CPU's in the future...

[Hulk]

Based on past progress where do you think we'll be as far as desktop processor performance in 10 years, 50 years, and 100 years? How far will the performance be pushed?

Will desktop processors be capable of "supercomputer" performance in the near future?

10 years ago the fastest computer of the day was the Pentium II 450. In CPUmark99 that processor did about 34. A fast (say 3.2GHz) quad core C2D running four instances of CPUmark99 could score about 1800. That's about a 50x increase in performance in 10 years. Could we see another 50x increase in performance in the next 10 years? 1M Superpi times in a fraction of a second? HD video compressed at 50x realtime? Compress a two hour movie in two minutes?

Of course these are all really rough estimates but I'm curious as to what other people in this forum think will happen in the next 10, 50, or 100 years? I can't even begin to estimate how fast cpu's will be in 50 or 100 years.

 

[v8envy]

10 years: 50x sounds doable.
50 years: what are these 'desktop computers' you speak of?
100 years: what are these 'computers' you speak of?

 

[Hulk]

Originally posted by: v8envy
10 years: 50x sounds doable.
50 years: what are these 'desktop computers' you speak of?
100 years: what are these 'computers' you speak of?

Ha! Yeah I never thought about it that way.

 

[Idontcare]

The performance capabilities are there for another 50x increase in the next decade. The engineers have the roadmaps to generate the tools to ensure this can occur.

The challenge though for the marketing department is to convince the executive decision making team that spending the money on the engineer's roadmaps for generating the tools (EUV for example) needed to create the trillion-transistor CPU of the future will actually result in a product that the marketing and sales team can forcast a profitable and lucrative demand for.

Right now the marketing teams are not doing a great job from the engineer's viewpoint.

Why is that? No one can envision a killer app that doesn't initiate spontaneous laughter when mentioned in brainstorming sessions. The visions for next-gen killer apps have to be practical to be believable, but if it is practical then it most likely already exists.

The adoption rate phenomena of the mobile phone, the digital camera, and the web were not exactly forcast by Intel and Microsoft in 1989. And the companies who did try and make these products the blockbuster hits that they eventually became (Motorola 1973, Kodak 1975, and for WWW think of where all the originals have gone, AOL...) have all failed to dominate any of the market segments they played such a pivotal role in creating.

So will we have 50x improved performance devices in 10 years? I put money on it. What will we be doing with devices with so much performance? I have absolutely no idea. Who will be making those devices? Odds are actually not to shabby that it won't be the players in today's markets.

 

[Viditor]

I'm actually expecting the first Quantum computers using CNT sometime in the next 10-15 years.
I also think that speed will become much less improtant than storage will by then (I am personally almost out of room now, and I have 15 TB of storage...)

 

[Idontcare]

Originally posted by: Viditor
I'm actually expecting the first Quantum computers using CNT sometime in the next 10-15 years.
I also think that speed will become much less improtant than storage will by then (I am personally almost out of room now, and I have 15 TB of storage...)

:shocked: Holy shizer that's a lot of "visual entertainment material" you've accumulated! :shocked:

 

[Viditor]

Originally posted by: Idontcare

Originally posted by: Viditor
I'm actually expecting the first Quantum computers using CNT sometime in the next 10-15 years.
I also think that speed will become much less improtant than storage will by then (I am personally almost out of room now, and I have 15 TB of storage...)

:shocked: Holy shizer that's a lot of "visual entertainment material" you've accumulated! :shocked:

All of the footage I work with is at 2k resolution, and it's uncompressed 36bit.
All told, 15TB is only about 15 hours of footage...

 

[trOver]

Quantum<20 years away imho

 

[Viditor]

Originally posted by: trOver
Quantum<20 years away imho

you may be right...but remember that they made the first CNT transistor almost 8 years ago now.

 

[Modelworks]

Quantum or Organic computers.
Organic has tons of potential.
Imagine 'feeding' your pc

 

[P2Mc28]

I suddenly picture "feed me Symore"

 

[Hulk]

Don't get me wrong here as I'm always for faster but I'm wondering what I would do with something more than a 50x increase in performance.

Pretty much my only cpu bound application right now is video editing. It's pretty easy to load up my C2D with a 1080i HDV stream with color correction and other effects so playback rate is like 1 or 2 fps. Multiply that by 50 and you have realtime playback no sweat, with room to spare.

I guess there will be an equal increase in OS overhead or other "futuristic" features that might actually work with enough cpu power. Like voice recognition and some decent AI, aka Star Trek. "Computer search all references for a TSB on the front engine guard for a Mazda 3 and let me know if there are any instances of the owner being compensated for this part under warranty." Well not exactly like that but you get the idea.

 

[taltamir]

Originally posted by: Idontcare
The performance capabilities are there for another 50x increase in the next decade. The engineers have the roadmaps to generate the tools to ensure this can occur.

The challenge though for the marketing department is to convince the executive decision making team that spending the money on the engineer's roadmaps for generating the tools (EUV for example) needed to create the trillion-transistor CPU of the future will actually result in a product that the marketing and sales team can forcast a profitable and lucrative demand for.

Right now the marketing teams are not doing a great job from the engineer's viewpoint.

Why is that? No one can envision a killer app that doesn't initiate spontaneous laughter when mentioned in brainstorming sessions. The visions for next-gen killer apps have to be practical to be believable, but if it is practical then it most likely already exists.

The adoption rate phenomena of the mobile phone, the digital camera, and the web were not exactly forcast by Intel and Microsoft in 1989. And the companies who did try and make these products the blockbuster hits that they eventually became (Motorola 1973, Kodak 1975, and for WWW think of where all the originals have gone, AOL...) have all failed to dominate any of the market segments they played such a pivotal role in creating.

So will we have 50x improved performance devices in 10 years? I put money on it. What will we be doing with devices with so much performance? I have absolutely no idea. Who will be making those devices? Odds are actually not to shabby that it won't be the players in today's markets.

That is because suits sit down and try to come out with a killer up... when really, what drives technology is the "toy" aspect...

The internet was driven by porn. Computer development was mostly pushed by video games. There are very distinct stages in the development... text, music, video, HD video... each one is progressively and exponentially more difficult.

Now when I say pushed forwards I mean the market size. Sure there are serious applications that REQUIRED the extra processing power... but millions of people will pay thosands of dollars for a new gaming/toy machine every year, they would not for something capable of revolutionizing physics.

I think the NEED for extra power is definitely there... Because the future is direct brain-computer interaction. And you want cyber Pamela Anderson to look as realistic as possible...

You want things that require massive computing power?
Requires NO new technology:
1. REAL physics in game. (the next 10 years will not longer be about pixel size, but PARTICLE size... how many particle "chunks" is the world divided into, exponentially more complex).
2. Capture and encode ALL the TV chanels AT ONCE.
3. Create a mobile PC the size of a wallet that can be carried in a pocket, and has at LEAST as much power as top of the line computers today (this is something I can use and WANT today... i wish it was possible with todays tech at a practical cost).
4. Cryptography. Both breaking and creating encryptions. DRM, etc, all require massive computational power.

Requires some revolutionary new tech:
1. Render directly to brain
2. AI
3. Cybernetics

 

[aigomorla]

shoot im just praying for a faster PDA phone within the next few years.

Today's PDA phones are a joke. There so dayam slow.

And you guys forgetting moor's law already?

 

[bryanW1995]

Originally posted by: Viditor
I'm actually expecting the first Quantum computers using CNT sometime in the next 10-15 years.
I also think that speed will become much less improtant than storage will by then (I am personally almost out of room now, and I have 15 TB of storage...)

that's a lot of pron...wow...uh, I can't focus right now b/c I'm thinking about that much pron in one place...uhhhhhhh...what were we talking about?

oh yeah, computers in 10,50,100 years.

10 years = another 50x performance
50 years = mind/machine interface. miniaturization to the point that anything you can see will be enormously more powerful than what we have today. No more desktops, just a personal VR lounge.
100 year= I AM GOD. well, anybody alive with lots of money then will be pretty well off at least.

 

[taltamir]

Originally posted by: aigomorla
shoot im just praying for a faster PDA phone within the next few years.

Today's PDA phones are a joke. There so dayam slow.

And you guys forgetting moor's law already?

double power ever 18 monthes (year and a half).

So that is 6 doubling plus almost a seventh... so 6-7 doubling.

Performance 10 years = performance(today) x 2^6 - 2^7 = x64 to x128 of todays power in 10 years...
The thing is... its not really moores law, its "moors rule of thumb". It is just a general estimate.

 

[Drsignguy]

I'm going to guess that speed will not be the issue as opposed to the efficiency of what will come. Remember the days of old that ( based on Intel ) the Pentium first came to be? 1000mhz, and on and on and on....to what we have today. I read somewhere in a column ( and I cant find it ) that 10Ghz wasn't out of the question...I guess as they have gone towards the more efficient approach. Correct me if I am wrong but processors today are more efficient by using far less power than ever before? Not to say that speed is out of the question but I don't see 10ghz before an 8 core processor. Which, 10 years ago, did we see a four core processor coming? Although, I get the notion that by reading most of the posts in "AT" that there are some whom have the knowledge and or the information of seeing the future . And BTW, you know who you are. Also I see "no moving parts" i.e.; HD, Optical drives. Remember, I am just guessing to have some fun with this. How far off am I, Who can foresee the future? Sure as hell it isn't me......

 

[taltamir]

you are confusing cycles with speed.

A 2ghz wolfdale is MUCH faster then a 2ghz P3. With things like SSE it can be orders of maginitude faster.
Wolfdale for example increases divx encode speed by 60% over same clocked Conroe, and by another 70% with SSE4 enabled, resulting in 2.3 times faster for the SAME Ghz.

Power efficiency and efficiency are not the same. Neither are Mhz and Speed.

 

[Xed]

Star trekkin....across the universe

 

[lopri]

What do you guys think of Intel's strategy? Obviously they realized their limit long time ago and have been pushing multi to many cores for some time. Now it's clear that they want 'x86 everywhere' for the company's future. (Atom, Larrabee, etc.) That agenda seems to be 'approved' so far because of Intel's superior manufacturing.

 

[bryanW1995]

intel has to adapt to survive and prosper just like any other company. they're betting a lot on x86 getting picked up from top to bottom, ie, from supercomputes to handhelds. if it sticks then intel could be the biggest and most profitable company in the world in 20 years. If it flops then intel could be a memory in 20 years.

 

[firewolfsm]

I'm pretty sure in a hundred years, chairs will all figure out your height and weight with some sensors and morph to fit everyone's body type.

 

[Greg04]

Originally posted by: Hulk
Based on past progress where do you think we'll be as far as desktop processor performance in 10 years, 50 years, and 100 years? How far will the performance be pushed?

Will desktop processors be capable of "supercomputer" performance in the near future?

10 years ago the fastest computer of the day was the Pentium II 450. In CPUmark99 that processor did about 34. A fast (say 3.2GHz) quad core C2D running four instances of CPUmark99 could score about 1800. That's about a 50x increase in performance in 10 years. Could we see another 50x increase in performance in the next 10 years? 1M Superpi times in a fraction of a second? HD video compressed at 50x realtime? Compress a two hour movie in two minutes?

Of course these are all really rough estimates but I'm curious as to what other people in this forum think will happen in the next 10, 50, or 100 years? I can't even begin to estimate how fast cpu's will be in 50 or 100 years.

It doesn't matter how fast CPUs get, Microsoft will be there to trip up even the fastest one. I can see it now, WINDOWS AEON (WINDOWS 22) Minimum System Requirements: 100 Core AI Gel Cubes, 40PB Holographic Storage, and at least a 500TB RAM Array (with 1KW of power dedicated to lighting the Bill Gates laser memorial on the moon). Of course, consumers will continue to revolt against this bloatware, and XP will have reached Service Pack 279 (where *almost* all the bugs have been worked out).

 

[Idontcare]

Originally posted by: lopri
What do you guys think of Intel's strategy? Obviously they realized their limit long time ago and have been pushing multi to many cores for some time. Now it's clear that they want 'x86 everywhere' for the company's future. (Atom, Larrabee, etc.) That agenda seems to be 'approved' so far because of Intel's superior manufacturing.

Like the borg, resistance is futile.

I am curious though what protection Intel has over their x86 IP. Is it merely patent protection (which expires in <20 yrs of patent award date) or some other mechanism which affords protection for an even longer time period?

At some point in the very near future this whole "x86 licensing embargo" should become a moot issue as it becomes public domain.

 

[bryanW1995]

didn't ibm try to get by without patenting their technology with the first pc? right about the time that compaq/etc came along and reverse-engineered the cpu iirc... but the point is that they didn't patent, maybe they wanted a copyright or some other protection?