Intel CPUs: Ten-year plan to includes 1.4nm and a two-year cadence
Click here to post a comment for Intel CPUs: Ten-year plan to includes 1.4nm and a two-year cadence on our message forum
ladcrooks
Hi Hilbert, and happy Xmas to you all. That is interesting and hard to get my head round. What wattage and what heat will be given in years to come? 😱
Cpu's have come a long way, well the whole tech has. A 286 and then an Athlon are few of the chips I have owned. My favourite Cpu for what it did has to be the i7920 @2.66. This in my mind was Intel's wonder chip, a 2.66Ghz that could go 4Ghz and higher. That was the OC's heaven. 😀
cryohellinc
Intel plans....
https://i.imgur.com/LOunszB.gif
Andy Watson
Perhaps that slide should be titled "No Moore we trust"
Seriously though, it's all well and good having these multiyear road plans but they are subject to change and are really just a best case scenario. This actually is more what the EUV manufacturer can provide than what Intel will be doing....
Kaarme
I kind of trust more in TSMC's plans than Intel's. Although one has to respect Intel engineers for being able to stretch the 14nm so far, as an achievement, not as a plan, especially since they didn't do much to the designs, sticking to the same old Core architecture all the way. In other words, they got the clocks so high they could still keep beating AMD in single-threaded performance, despite AMD refining a modern architecture on 7nm. In still other words you could say the Intel engineers could pull it off with both hands tied behind their backs by the not overly smart but all the greedier Intel execs and CEO (who have since been replaced, I believe).
sverek
Intel in 2030 : We put dot in "1.4nm" by mistake.
Also Intel : Instead let us introduce this awesome 14nm+++++++++++++++ Coke-Cola Lake!
Cyberdyne
Presumably Intel has (hopefully) learned a lot from 10 nm R&D and production. They seem to finally be over that hill, so maybe it will be much smoother for them to continue to shrink. Who knows.
Robbo9999
Well, if it's true that Intel can hold to this 10yr plan, then what they're basically saying is that they will need to move to a completely different technology after that 10yr point because they ain't gonna be shrinking it below 1.4nm using silicon. A fundamental change would need to happen from that 10yr point in order to see further improvments - either by linking together many individual chips (which is on the plans anyway before that 10yr time frame comes up), but there are limitations with linking chips, you can only link so many chips together until you start running out of physical space and cooling/power abilities. So it looks to me that Intel is committing to a completely different technology (away from silicon) sometime relatively soon after this 10yr plan - I think that's the upshot of all this. Therefore, perhaps in 2035 we won't be using silicon chip technology for cutting edge performance applications and it will have moved onto something else....to put that into perspective 2004 was not that long ago right!
toyo
This looks like a bunch of coping from Intel. As a happy 8700K owner, I don't trust them to deliver after their endless shitty 14nm++++++++++ crap.
Fewer words, more actual products, and shut that embarrassing marketing team and its crap.
insp1re2600
1.4nm Ricki Lake
Astyanax
lol, you're a good laugh.
nevcairiel
Their 10nm problems are the real crap, the amount of performance they managed to get out of their 14nm process to compensate for the 10nm problems is quite unparalleled in the industry and makes their 14nm one of the best nodes ever. Its just at the end of its lifetime now.
asturur
The fact that AMD can produce at 7nm, and reach or surpass a 14nm from intel, does not make 14nm process a crap.
You guys, again, you want a fast CPU, you do not want a piece of technology.
Some of you would buy a 5nm 386 (at least is 32bit ) if it was available.
DeskStar
HAAAAAAAAAAAAAHAHAHAHHAHHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHHAHAHAHAHAHAH
Silva
Oh Intel, make it 5 years gap, not 2.
H83
I think Intel is being too ambitious but okay...
JamesSneed
At least they have learned something from 10nm. Notice all that back porting? They are going to design new products on the old and new nodes which could have saved them if they had done the same with IceLake. Considering each node will be progressively more difficult and more importantly costly, this is a rather good plan. People are laughing since Intel was way to aggressive with densities on 10nm however I think Intel will be validated over time.
slyphnier
are u drunk ? when they even : https://www.tomshardware.com/news/intel-increasing-use-of-third-party-fabs-for-non-cpu-production-14nm-chip-shortage
manufacturing isnt all about ahead the race dude
also did u already read : https://www.techpowerup.com/259289/tsmc-trembles-under-7-nm-product-orders-increases-delivery-lead-times-threefold-could-hit-amd-product-availability
they behind the "nm-race" atm, but its not make their fabs stop producing
beside that without their fab, how many billions they spent for outsourcing their chips ?
Alessio1989
1.4nm.. in what numeric system? xD nm.. = nano-miles? xD
schmidtbag
Yeah I was wondering the same thing. Perhaps it's a different transistor structure, and works at an absurdly low voltage.
Robbo9999
However, TSMC's 5nm does not equal Intel's 5nm - the densities can be quite different (ie transistors per mm^2)..... currently Intel has a higher density per mm for any given Xnm node. But yes, I'm not saying that Intel will be able to match the race against other fabs though.
