Intel is skipping 10nm and going straight to 7nm, competition = nom nom for my tummy?

[Space Lynx]

https://www.techspot.com/news/79131-intel-cpu-shortages-worsen-second-quarter.html

Intel previously said it would begin mass production of 10nm Ice Lake parts in the second half of 2019 although according to Digitimes Research’s upstream supply chain sources, Intel’s production schedule for the 10nm process still has lots of issues… so many that Intel may skip over 10nm entirely and dive into development of the 7nm process.

article also mentions AMD taking a heavy bite out of Intel's market share - so Intel is def going into hyperdrive mode I think hehehehehe this is great news for all of us. though, I am tired of waiting so I will be doing a 3700x probably this summer.

 

[Metroid]

I am tired of waiting so I will be doing a 3700x probably this summer.

Count me in, the 3700x will seal intel's fate, 3 to 6 years for intel to recoup this extreme direct attack.

 

[R0H1T]

It's an open secret that 10nm is a virtual dead end. The question for Intel is whether 7nm will be good enough, given their aggressive node shrinks or whether they're jumping on the marketing bandwagon w/7nm ala TSMC or Sammy? Lastly the most important factor - yields!

 

[GreiverBlade]

Count me in, the 3700x will seal intel's fate, 3 to 6 years for intel to recoup this extreme direct attack.

nr.3 for me ... them disabling OC on my 6600K was the last drop (even if it was by mistake)

 

[FordGT90Concept]

They should have made this call years ago. It's like NetBurst all over again. Intel likes beating dead horses.

 

[R0H1T]

What do you mean disabling OC on 6600k

 

[las]

Would be nice. Glad I'm not playing the waiting game tho.

 

[eidairaman1]

What do you mean disabling OC on 6600k

Could be a MC update

 

[John Naylor]

Oh darn.... I was so hoping on freeing up those few nm in my PC so I could install bigger radiators.

 

[eidairaman1]

You can only hope they dont have milling problems at 7nm even...

 

[TheoneandonlyMrK]

It took even less time for it to go as i said then i expected , They were never going to compete with 7nm at 10nm,, from a marketing POV, it wasnt going to work.

plus they're optimistic they'll get their chipzilla title back from samsung this year ,but realise it wont be an easy sell post Ryzen.

 

[eidairaman1]

It took even less time for it to go as i said then i expected , They were never going to compete with 7nm at 10nm,, from a marketing POV, it wasnt going to work.

plus they're optimistic they'll get their chipzilla title back from samsung this year ,but realise it wont be an easy sell post Ryzen.

Advanced Micro Devices is working on this current phase of Ryzen 2, Ryzen 3 might be a design change, but socket compatible with Gen 1 and 2 boards. Ryzen+ is 1.5.

 

[SL2]

Intel could end up skipping over 10nm and diving right in to 7nm development

Intel’s production schedule for the 10nm process still has lots of issues… so many that Intel may skip over 10nm entirely and dive into development of the 7nm process.

Digitimes Research believes Intel could shift its investments directly to 7nm process development, skipping 10nm.

Speculations? FUD?

Intel has had severe issues for years with more than one process node, that's not really news. The first 10 nm was launched half a year ago, although a crippled one.

 

[kastriot]

Too late, too late intel you should pull up your head from ass long time ago not now.

 

[R0H1T]

We're talking about mass market chips here & 10nm is pretty much dead for those! Intel has released CNL U but it's so bad that the IGP is still not functional

 

[SL2]

I mean, I wouldn't be surprised if Intel actually skipped 10 nm, but I need to see more evidence than speculations.

June 2018: https://www.reddit.com/r/AMD_Stock/comments/8t8yvw
October 2018: https://arstechnica.com/gadgets/201...ancelling-its-10nm-process-intel-no-were-not/

We're talking about mass market chips here & 10nm is pretty much dead for those! Intel has released CNL U but it's so bad that the IGP is still not functional

I'm not so sure about that, 10 nm looks very competitive on paper. That GPU could most likely get fixed in time for mobile Ice Lake in Q4.

 

[R0H1T]

It's competitive on paper because there's barely 50% of the chip working, when in fact 100% was the intended target.

 

[SL2]

It's competitive on paper because there's barely 50% of the chip working, when in fact 100% was the intended target.

No, I meant process node specs. It's closer to TSMC's 7 nm than you might think.

 

[Space Lynx]

Would be nice. Glad I'm not playing the waiting game tho.

you have a 9900k on custom water at 5,2, your pretty much set for ten years unless we get a big IPC jump lol - your statement is a null statement, of course someone who has that kind of setup doesn't have to play the waiting game.

 

[eidairaman1]

10 is 10, 7 is 7

 

[R0H1T]

No, I meant process node specs. It's closer to TSMC's 7 nm than you might think.

Sure & that's why I said that 7nm from Intel might be more of a marketing term as TSMC & Samsung do with their nodes. The process would be dense but how much wrt their own 10nm or the process characteristics - I'd say Intel would probably go on the conservative side this time around.

 

[hat]

10 is 10, 7 is 7

Not quite. "10nm" and "7nm" are mostly marketing terms. The actual measurements such as fin pitch and whatnot are pretty close, despite 7nm being "30% smaller" than 10nm.

Of course, I'm no semiconductor engineer, but that's what I've read time and time again in these discussions about Intel's 10nm. The most important difference, though, is TSMC's 7nm works, and Intel's 10nm doesn't...

 

[las]

you have a 9900k on custom water at 5,2, your pretty much set for ten years unless we get a big IPC jump lol - your statement is a null statement, of course someone who has that kind of setup doesn't have to play the waiting game.

I were playing the waiting game for a few years tho, I came from 3770K @ 4.8 GHz, shortly had a Ryzen 2700 @ 4.15 GHz and then jumped on 9900K on release.

Still looking forward to see what Zen 2 has to offer. I'll get Zen 2 in PS5 when it launches I guess

I fear that Zen 2 is mostly more cores. AMD needs a decent clockspeed gain to really bring it. Especially for gaming. Many gamers, pretty much all high refresh rate gamers, still only considers Intel at this point.
In many games my 2700 @ 4.15 GHz performed worse than my 3770K @ 4.8 GHz. I was targetting 100+ fps, hence not 100% GPU bound like some people might be.

I'll probably pick up B550 + Ryzen 3600/3700 for my homeserver at some point.

 

[notb]

Next node (lets call it 7nm) must have been in development for a while now. 10nm is already over a year behind the schedule. 7nm team may have more luck and deliver on time (2021) or even earlier.
So yeah, if that's the case, why not? Intel can certainly sail on the 14nm+++ for another year. They may be able to squeeze 10 cores on consumer socket (with or without IGP) which will be competitive against AMD 16 cores (if no serious IPC gain happens there).

Even if this scenario resulted in another 5% market share lost for a year, it will likely be better than putting even more money into 10nm development and building semiconductor plants for a process that will be obsolete within 2 years.

And yeah... I'm pretty sure they're also developing a CPU that could be made on TSMC or Samsung 7nm (instead or alongside the 10nm). Because why not?

Also, they may have simply learned how much improvement TSMC 7nm gives AMD and are not impressed. We've all seen Radeon VII... if 7nm EUV isn't much better, 14nm++++++ will be good enough.

 

[las]

Next node (lets call it 7nm) must have been in development for a while now. 10nm is already over a year behind the schedule. 7nm team may have more luck and deliver on time (2021) or even earlier.
So yeah, if that's the case, why not? Intel can certainly sail on the 14nm+++ for another year. They may be able to squeeze 10 cores on consumer socket (with or without IGP) which will be competitive against AMD 16 cores (if no serious IPC gain happens there).

Even if this scenario resulted in another 5% market share lost for a year, it will likely be better than putting even more money into 10nm development and building semiconductor plants for a process that will be obsolete within 2 years.

And yeah... I'm pretty sure they're also developing a CPU that could be made on TSMC or Samsung 7nm (instead or alongside the 10nm). Because why not?

Also, they may have simply learned how much improvement TSMC 7nm gives AMD and are not impressed. We've all seen Radeon VII... if 7nm EUV isn't much better, 14nm++++++ will be good enough.

I agree, more cores alone won't put much heat on Intel.
For the most part, consumers don't really need more than 6C/12T. Maybe next-gen consoles will change this fact, but I don't expect it to change for 2-3 years tho.

I really hope that Zen 2 does 4.5 GHz single core boost out of the box and reach 4.6-4.8 on all cores. This would put Intel under more pressure and even high fps gamers would maybe start considering AMD as an option.

On another note, AMD needs more mobile market share. Intel dominates this market. Just like the enterprise market (I feel like AMD is gaining ground here tho).