• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Intel unveils it's 3D stacked transistor technology, increasing density 10 fold

Soldato
Joined
6 Feb 2019
Posts
19,402
Using RibbonFet technology on its 20a process node Intel believes it can achieve a 10 times increase in transistor density over current technology by stacking transistors.

It's early days, Intel is only planning to implement this technology into new CPUs between 2025 and 2027

 
Last edited:
Using RibbonFet technology on its 20a process node Intel believes it can achieve a 10 times increase in transistor density over current technology by stacking transistors.

It's early days, Intel is only planning to implement this technology into new CPUs between 2025 and 2027


So 2035 then going from Intel's history.
 
It’s interesting Intel have backtracked on 3D. Question is what changed their minds and why announce now.

Has something about about AMD’s 3D-V cache spooked them?
 
Amazing technology. Now that Intel's fixed it's 10nm ('Intel 7' process) I think it's onwards and upwards for Intel's fab business.

AMD would kill to have their own fabs on such a process.
 
I don’t think AMD would kill to have their own fabs just now. AMD’s focus is building cash reserves and stetting up a fabrication business would take decades to see a return. If AMD continue to grow over the next decade as they have over the past I could see the temptation to bring production in house or maybe do some kind of partnership with TSMC.

Can Intel compete with TSMC and how long it takes is the question. It make sense for Intel to invest in fabrication now, they have the cash and the push into GPU’s with TSMC should offset the future losses in its CPU business. Intel just need to be carful not to over produce when it’s new fabs are up and running.

It’s good that Intel has had change of mind on its 3D tech. The timing of the announcement and how far off it gate tech is from becoming a reality is odd though. The fabs required to use this tech aren’t even past the planning stages.
 
Last edited:
I don’t think AMD would kill to have their own fabs just now. AMD’s focus is building cash reserves and stetting up a fabrication business would take decades to see a return. If AMD continue to grow over the next decade as they have over the past I could see the temptation to bring production in house or maybe do some kind of partnership with TSMC.

Can Intel compete with TSMC and how long it takes is the question. It make sense for Intel to invest in fabrication now, they have the cash and the push into GPU’s with TSMC should offset the future losses in its CPU business. Intel just need to be carful not to over produce when it’s new fabs are up and running.

It’s good that Intel has had change of mind on its 3D tech. The timing of the announcement and how far off it gate tech is from becoming a reality is odd though. The fabs required to use this tech aren’t even past the planning stages.

I could see them more likely offering TSMC a partnership in a new fab, especially given that TSMC are looking to expand further out of Taiwan. That would be more cost effective than starting again from scratch.
 
I could see them more likely offering TSMC a partnership in a new fab, especially given that TSMC are looking to expand further out of Taiwan. That would be more cost effective than starting again from scratch.

I can see that happening. Intel will move away from TSMC and Apple only want the latest node every time. That leaves AMD being the main customer across many nodes.
 
lets hope it doesnt come with a 10 times increase in power consumption. not everyone has 3 phase.

I know we all go on about power but I used to have nearly 800 watts of lightbulbs running in the kitchen and dining room untill a few years ago
 
800 watts? That must be a hell of a kitchen and dining room you have.


not really considering I had 4 80watt spots , 3 smaller ones above the sink , and 4 tubes under the cuboards and 1 100watt down the end

to be fair it would mostly be the 4 halogen spots and the 100w running the rest was optional but still 420w

now its all led of course and uses bugger all
 
In saying that. Going from 100w incandescent bulbs to sub 10w led bulbs there doesnt seem to be much difference in electric bill. If anything its gone up ??
 
Deja Vu ?
Intel 2015 Press release regarding 3D Xpoint (yes a different tech, however relevant to huge claims)
The technology is up to 1,000 times faster and has up to 1,000 times greater endurance3 than NAND, and is 10 times denser than conventional memory.
https://newsroom.intel.com/news-releases/intel-and-micron-produce-breakthrough-memory-technology/

Whilst 3DXpoint imo is a interesting tech, and likely will continue to improve in time (I'd love some Terrabyte Dimms as an additional cache). However it's not close to the claims made. Nor even after the 5+ years since the announcement has much materialised.

So to conclude: Take such 10x... 1000x claims with a mountain of salt. And the timescale 5 years = 10-15 years. Based on historical announcements.
 
In saying that. Going from 100w incandescent bulbs to sub 10w led bulbs there doesnt seem to be much difference in electric bill. If anything its gone up ??

uh huh

I don't know what size home you have so I'll use mine as an example. We've got 24 bulbs across the two floors, if i assume that on average half of these bulbs ran for 3 hours each day and they went from 100w to 10w then ((12 * (100-10) * 3) * 365) /1000 = 1182kWh saving over 1 year.

There is no way you wouldn't notice that if you were paying attention to your usage. You say your bill has gone up, it's probably because you're using some other device more than before or the price per unit has increased giving you then illusion that the bulb swap did nothing
 
Back
Top Bottom