2017/09/14 03:11:20
Xavier Zepherious
 
if you read it at Semiaccurate
 

SemiAccurate digs out Intel’s 10nm process problems

https://semiaccurate.com/2017/09/11/semiaccurate-digs-intels-10nm-process-problems/
 
 
the real interesting stuff is easier to find here
https://www.reddit.com/r/AMD_Stock/comments/5k31sr/i_bought_canard_pc_there_you_go/
 
  • They say Intel's in the most precarious time ever in its existence.
  • Employees blame Krzanich as the main cause, he's impatient and eager and keeps changing/canceling projects before they get anywhere.
  • Many employees/engineers got fired in recent times and they say Krzanich treats engineers (replacing them) like supermarket cashiers, many new hires are inexperienced and not getting trained well enough. The mood in the company is at an all-time low and many employees are in fear for their jobs.
  • Murphy Renduchintala is said to actually be a glimmer of hope in the whole mess as since he's been brought on he's been able to focus R&D properly and actually has technical background and makes decisions that engineers are able to get behind on. Some say they're hoping he'll replace Krzanich as CEO once they hope he'll get the boot from the board.
  • 10nm is suffering all kinds of problems, both technical as well as management problems. The first samples that got in apparently were extremely disappointing in performance, not to talk about yields.
  • Kaby-Lake as well as Cannon-Lake were supposedly late designs just to save face and management was just hoping AMD wouldn't compete. Cannon-Lake is said to bring almost no architectural improvements.
  • 10nm will sample in late 2017 but production won't happen till 2018.
  • Intel has nothing concrete till 2019.
  • Long-term, Intel is said to be very worried about x86 as ARM is gaining ever more. If they can't keep Apple happy in the Macbooks that would be the first sign of the decline of x86.
  • Because of the above, they say that they have one CPU project which can do hybrid x86/ARM execution and they say they'll have the first prototype wafers with this within the next few months.
  • There's an Intel CPU AMD GPU MCM in the works.
 
 
 

Intel’s next generation chip plans: Ice Lake and a slow 10nm transition

https://arstechnica.com/g...-slow-10nm-transition/
 
---------------------------------------------
 
so there ya go..... like i said they were caught with their pants down
2017/09/14 04:42:44
PietroBR
Sad to see that, Intel, being the huge company it is, can't manage first the happiness of their employes, for then start showing some new results.
For me, looks like Coffee Lake might be the last build on Intel I will be doing if they can't organize thenselves.
 
In a few more years, AMD's CPU will have better compatibility and performance in games (since this is my main focus) than Intel. Other than this, we might just see 14nm++++++ process or new sockets from Intel.
2017/09/14 04:49:07
rjohnson11
You're not really posting an article from Charlie Demerjian are you? This guy posts anti-NVIDIA and anti-Intel article every chance he gets so the article should be taken with a grain of salt. In fact his articles angered so many people that he no longer allows anyone to post comments to the articles he writes. 
2017/09/14 06:47:10
redleader00
Intel is having problems with 10nm, same as everyone else.
2017/09/14 08:19:55
quadlatte
rjohnson11
You're not really posting an article from Charlie Demerjian are you? This guy posts anti-NVIDIA and anti-Intel article every chance he gets so the article should be taken with a grain of salt. In fact his articles angered so many people that he no longer allows anyone to post comments to the articles he writes. 


thats what i thought when i seen who wrote it, dude is the tech industries weekly world news tabloid, his stories have been proven false so many times that i dont know how anyone can even read a word he says 
2017/09/14 09:23:25
bcavnaugh
Looks like Intel is headed down the same road map as Microsoft, or maybe Microsoft has forced Intel down this Road Map.
2017/09/14 09:27:48
Xavier Zepherious
the information from canard PC is accurate - because they interviewed ex-Intel and working intel employees for the story. although none would go officially on record for fear of reprisal
 
as for Semiaccurate why do you not think I included the extra stories in for add credence to the story
 
Intel dropped the ball...TSMC and Samsung will surpass Intel if they don't get a handle on it
A bad CEO can do wonders to destroy a good company...intel needs new leadership
 
TSMC 7nm is equivalent to 10nm Intel or very close to
which is why ARM is a threat to intel under 7nm Finfet
2017/09/14 10:53:15
rjohnson11
Xavier Zepherious
the information from canard PC is accurate - because they interviewed ex-Intel and working intel employees for the story. although none would go officially on record for fear of reprisal
 
as for Semiaccurate why do you not think I included the extra stories in for add credence to the story
 
Intel dropped the ball...TSMC and Samsung will surpass Intel if they don't get a handle on it
A bad CEO can do wonders to destroy a good company...intel needs new leadership
 
TSMC 7nm is equivalent to 10nm Intel or very close to
which is why ARM is a threat to intel under 7nm Finfet


Samsung will also be at 7nm soon
2017/09/16 03:18:47
QuintLeo
Xavier Zepherious
 
  • They say Intel's in the most precarious time ever in its existence
  • Many employees/engineers got fired in recent times and they say Krzanich treats engineers (replacing them) like supermarket cashiers, many new hires are inexperienced and not getting trained well enough. The mood in the company is at an all-time low and many employees are in fear for their jobs.
  • There's an Intel CPU AMD GPU MCM in the works.
 

 
 
 They were WORSE off when Zilog started selling the Z80 - the only thing that saved Intel at THAT point was IBM deciding to use the 8088 in the IBM PC.
 Some of the same "engineers getting fired" or in more cases QUIT were what resulted in Zilog existing in the first place.
 
 They took significant hits from the original Athlon and the Athlon64 - they should EXPECT AMD to come up with something good every so often.
 Ryzen is NOT in the position of dominance (yet) that the Athlon64 was in - Intel had NO ANSWER AT ALL to that for almost 3 years, Ryzen they're still leading just by a very narrow margin.
 
 Intel GPU. Yawn. Intel has NEVER been a significant factor in GPUS, there have been times they didn't even crack the top *5* in graphic performance by a company (right now the only reason they MIGHT be in the top 5 is that there aren't many graphic companies LEFT).
 
 The transition to 10nm (much less the 7nm others have announced) is looking to be a slow one - and the next node after that is looking to be a nightmare for EVERYONE.
 
 Also keep in mind that what a lot of folks are CALLING "7nm" is probably very very close to what Intel is calling "10nm", based on a LOT of rumblings out of the industry.
 
2017/09/16 12:26:03
lehpron
Even semiaccurate rehashes their own articles, I've seen those points made before.
https://semiaccurate.com/...-intels-10nm-problems/

Even if those talking points of Intel were accurate (no pun intended), it's old news, I doubt those attitudes are still there now.  10nm is late, so much that Skylake is being refreshed twice as Kaby and Coffee.

The only reason Coffee is coming, if you think about it, is that there is no room in nomenclature series to put a 6-core into mainstream if 7700K and 7740K is top end quad, because 7800 is in HEDT.  It would have introduced a confusing overlap, so they introduce a new number sequence with it's own codename and gives them the opportunity to sell another chipset.  

Cannon should have been the 7-series after 6-series Sky, per early Client Group roadmaps with no delays.  Add my theory that Intel would have wanted to consolidate all unlocked to one platform while locking out Client, 6700K should have been the last unlocked SKU for the Z-chipset, if they didn't retire Z for just the H-series while putting all unlocked quads on up into X299.  They could all been called Core i7 if AMD never came back.

Furthermore and getting back to topic, the next lean x86 architecture is also delayed (possibly due to starting late, i.e. right after AMD revealed Zen in May 2015), which is why Cannonlake has a pair of refreshes at 10nm called Ice and Tiger-- I'm not dumb, those are both still lake architectures.  It's after Tiger we see something new, which I why I recommend no one upgrade to any Lake processor unless you already have one.  It'll be worth the wait, it's going to be like Core 2 again.  

How do I know?  Intel needs a wide berth from Zen and all iterations to allow the refresh cycle to start again, that's what Core (80% over Netburst), Nehalem (25% over Core) and Sandy Bridge (17% over Nehalem) gave Intel against AMD last time.  AMD took 5 years to come back, Intel wants that again.  This time they will try for an extreme boost like well over 50% IPC off of Skylake, I'm thinking 70%.  It will allow running lower frequencies and better binning with lower power, and still be better than last gen.  

I'd like to think AMD isn't sitting on their hands, even they said Zen would last four years, that gives Intel at least 1-2 years of dominance with their new arch; after that they better have something, but without Jim Keller, I don't know.  I figure Intel is betting a Bulldozer 2.0 will come after Zen in order to have their refresh rebrand cycle throughout the 2020's.

Use My Existing Forum Account

Use My Social Media Account