3 2018 it will be obvious which direction their stock is headed. Down, down and down.

https://www.cnbc.com/2017/12/19/intel-ceo-in-memo-we-are-going-to-take-more-risks.html

Intel CEO to employees: 'We are going to take more risks'

• Intel's Brian Krzanich wrote in a memo to employees on Tuesday that the company faces "an exciting challenge" in areas where it's an underdog.

I assume by 'we' he means the employees. After selling the maximum amount of stock legally possible his risk profile is looking decidedly smaller at the moment.

 

[Phynaz]

Brian Krzanich abandons ship? Sell $54M worth of stock and options

ruh roh

The actuality -

https://www.sec.gov/Archives/edgar/data/50863/000112760217033679/xslF345X03/form4.xml

 

[Phynaz]

Got to cash those shares before it becomes a lot less valuable. I think by Q3 2018 it will be obvious which direction their stock is headed. Down, down and down.

Intel is doomed.