"Moore’s Law, the empirical observation that the number of transistors in a computer chip doubles approximately every two years.
The number of researchers required to double chip density today is 18 times larger than those required in the early 1970s. In the case of Moore’s Law, research productivity has been declining at a rate of about 6.8% per year."
There's far more than 18 times people using IT nowadays than there were in 1970s so all is fine.
Stop your whining and smoke some, bitch.
Henry Hughes
Research is now about corporativism and diversity issues. Take google for example, half of their staff are either lazy ass overqualified PHDs or degenerate liberal scum students who earn too much to do trivial shit.
Tech is not about progress anymore, its about money. Anything goes when money is the game.
Carter Perez
Because the smart people who engineered all the tools we use today are dead and now it's just us.
Christian Jackson
womyn
Brayden Russell
This thread is unironically funny.
Connor Ward
So what ever happened with graphene?
Gavin Young
We humans need to increase our brainpower with eutelegenesis so we can reach the next level.
Tech in the ~70s was never about progress. It was about the Cold War (read: better missiles), which bled off to third parties and research firms using the technology to launch commercial and consumer enterprises.
Isaac Reed
physics a bitch
Nicholas King
It's actually not that bad of a thing. Finally companies will start optimization and and make programs run better on all hardware. Gone are the days when of you don't update your CPU every 2 years you when left with unusable crap.
Liam Wood
yea, you wish..
Cameron Perry
Real scientists work to produce results that can be used to create products that enhance our lives.
The leftists have pseudoscientists whose job is to promote the ideology, provide fairy tale scriptures (man-made climate change, eugenics, abortion, sexual deviancies, Darwinian Evolution, etc), and bring in the money and worshippers for their cult.
Old out-dated early 20th Century thinking: >Raise children to be normal healthy heterosexuals aspiring to monogamous relationships that generate lots of children within stable married families
Fresh new evovled & enlightened progressive 21st Century thinking: >Raise children to be pansexual polyamourous genderconfused ego-broken freaks whose life goals are nothing more than arranging who stuffs xers bumhole this week on grindr/tinder, or figuring out what porn they'll jerk off to while sitting alone in their empty studio apartments.
Accept Progress!™
Nolan Lewis
>Real scientists work to produce results that can be used to create products what the fuck mate science isn't just product and tech rnd
Blake Brown
If intelligence is genetic and directly linked to success in life then why do all women not use the Repository for Germinal Choice?
Justin Butler
This is a thread about Moore’s Law not about your fantasies of getting your bumhole stuffed.
Asher Howard
We have already gotten around this
They are developing Graphene to replace copper as the transistor and they are moving to a 3nm GAA design (gate all around) design to replace the Trigate designs we have today which got around moores law by replacing the planar transistors we had 10 years ago.
GAA will drastically reduce leakage allowing higher clocks at lower voltages. ARM designs are becoming more advanced and powerfull giving more performance at lower voltages and eventually, mass-produced graphene will replace copper as the transistor material resulting it even greater conductivity and higher clockspeeds.
A chip with graphene transistors will probably be running at a 600-700ghz clockspeed.
Thats quite a bit more than double performance.
William Myers
Intelligence is not linked to succes in life.
Anthony Torres
never need more than 640mb RAM chips anyway
Ethan Lee
Because we're almost at the next huge leap
Nathan Turner
Women are not intelligent.
Carter Perry
my gf is smart
Thomas Jenkins
Clearly not if shes dating you.
Aaron Evans
you don't mean that
Eli Collins
Anime isn't real.
Elijah Cook
Not really. It's true that everything is about pure money now. The real issue is that these companies hire PhDs to "park" them (i.e. prevent them from working at competitors). They don't actually care what these people do or don't do, but to get some return on investment, they typically prevent them from doing real research and instead employ them as engineers by constructing research-averse situations like 3 months mandatory turnovers. Trouble is academia doesn't pay shit without being in the right field at the right time and keeping at it for years with little to no (literally - you have to get grants to be paid more than a literal $0 per year) salary, so of course most scientists escape to industry. Part of this is a consequence of how economy has shifted from people being able to afford a house after a year of work to not being able to afford one with 20 years of full-time work, it has affected academia even more yet human lifespan hasn't increased proportionally.
Cooper Morales
>Why are we stagnating Jow Forums? ASML fucked up
Ian Rogers
Google isn't a tech company, it's a marketing company. Same with Facebook and Apple.
Real tech companies still invest heavily in research. The general public just doesn't care about them because they only care about consumer end products and clickbait apps.
Daniel Miller
The resolution of lithography depends on 3 factors: - wavelength of the light used. - numerical aperture of the optical system. - refractive index of the air/liquid above the wafer.
Traditional lithography hit a brick wall about a decade ago: - smaller wavelengths can't pass through lenses. - lenses can't be made bigger. - nothing exists with a much higher refractive index than the water they use.
EUV lithography was supposed to replace traditional lithography years ago. It allows much smaller wavelengths by replacing glass lenses by mirrors and placing everything inside a vacuum. But it's an engineering nightmare and still can't be used commercially.
Nolan Taylor
because the smart is a trait that can not be easily physically identified directly, which is why many women are sexually attracted to taller men because in the centuries past the poor could not get high quality nutrition compared to nowadays the richer people who could afford to eat were naturally tall as a result of better nutrition which caused tallness to become a sign of wealth, which correlated to being able to raise a large family.
Angel Gutierrez
>let women and niggers into stem >invest heavily into diversity >golly gee why isn't it paying off >Prof. Dr. Goldberg what should we do Yeah.. I wonder why.
Dominic Peterson
I bet that girl is cute.
David Moore
okay pol
Jose Fisher
The power of intel monopoly for the last decade.
Joseph Ortiz
> 2 legitimate /g posters >10000 salty /yungpol posters Can this board ever cure itself
moore's law isn't a law its just some guy going "oh wow the number of transistors is doubling every year". it doesn't apply anymore now that we are literally reaching the physical limit of how small transistors can be
Parker Ward
Do you show her your Rick and Morty tattoo?
Luke Campbell
Your opinion tweet is really the contribution we needed. Thanks for your service to the board.
EUV is already in widespread commercial use, sherlock.
Robert Gomez
I think it is. Albiet kind of redundant at this point. I have nothing to add except, Where is the economic value in faster computations, maybe for render farms for the movie industry, and for something like shadow? I think quantom computing will find implimentation in 'real' simulated 3d for gaming, for very dense universes, however it would take competent scientists to implement such a thing
Bentley Perez
>I think it is. We are in disagreement but that was no reason for me to be so harsh. I'm sorry.
Christian Brooks
"Moore's Law" has been nothing more than an observation and a corollary of exponential growth. Physics just caught up and economic realities make it difficult to miniaturize semiconductors/ICs any further. 10nm/7nm might be the last economically viable (albeit expensive as hell) nodes.
Michael Cooper
7nm is already too expensive for the vast majority of IC designers. 14nm was too expensive for a lot as well. The SOI market now is exploding because FinFETs are just out of reach for smaller firms.
in transitioning to be let in on the other side of the counter
Ryder Carter
>although you might get something out of it that i do not. If you're interested, it stems from old imageboard etiquette. There's no reason we cannot all share these boards and threads in harmony. user angry at fellow board members, is the worst thing to witness, in every perspective.
You posting a vibrant image only to yell at us, is not necessary. It's also an unfair generalization.
Nathaniel Barnes
You should seek help or write an operating system
Tyler Wilson
>read: better missiles That's progress, though.
Cooper Jenkins
I suppose. My image added unneeded harshness to the post, it was sort of unthoughtful. I suppose i can just accept everyones comments, as the chan dictates. Very true.
Aiden Mitchell
Moore's Law is an observation, not some kind of natural law. Intel in 2018 proved Moore's Law to be dead
Eli Baker
yes it is, not in all cases, but on average
Gabriel Thompson
Each new node should be cheaper per transistor than before. That's the entire reason they kept shrinking: to save costs.
Lower power consumption and high clock speeds are a nice bonus, but not the driving force.
If 7nm remains more expensive than 14nm then manufacturers will stick with 14nm and look for cost cuts elsewhere.
Carter Flores
It's more than just an observation. For a very long time it was a guideline for the industry to plan ahead, and therefore becoming a self fulfilling prophecy.
Designing a new CPU takes years. You want to know in advance which nodes will become available when your design is finalized. And at the same time the people working on new nodes, which also takes years, want to know what their customers will be looking for by the time they are ready.
Xavier Nguyen
No it isn't.
It was SUPPOSED to be starting with 10nm but they still use immersion lithography even for 7nm.
Jeremiah Moore
Dont mind me, i'm just the janitor.
Christian Flores
Samsung 7nm EUV is in volume production right now. TSMC has their 7nm EUV lines online as well. 5nm from each foundry will have dramatically heavier utilization of EUV.
Stop talking out of your ass.
Ryder Cook
The real problem is babyboomers are retiring and millennials are too retarded to work in tech. So gen X has to do all the work by itself, all tech companies are severely understaffed.
Adrian James
What's next after EUV? Gamma rays? Monolithic circuit blocks?
Justin Sullivan
The section of spectrum referred to as EUV provides wavelength tight enough to etch 1nm lines potentially. Where the industry needs to advance is increasing the intensity of the light sources, their power output.
Caleb Anderson
"volume production" isn't the same as mass production.
Samsung's 7nm isn't economically viable, so they don't mass produce. TSMC's 7nm mass production uses immersion.
Hunter Russell
Shorter wavelength = moar power generally right? So as the EUV gets tighter it increases automatically in power.
Austin Young
>I'm talking out of my ass trying to make excuses for my ass pulled nonsense TSMC's early 7nm is not EUV. Their EUV line has already accepted tape outs, passed risk production milestone. They'll be in full volume production before you can type up your next shitpost. Samsung didn't need a full SAQP non EUV 7nm line, they waited until they had EUV debugged, so they came right out the gate for it. Prices are in line with non EUV 7nm because it reduces cycle times. TSMC already has 90+ customers at 7nm, Samsung has a handful themselves, including IBM who is producing two different product lines on the process.
Tbats mass production. They'll be shipping thousands of wafers of 7nm EUV chips per quarter.
Jason Cox
>mass-produced graphene Fucking when, though? How can it be done, even theoretically?
Noah Brooks
>Tech is not about progress anymore, its about money. It's always been about money, what happened is that the money is not on progress anymore.
Connor Jones
retard
Jace Thomas
It's not a matter of "letting" them into stem. It's forcing them into stem that will cause problems.
Juan Watson
That one nigga checking his phone
Owen Roberts
nothing before 2024 if you ask me
Zachary Wright
What hes talking about likely isn't full graphene as an interconnect, its copper inside of a graphene sheath. It allows copper to carry signals that would ordinarily be beyond its capacity, at signaling frequencies that would blow out the interconnects. Constraining them with a boundry of graphene prevents them blowing and shorting out at more extreme potential rates.
I've got no idea how they actually produce and deposit the graphene, or how they'd grow it in place. That process would be highly guarded trade secret.
>I have no idea what I'm talking about at all so I'll just shitpost all over Jow Forums like a stupid nigger Good one.
James Jackson
The problem is the high energy light source. It destroys the mirrors and the plasma spreads contamination everywhere.
Jacob Cooper
Poojeets.
Justin Rivera
Just stop posting you low-IQ nigger, you clearly know nothing about this subject.
Levi Nguyen
There's hardly anything surprising in the fact that it gets harder and harder to refine something further the more refined it is. That applies to most things in life. Diminishing returns, and all.
Jack Reyes
The fundamental design of the semiconductor transistor CPU has not changed in a long time. We need something new. Or we need something to compensate for a lack of CPU advancement.
Bentley Richardson
Still in development.
Kevin Williams
>moving to a 3nm GAA design Any idea of the complexities in this? Just getting masks for this is a major undertaking.
but thanks for suggesting writing an Temple OS 2.0
Brody Russell
It actually "died" back in mid-2000s. That's when the semiconductor companies start to jump ship en mass and only the big player stay in the game. The big players are now starting to throw in the towel.
Zachary Ross
Nobody raises their children like that Get the fuck out pol gook nigger
Brayden Hernandez
Actually, it increase yields per wafer by making IC/transistor density which does save costs when you trying to mass produce units.
That's the main reason to keep shrinking IC/transistors. The problem now is that tools for cutting edge techniques requires massive capital expenditure. The big players are hesitant to jump to 7nm/10nm since it is unlikely they'll get a ROI in the near-future for it. The yields aren't that great either and there's no guarantee that they will improve significantly.
TL:DR version, the golden era of cheaper, faster digital computing is completely dead.