[Boards: 3 / a / aco / adv / an / asp / b / bant / biz / c / can / cgl / ck / cm / co / cock / d / diy / e / fa / fap / fit / fitlit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mlpol / mo / mtv / mu / n / news / o / out / outsoc / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / spa / t / tg / toy / trash / trv / tv / u / v / vg / vint / vip / vp / vr / w / wg / wsg / wsr / x / y ] [Search | Click for more| Home]

Moore's Law is dead. Chips haven't gotten faster for

This is a blue board which means that it's for everybody (Safe For Work content only). If you see any adult content, please report it.

Thread replies: 126
Thread images: 10

File: tinychip.jpg (18KB, 400x300px) Image search: [iqdb] [SauceNao] [Google]
tinychip.jpg
18KB, 400x300px
Moore's Law is dead. Chips haven't gotten faster for shit lately. It will stay dead. Prove me wrong.
>>
>>45416530
i still have an almost six years old intel c2d and a 8800gt. and before that i bought a new pc every 2 years. what's going on?!
>>
>>45416580
>what's going on?!
The physical limits of silicon transistors.
>>
>>45416617
AAA IM PANICKING
>>
Silicon can only get so hot and current transistor gate manufacturing processes limit us greatly. We need new material that is more heat resistant and new way to compute data and rewrite all the code and shit. Bro . We are all gonna die
>>
stop buying the shitty new CPUs and eventually jewtel will be forced to make something new, maybe not silicon based
>>
rip
>>
>>45416530
I can't.
http://www.extremetech.com/computing/165331-intels-former-chief-architect-moores-law-will-be-dead-within-a-decade

http://news.techworld.com/operating-systems/3576581/moores-law-is-dead-says-gordon-moore/
>>
>>45416654
>when will quantum computing happen
>>
Maybe you should stop listening to corporate shills, you know, its pretty stupid that they created a 'scientific law of why you should invest in intel' and people just went with it
>>
>>45416530
Moore's law was almost dead.

But ASML just figured out how to do EUV lithography.
It's going to take off again starting 2016, and will continue for at least another decade.
>>
Moore's "law" doesn't account for market forces.

Never got why it was called a "law" in the first place.
>>
>>45416654
>>45416696
Based Graphene will save the day.
>>
>>45416738
Yeah when we can get working transistors out if it maybe. I do hope graphene is the solution, but we need to find out how to make that happen first.
>>
>>45416729
>>45416702
>>
>>45416707
intel will go into mass production of euv in 2017
>>
>>45416782
http://www.extremetech.com/extreme/175727-ibm-builds-graphene-chip-thats-10000-times-faster-using-standard-cmos-processes
>>
Isn't IBM and HP working on a new type of computer that imitates the brain basically? Just stacking a whole bunch of processing chips into a cube or whatever the fuck and apparently it's really power effecient
>>
>>45416782
http://www.extremetech.com/extreme/164301-graphene-transistors-based-on-negative-resistance-could-spell-the-end-of-silicon-and-semiconductors
>>
>>45416530
I agree. Parallelism is the near term future until better chip designs or a totally new way of computing comes along.
>>
>>45416738
>>45416679


>But the old way — the old promise — of a perpetually improving technology stretching into infinity? That’s gone. And no one seriously thinks graphene, III-V semiconductors, or carbon nanotubes are going to bring it back, even if those technologies eventually become common.
>>
>>45416895
>>45416918
isn't graphene expensive as shit? will not take off this way
>>
>>45416945
not really. some dude found a way to make it using a fucking cd drive or some shit

also it's just carbon, which is found everywhere
>>
>>45416976
This, I welcome our new graphene overloads. Silicon market can now dedicate itself to providing material for big fat tits. New golden age.
>>
File: C.gif (2KB, 144x182px) Image search: [iqdb] [SauceNao] [Google]
C.gif
2KB, 144x182px
how can one element be so based?
>>
>>45417047
It also is making up muh dick which is now diamond hard
>>
My PS4 is the world's most powerful console so I think you are lying.
>>
>>45417047
Just like in programming, C++ (Carbon 14) is more useful in science than C (Carbon 12) because it can be used to date things, something that freetards never do.

Stay mad /g/.
>>
>>45416976
Making some isn't hard.
But making lots and purifying it is.

With most methods you get only a tiny fraction of a percent graphene amongst shitloads of other stuff.
>>
>>45417005
>Silicon
>Silicone
>>
>>45417096
>My PS4 is the world's most powerful console

tallest midget
>>
>>45417115
Isn't there silicon in silicone?
>>
>>45417115
>bickering over UK/US spelling differences
>>
>>45417172
They are two different things.
Silicone is what you seal your windows with.
Silicon is the element.
>>
>>45417197
Yes, the element that makes up the stuff you seal your windows with.
>>
File: emperor.gif (493KB, 500x263px) Image search: [iqdb] [SauceNao] [Google]
emperor.gif
493KB, 500x263px
>>45416530
>He thinks moore's law is about processors getting faster.
It's just about the count of transistors which does not equal the "speed" of the processor.

p.s. I bet your one of those kids who say they have a 10.8 Ghz CPU just because it says 4x2.7Ghz.
>>
>>45417212
>Silicones are polymers that include silicon together with carbon, hydrogen, oxygen, and sometimes other elements.
>>
>>45417240
So silicon is indeed a component of the material that makes fake tits. So anon was right.
>>
>>45417253
So what was the point the the anon that was right pointing that out?

Saying silicone was still wrong.
>>
>>45416617
no. it's the marketing demand.
evolution is flawed because the need for CPU resources is stagnating. Tablets and smartphones are selling, and requires small chips but not real chips. it's technology that is the equivalent of a ten years old laptop/office desktop, and even...

there's no real need for insane cpus, since most of the thing done with PCs relies on the internet connection. the only recent thing Intel brought was just rebranding core2quad into i5, i7... AMD doesn't care much, they're just continuing on selling.

Most of the high performance computing's demands are basically servers. AMD makes cheap opterons, what Cray fills entire boxes to make supercomputers. Of course we still have Fujitsu, NEC, SGI and IBM and a few brands that makes monstruous processors for vector machines. (altough even SGi now are filled with Xeons.) add a few custom corprocessors like a rack of nVidia teslas and you got something perfect for scientific simulation.

There's nothing requiring real processing power in matter of software, even if we wouldn't mind a few more gigaflop/s on our machines to improve a few things. GPUs does a lot of work for all 3d tasks.

the cancer lies in the cheap, and so-called "green computers", macs, all in one desktops, touch screens runing windows8. the goal of course is to destroy personal computing. People being equipped with tablet or other terminals is the goal of this market-controlled and technological degeneracy.

since it's better for those who aims to control the internet to not have you owning your own personal storage device. many faggots claims the cloud is the fucking future, these are a part of the problem. using ipads/chromebooks/macbooks, things you buy and throw away after the battery no longer holds a charge (2 years max)

the goal is to destroy anonimity, privacy, possibility for all to create, hack, keep all kind of downloaded movies and media at home, including all kind of illegal books and docs.
>>
>>45417288
>So what was the point the the anon that was right pointing that out?

Read this sentence to yourself and try to understand what the fuck it says, because I certainly can't.
>>
>>45417312
I understood fine.

What was the point in the anon who you are referring to being correct? It could be assumed that that anon was being sarcastic.

You could also interpret what that person said as being incorrect, since they might have been trying to say that silicon were the same thing.

Silicone != Silicon
>>
Transistors are still getting smaller, its just architecture design is being focused on energy efficiency rather than speed.
>>
>>45417372
Jesus fuck, trying to follow this train of thought is harder than undoing my bro's box of cables.
So, let me get this straight: The one anon said that silicon providers can now focus on providing silicon to make implants, correct? Silicone contains silicon. Silicon don't come from thin air, it has to be provided by someone.
>>
>>45417301
That actually kinda makes sense.
>>
>>45417404
Ok I lied I am confused.
>>
Whatever happened to photonics? I heard back in 2006 we'd be getting photonic hardware within a few years.
>>
>>45417404
SILICONE IS NOT AN ALTERNATIVE SPELLING TO SILICON.

Also everyone being named anon is confusing.
>>
>>45417429
>Also everyone being named anon is confusing.
You'll get over it.
>>
>>45417429
I never said silicon is silicone. Silicon is an element, silicone is a polymer with silicon in it.
>>
>>45417422
Just another sensationalist article to catch eyeballs based on somebodies maybe or out of context quote.
>>
>>45417422
Vaporware, just like this graphene business will be.
>>
>>45417454
>>45417475
It seems like they're actually making headway with tangible gains, though. To me it looks like it shouldn't be too far off.

http://www.intel.com/content/www/us/en/research/intel-labs-silicon-photonics-research.html
>>
>>45417429
I know rite? I dunno why the IDs that /b/tards get are omitted everywhere else. Sure, less shitposting elsewhere, so I suppose they wanna roll with the full anonymity thing, but it does get confusing if you don't know who the heck you're even supposed to be arguing with.
>>
>>45416696
PCs will not benefit from quantum computing
>>
>>45417407
Samefag
>>
>implying a switch from silicon transistors to CNT won't continue moore's law
>>
>>45417680
Anything that benefits from encryption will benefit from quantum computing, am I wrong?
>>
File: Y3_9b.png (245KB, 498x544px) Image search: [iqdb] [SauceNao] [Google]
Y3_9b.png
245KB, 498x544px
>>45417453
It is just anglotards who decided to not call it silisium
>>
>>45417761
Oh fuck off. If it were called that we'd have some close name for the polymer too.
Don't act like it would fix anything.
>>
>>45417786
actually the polymer is called essentially "silicone" in the very same languages.
>>
>>45417301
wow, this seems about right
>>
>>45417567
maybe the point is that internet arguments are stupid
>>
>>45417761
szilícium masterrace
>>
File: roger-moore[1].jpg (139KB, 822x1023px) Image search: [iqdb] [SauceNao] [Google]
roger-moore[1].jpg
139KB, 822x1023px
>Moore's Law is dead
>>
>>45417796
What I'm saying is, if we did go with that name, the polymer would have a different name that was similar. If history was changed, it would fix nothing.
>>
>>45417840
Well yeah, but that's what this place is for
>>
>>45417702
nope, this second one was me.
this is also me:
>>45417567
And this FURTHER proves my point about IDs.
>>
>>45417859
but -um is for elements pretty much maybe it would have been called siliciate but that is way less retarded than silicon[e]
>>
>>45416530
>Moore's Law is dead
You must be retarded.
>>
>>45416738
>>45416782

Graphene can do anything except leave the lab.
>>
>>45417301
Since there's no need for insane CPUs, shouldn't development be focusing on cheapening them?
>>
>>45417931
naw man, we need moar cores
>>
Moore's law is about transistor count~
>>
>>45417931
The focus is on making them more power efficient.
Which does have as a side result them become more cheaper as processes mature.
But the overall goal is battery life, not cheapness of silicon.
>>
>>45417931
Corporations produce goods for profit, not for the good of humanity.
>>
What's better, diamond or graphene? I know they're basically the same elemental composition but diamond makes an even better semiconductor and is far more practical.
>>
>>45418016
And the cheaper they make them, the more profit they can make.
>>
>>45418064
Why do you think it costs like $4 to make Beats By Dre™ and they sell it for $250? The cost of production will become cheaper, but it won't reflect on the price consumers pay.
>>
>>45416530

>Moore's Law is dead. Chips haven't gotten faster

When did Moore ever mention "speed", dipshit?

You've only proved your poor reading comprehension.
>>
>>45417301
Lies.

It is physically impossible to make a functional transistor out of silicon that is smaller than 5 micrometres in size because of leakage and the uncertainty principle.

The smaller it gets the harder it is. Cutting edge chips are currently about 14 micrometres. We are very close to the end, and each micrometer is more expensive than the last.
>>
>>45418081
That's because people are dipshits and they just want them so they can think they're cool like everyone else when really they just spent $250 to look like a complete retarded sellout with no concept of the value of money. It's a stupid trend just like the rubber band animals, it will pass when people realize how stupid they are.
>>
>>45416679
What fuck is the intel guy taking about. Moore's law is already dead even wikipedia says that it's slowing down by the end of 2013.
>>
File: photo.php.jpg (2KB, 54x71px) Image search: [iqdb] [SauceNao] [Google]
photo.php.jpg
2KB, 54x71px
>>45416936
>>
I need moore's law because my new laptop with i7 extreme mobile processor STILL can't smoothly decode, interpolate, and upscale 10-bit 1080p BDrips at the highest quality using madVR/SVP
>>
>Thinking that moore's law wasn't just an observation
>thinking that corporations make better profits by slowing down the evolution of hardware.
>>
>>45418081
But the products themselves are cheaper. Don't compare processors with beats, it makes you look like a retard to compare products that can be wholely quantified against those that are 90% qualitative.
Apple's entire success is based on making their products depend on qualitative properties rather than quantitative ones.
Before apple, computers were sold with gigs and bytes. After, they're sold with 'style, thiness, it just werks, etc'. See what I mean?

Anyway processors can be objectively quantified in terms of performance so competition can actually send the prices down.
Because how are you supposed to make money if no-one buys your products because they're more expensive?

Well you give retail discounts to people who work in retail, so they are indoctrinated into selling intel.
And by marketing intel as a brand. And by paying OEMs to not use AMD chips etc.

Business practices aside, if Intel had an actual competitor (they are actually competing with ARM in the mobile space at the moment, so they have real competition instead of the sincere but lackluster AMD)
Mobile space processing is competitive.

And this is why low power processors are the future. don't expect great performance leaps until a change in substrate.
>>
>>45418111
There will eventually be an end of course. You can't pack an entire computer into the size of a pearl. It's just not practical. However, each micrometer they shave off will increase speed more than the last. So it's slightly expodential. I can't see silicone based transistors taking processors past 10ghz tho.
>>
>>45417301
>using ipads/chromebooks/macbooks, things you buy and throw away after the battery no longer holds a charge
Or, here's a crazy thought, you repair them.
>>
>>45418174
Look at the GPU space. Hardware accelerated decoding. Problem is you're doing 10bit. If you were doing 8bit, the hardware acceleration on the iGPU would kick in and make it smooth. The actual processing for that sort of thing is cheap so long as the processor is dedicated to that task. CPUs are general processors.
>>
>>45418192
This. Intel's R&D costs have skyrocketed over the past decade. It's simply getting harder and harder to get better performance and smaller dies.
Hence why AMD is 3 years behind.
>>
File: carbon.jpg (214KB, 1516x704px) Image search: [iqdb] [SauceNao] [Google]
carbon.jpg
214KB, 1516x704px
>>45418062
>basically the same elemental composition
No they're not. Graphene is more like Graphite than like diamond.
>>
>>45418174
Your autism will be fixed by a cpu
>>
>>45418214
>Apple warranty runs out
>Average apple user doesn't even know how to unjam a pencil sharpener
>Throw away buy new
Repeat
>>
File: quote.jpg (324KB, 1200x1200px) Image search: [iqdb] [SauceNao] [Google]
quote.jpg
324KB, 1200x1200px
>>45417301
QFT.
>>
>>45418239
I mean they're both carbon. Diamond is a crystalline structure and graphene in planar.
>>
>>45418216
I would rather pay $500 more for a better processor than wait for transcodes, too bad they don't exist.

>>45418243
yes, yes it will.
>>
>>45418281
thank you for this laugh
>>
>>45416530
It's not dead. You're retarded. Ram is getting faster, graphics better, cpu's smaller and faster. You're just getting old.
>>
>>45418281
thanks for my new wallpaper
>>
every single year someone says
>ITS OVER MOORE IS FINISHED
and every single year they are proven wrong
>>
File: Bravo.gif (1011KB, 225x116px) Image search: [iqdb] [SauceNao] [Google]
Bravo.gif
1011KB, 225x116px
>>45418281
>>
Does anyone else remember the run from the 386 to the Pentium 4 bring like HOLY SHIIIIIIIIT

Every 12 months your $3,000 desktop was an obsolete piece of shit and the new hotness was like 10x as fast.

I kind of miss it now even if prices are better.
>>
>>45418214
they are designed not to be repaired.

You can fix a good old PC, it will never break, it will still access the internet as long wifi or TCPIP technology exists. but you cannot fix a sandwich of LCD display, chips on a plate, battery. Soon these will be sealed in plastic and not even have screws, what's already the case for a lot of apple products, cheap chinese screens and printers. Things ARE designed to break.

Things are designed to move people to the cloud and destroy personal computing and all the things people like us used to be able to do. The coming up generation will probably not even be able to use a proper keyboard, or bypass a simple DRM. This is idiocratic as fuck!
>>
http://arxiv.org/abs/1411.4798

Paradigm shift?
>>
>>45417301
See >>45418192
>>
>>45417237
/thread
>>
>>45418248
That's their problem. You act like a more 'repair friendly' PC would be of any use to them.
>>45418382
>they are designed not to be repaired.
No, they're just not designed to be EASILY repaired by end users. I've had no problem replacing batteries in unibody MacBooks, but that's because I'm not an idiot. Do I wish it was as simple as in the older Apple laptops? Absolutely. But Apple went with the internal battery for one reason or another, probably due to 'MUH THIN'.
>>
>>45418316
Stagnating under an average of 4GHz since 2004.
Silicon still have a bright future before quantum tunneling and other electronic weirdnesses tend to have any effect, but we're far from the absolute limit, for this material at least. and there's plenty of room for larger chips, just on the surface of a few inches's LGA. . .

The limitation of technology is a made up lie, it's just a mass consumption system. If we had computers built to last, like the thinkpads or HP elitebooks, or even a Dell Vostro, we would keep our computers for like, 5, 7, 10 years and they would not be much obsolete as long we keep it for watching our movies, writing, working.
>>
>>45417301
>There's nothing requiring real processing power in matter of software
Except for games wich are the fuel of evolution of hardware
>>
>>45418436
Exactly. It's easy to replace shit on Mac books. iPhones aren't that hard either.
But if you were smart you'd buy your own parts and build your own pc though.
>>
>>45417387
That only happens because it's easier and more possible.
>>
We have the technology though.

>We can make the same old shit barely any faster and people will still buy it at almost the same rates!

Meanwhile my 2600k is still more than adequate for just about everything still. :^)
>>
>>45418488
>Exactly. It's easy to replace shit on Mac books. iPhones aren't that hard either.
Okay, so what's your point?
>But if you were smart you'd buy your own parts and build your own pc though.
Yea, I'll get right on building my own laptops.
>>
>>45418472
Games makes you buy like; one GPU per year now, despite technically, if the developers wanted to make an effort, the most modern games would need 2Gb of RAM, a dualcore 2GHz and a 9600GT. I think the PC gamers and professionals will prefer to have their software on a decent battlestation rather than a cloud-based, disposable device.

>>45418436
Of course someone will still be able to fix it, to hack newer stuff but everything will be made to discourage such activity. In the 1980s, when you bought a hifi, they offered you the diagrams for maintenance, in case something failed. My 80's Marantz still works like a charm. I can't tell the same about my phone. We really have to fix this system.
>>
>>45418594
>In the 1980s, when you bought a hifi, they offered you the diagrams for maintenance, in case something failed.
What's your point? We have SMDs, now. Can't repair those at home. Audio equipment isn't computers, either.
>>
>>45416580

Cause there's no practical use for more CPU processing power in general usage.
I suspect even older ddr1, or even sdram cpus will work fine, if they were provided with sufficient memory pool and bandwidth.
With gpus it's a bit different story since incompetence and terribly inefficient 3d engines are common and need brute forcing if you intend to play the games at maximum (which is silly and very expensive imo).
>>
This is good news, my haswell i3 should be good for a long time to come. All I need to do is buy an nvidia card and 4 more GB of ram then I'll be set for 10 years.

The drawback is I doubt that computers will be as intelligent as humans by 2020 like was expected.
>>
File: 1417211120317.jpg (165KB, 1920x1080px) Image search: [iqdb] [SauceNao] [Google]
1417211120317.jpg
165KB, 1920x1080px
>>45418638
bro, computers from the eighties still works fine today, the manufacturing is the same since chips are sealed. Everything manufactured decades ago will certainly still work today, while shit you buy today is designed to break.

if you bought a phone recently that's made in China, don't expect it to work in a couple of years. While a Nokia, Motorola or Blackberry will still be operational.
>>
anyone got actual stats of the last 4 years in cpu's and their power. Its definitely has not doubled^2. My ye olde i5 2005k barely has been trumped by the most extreme end intel ones.

for its price range it what, got a 50% increase in speed over these 4 years? Barely 25% per 2 years?
>>
>>45418792
Look, I even heard retards who say that the CPU is loosing MHz with time. It's like audiophile crap. my core2quad + GTX480 still plays modern games. Win7 ultimate and everything, I can hardly find an equvalent today, I'm not interested to upgrade.
I still prefer my 1920x1200 ASUS display versus an IPS 1920x1080. And I'm completely right about it.
>>
>>45417099
>C++ is an unstable isotope
Not a bad comparison indeed.
>>
>>45418897
what? I am talking about the literal interpretation of moore's law. The double of transistors every two years.

All I am asking for is for accurate stats and all I am giving is some rough estimation of what it could be.
>>
>>45418786
>bro, computers from the eighties still works fine today
Do what they were designed to do? Sure. Compete with modern products? No.
>while shit you buy today is designed to break.
Go on believing that. I've got plenty of modern technology that has worked for many years and is showing no signs of stopping.
>>
>>45419124
>Release cycle on the most popular products shortens to a borderline braindead degree
>New phones every year, shit always improving
>Development of applications and all that jazz focus on the new products
>Older devices become obsolete fast and completely worthless even faster.

Simple as that, "designed to break" is just shit retards spout because new stuff won't work on old stuff. That and shit is much more delicate and people are clumsy cunts who can't take care of their shit.
>>
>>45419219
>new stuff won't work on old stuff
Old stuff stops working when you update them and besides that cellphone gets old way faster than pc.
Thread posts: 126
Thread images: 10


[Boards: 3 / a / aco / adv / an / asp / b / bant / biz / c / can / cgl / ck / cm / co / cock / d / diy / e / fa / fap / fit / fitlit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mlpol / mo / mtv / mu / n / news / o / out / outsoc / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / spa / t / tg / toy / trash / trv / tv / u / v / vg / vint / vip / vp / vr / w / wg / wsg / wsr / x / y] [Search | Top | Home]
Please support this website by donating Bitcoins to 16mKtbZiwW52BLkibtCr8jUg2KVUMTxVQ5
If a post contains copyrighted or illegal content, please click on that post's [Report] button and fill out a post removal request
All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
This is a 4chan archive - all of the content originated from that site. This means that 4Archive shows an archive of their content. If you need information for a Poster - contact them.