General discussion

Locked

The World's Fastest Computer but..is the human brain more intelligent????

By Black Panther ·
can it beat Gary Kasparov in Chess?? ( possibly the greatest chess player of all time! ) or even conquer Human's in other classic games like "Go"??

Gary Kasparov proved how intelligent the human mind is by conquering Deep Blue and matching Deep Thought and X3D Fritz.
In 1965, the Russian mathematician Alexander Kronrod said, "Chess is the Drosophila of artificial intelligence."

These previous SuperComputer's relied on massive calculation "grunt" ( over 1 million per second ) whilst the human mind struggled to calculate 10 moves ahead but... the human mind had an advantage over the supercomputer by using the Human Brains pattern recognition capabilities.

Will SuperComputer's ever be capable of matching the human brain?????

Is the Human Brain more intelligent by 'design'???

How powerfull is the Human Brain compared to these Computers???

What do you think????


IBM's Blue Gene/L supercomputer has doubled its own performance record by doubling in size; the machine has now performed 135.5 trillion calculations per second

The system, which is in the process of being installed at Lawrence Livermore National Laboratory, already topped a list of the 500 fastest supercomputers last November with a sustained performance of 70.7 trillion calculations.

The latest performance increase was achieved by doubling the number of racks in the system to 32. If the design continues on this path, the final machine with 64 racks will perform at about 270 teraflops later this year.

Each rack contains 1,024 processors. So 32 x 1,024 Processors !Each processor, a special variant of IBM's Power family, has dual-processing engines called cores. For running the basic performance test used to rank the top 500 computers (a test known as Linpack), each core can perform calculation work, but in many tasks one of the cores will be devoted to communications.

Blue Gene began in 2000 as a research project to build a system that could perform 1 quadrillion calculations per second--a petaflop--but IBM is trying to make a business out of the machine. It's begun selling the Blue Gene/L machines for about $2 million per rack and is renting out access to one of its own machines.

Blue Gene/L is one of several products stemming from IBM's focus on high-performance technical computing. The company is trying to secure the top spot in the market from Hewlett-Packard and keep its machines ahead of high-end rivals including Silicon Graphics and NEC.

IBM has said the full system will be installed by May, and Livermore Lab spokesman Don Johnston said it should be up and running in July.

Livermore's Blue Gene/L initially was expected to be used for nonclassified work, but its mission expanded to include weapons research as the lab realized it could be useful there too, Johnston said.

The computer has been used to simulate the interactions of 16 million atoms in a sample of tantalum that's solidifying under pressure, but Blue Gene/L isn't suited for all supercomputing tasks.

DOE purchased Blue Gene/L as part of a $290 million deal that also included a system now called ASC Purple. Purple uses fewer, more-powerful processors with more memory, a design that makes it better suited to its primary purpose: complex simulations of nuclear weapons physics.

ASC Purple is based on p5-575 servers. IBM will start delivering ASC Purple to the Livermore lab in April, and it should be complete in July or August, Johnston said.

Specs

Blue Gene At a Glance

Attribute Description Benefit
Processor Power PC 440 700MHz,two per node Lowpower allows dense packaging; better processor-memory balance
Memory 512 MB SDRAM-DDR per node
Networks 3D Torus ? 175 MB/sec in each direction
Collective Network ? 350 MB/sec; 1.5 usec latency
Global Barrier/Interrupt
Gigabit Ethernet (machine control and outside connectivity) Special networks speed up internode communications; designed for MPI programming constructs; improve systems management
Computer nodes Dual processor; 1024 per rack Double FPU improves performance
I/O nodes Dual processor; 16 per rack (additioanl nodes optional) Strengthens systems management
Operating Systems Compute Node ? Lightweight proprietary kernal
I/O Node ? Embedded Linux
Service Node ? SuSE SLES 8 Linux
Front End Nodes ? SuSE SLES 9 Linux Kernel tailored to processor design;
industry-standard distribution
preserves familiarity to end user
Performance Peak per rack (virtual node mode) ? 5.73 teraflops
Peak per rack (coprocessor mode) ? 2.86 teraflops
Linpack per rack (VN mode) ? 4.53 teraflops Highest available performance
benefits capability customers
Power 28.14 KW power consumption per rack (maximum)
208 VAC 3-phase; 100 amp service per rack Lowpower draw enables packaging
Cooling Air conditioning 8 tons/rack (minimum)
2800 CFM (compute rack); 350 CFM (power supplies) Lowcooling requirements enable extreme scale-up
Acoustics 9.0 LwAD and 8.7 LwAm
Dimensions
(include air duct) Height ? 1958mm
Width ? 915mm
Depth ? 915mm
Weight ? 759Kg Design allows "brickwall" layout for better floor space utilization


~~~~~~

Japan is aiming to develop a supercomputer it hopes will be fast enough to help it regain the top spot it lost to U.S. makers last year in an industry that is often seen as a proxy fight for technological supremacy.

The government wants to develop a supercomputer that can handle over a quadrillion calculations per second as early as the fiscal year ending in March 2011, an official at the Ministry of Education, Culture, Sports, Science and Technology said on Monday.

That would compare with the 135.5 trillion calculations per second in independent tests earlier this year for IBM's Blue Gene/L, currently the world's fastest computer, built for the U.S. Department of Energy's National Nuclear Security Agency.

Japan's fastest machine currently is NEC's Earth Simulator, which boasts nearly 36 trillion calculations per second.

It had been the world's fastest supercomputer until last year, when it was overtaken by a machine made by SGI and by Blue Gene/L.

The U.S. and Japan have battled for a number of years over supremacy in supercomputers, machines that have massive processing power and which are used in advanced climate forecasting, medical research and other areas.

Japan's new supercomputer would enable medical researchers to conduct comprehensive simulations on how a medicine is dissolved and carried through a human body and how it affects a specific organ, for example, the ministry official said.

It would also help provide weather forecasts with improved accuracy, the official said.

NEC, Hitachi, the University of Tokyo and Kyushu University were chosen by the ministry earlier this month to develop critical technologies to make the ultra-fast computer possible.

Details such as how much total investment will be needed for the project and which organizations will be involved in the actual development has yet to be decided.

`````````````

"If you were to fully develop the entire tree for all possible chess moves, the total number of board positions is about 1, 000, 000, 000, 000, 000,000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, 000, or 10(power of 120 ), give or take a few. That's a very big number. For example, there have only been 10( power of 26 )nanoseconds since the Big Bang. There are thought to be only 10( power of 75 ) atoms in the entire universe.

When you consider that the Milky Way galaxy contains billions of suns, and there are billions of galaxies, you can see that that's a whole lot of atoms. That number is dwarfed by the number of possible chess moves. Chess is a pretty intricate game!

**No computer is ever going to calculate the entire tree. **

What a chess computer tries to do is generate the board-position tree five or 10 or 20 moves into the future."

and never mind if the Computer's take over -- you will always have Governor Arnie to deal with those 'beasts' :)


ps I heard it Plays Half Life 2 well :) Will it be our home PC in 10 years? :) :)

This conversation is currently closed to new comments.

25 total posts (Page 1 of 3)   01 | 02 | 03   Next
| Thread display: Collapse - | Expand +

All Comments

Collapse -

Go

by gene.fellner In reply to The World's Fastest Compu ...

I'm a go player. The reason is that in 1965 one of my professors who wanted to write software to play go trained his students to be its opponents. An Asian businessman offered a million-dollar prize for a program that could simply play the game at all, following the rules and not making suicide moves.

Thirty years later, after computers "solved" checkers as a puzzle and learned chess well enough to beat all but the best human players, go programs were finally available that were only marginally stronger than the original "don't-break-the-rules" model. I don't know if the prize was still available then.

Today, the strongest go software is rated around 10-kyu (professional scale), which I'd say is roughly equivalent to 1100-1200 in chess. They're useful for teaching beginners (the lowest rating is about 30-kyu) -- but then again they're not really. They don't play like humans. Obviously there's nothing wrong with that, but they don't teach strategies and tactics that will hold up as the student becomes stronger and has to hold up against stronger opponents. Their approach to the game is flawed. I don't think they'll get much stronger. The next generation of go software will have to be thought out from the ground up.

One of the most basic skills to learn in go is pattern recognition. Your eyes have to analyze the shape of a group of stones and send the results of the analysis to your brain. You can't possibly think through that analysis consciously or you'll take ten hours to finish a game. The patterns are taught in typical Oriental fashion with lovely names and metaphors referring to war and nature. Not something easily digitized.

The most important thing I teach my go students is that it is not too difficult to figure out the right moves. There are always several nice ones available at every turn. What's hard, and what determines the outcome of the game, is figuring out which one is the most important one to play first. I think that's probably the most difficult thing to program.

A go board has 361 spots, each of which can be unoccupied or occupied by a white stone or a black stone. Solving that "problem" as a tree is beyond the capability of any foreseeable computer. The position must be decomposed into components, which must first be recognized and identified.

I think that humans will have the upper hand at this type of mental activity for some time to come.

Collapse -

Have any of these computers

by Tony Hopkinson In reply to The World's Fastest Compu ...

invented a human ?
There you go.
Problem solved. I can beat any chess program in the world. I'll just program it to lose when it plays me.
Now who's clever ?
Computer are neither intelligent nor stupid, they are well programmed or not. To get past that barrier they won't be computers amymore they'll be AIs.

Collapse -

and the subliminal message is.....

by Black Panther In reply to Have any of these compute ...

How smart is the brain! :)

Collapse -

Well mine

by Tony Hopkinson In reply to and the subliminal messag ...

is quite smart.
There are a lot of people who are more intelligent than me though.
How do you measure how intelligent someone is. Someone who is capable of walking into a room of strangers and getting everyone there to like talking to them could be considered to have a very high intelligence compared to a mathematical genius who failed extremely badly as the same exercise. Put them both in his environment and the mathematician wins hands down.

There is no comparison between a computer and a human though, the computer is the program(s), and they are as smart as the people who programmed them.
Any computer program playing a game always plays to the best of it's ability, a human will sometimes through inattention play worse and sometimes through intuition play much better.

In an environment with a limited set of fixed rules and a finitely approachable set of probabilities, we are cabapable of writing programs to look smart in a very limited way.

We have n't come close to programming a computer for pattern recognition, sure we can it have 'notice' patterns we've programmed it to recognise, but thats about 1% of the story.

As for how smart we are really, well for a race that can clone animals, split protons , genetically modify other organisms and reach space. We also came up with slavery, atom bombs, biological weapons and tip our rockets with bombs we are afraid to use.
So all in all we are as dumb as the hole in a cows backside on average.

Collapse -

Human vs Computer ( Computational Speed )

by Black Panther In reply to Well mine

Processing Power and Speed
The human brain - We can only estimate the processing power of the average human brain as there is no way to measure it quantitatively as of yet. If the theory of taking nerve volume to be proportional to processing power is true we then, may have a correct estimate of the human brain's processing power.

It is fortunate that we understand the neural assemblies is the retina of the vertebrate eye quite well (structurally and functionally) because it helps to give us a idea of the human brain's capability.

The retina is a nerve tissue in the back of the eyeball which detects lights and sends images to the brain. A human retina has a size of about a centimeter square is half a millimeter thick and is made up of 100 million neurons. Scientists say that the retina sends to the brain, particular patches of images indicating light intensity differences which are transported via the optic nerve, a million-fiber cable which reaches deep into the brain.

Overall, the retina seems to process about ten one-million-point images per second.

Because the 1,500 cubic centimeter human brain is about 100,000 times as large as the retina, by simple calculation, we can estimate the processing power of a average brain to be about 100 million MIPS (Million computer Instructions Per Second ). In case you're wondering how much speed that is, let us give you an idea.

1999's fastest PC processor chip on the market was a 700 MHz pentium that did 4200 MIPS. By simple calculation, we can see that we would need at least 24,000 of these processors in a system to match up to the total speed of the brain !! (Which means the brain is like a 168,0000 MHz Pentium computer). But even so, other factors like memory and the complexity of the system needed to handle so many processors will not be a simple task. Because of these factors, the figures we so childishly calculated will most probably be a very serious underestimate.


The computer - The most powerful experimental super computers in 1998, composed of thousands or tens of thousands of the fastest microprocessors and costing tens of millions of dollars, can do a few million MIPS. These systems were used mainly to stimulate physical events for high-value scientific calculations.

Here, we have a chart of processor speeds for the past few years.


Year
Clock Speed (MHz)
Instruction Rate (MIPS)

1992
200
200 (400)

1993.5
300
300 (600)

1995
400
800 (1600)

1996.5
500
1000 (2000)

1998
600
2400 (3600)

1999.5 700 2800 (4200)
2000
1000
?


From the chart above, we can observe some break through s in microprocessor speeds. The current techniques used by research labs should be able to continue such improvements for about a decade. By then maybe prototype multiprocessor chips finally reaching MIPS matching that of the brain will be cheap enough to develop.

Improvements of computer speeds however have some limitations. The more memory it has, the slower it is because it takes longer to run through its memory once. Computers with less memory hence have more MIPS, but are confined to less space to run big programs. The latest, greatest super computers can do a trillion calculations per second and can have a trillion bytes of memory. As computer memory and processors improve, the Megabyte/MIPS ratio is a big factor to consider. So far, this ratio has remained constant throughout the history of computers.

So who has more processing power ?
By estimation, the brain has about 100 million MIPS worth of processing power while recent super-computers only has a few million MIPS worth in processor speed. That said, the brain is still the winner in the race. Because of the cost, enthusiasm and efforts still required, computer technology has still some length to go before it will match the human brain's processing power.

Collapse -

Simplistic to the point of disappearing

by neilb@uk In reply to Human vs Computer ( Compu ...

You seem to be suggesting that the human brain works in the same was as Intel's finest in some linear processing fashion. It doesn't! So your comparisons are about as relevant as comparing, say, the airspeed of a peregrine falcon and a brick and concluding that the brick is a better flyer.

Can we have a little logic, here?

Collapse -

Not at all

by Black Panther In reply to Simplistic to the point o ...

It was only a 'far fetched' similation and to be taken in that light! :)

Collapse -

Processing speed

by Tony Hopkinson In reply to Human vs Computer ( Compu ...

What has that got to do with intelligence ?
The average human processes more information in a day, than the every computer on the planet in a year.
And most of it is pattern recognition. I can hear the telly the missus is watching, my keys clicking and and the hum of my not as silent as made out processor fan. That's not to mention the smell of coffee, the taste of buttered toast, the heat on my skin, the itch that indicates a shave is in order and I haven't even started on my various visual inputs. On top of that I'm communicating with you through a vicarious medium. There isn't a computer in existance that can match that feat.

Collapse -

as you say...

by Black Panther In reply to Processing speed

It's just one small part of it!

Collapse -

The Rationale behind Brain and Computer thinking!

by Black Panther In reply to The World's Fastest Compu ...

A lot of research has been able to show us which parts of the brain are active during certain
activities, but not how those parts actually function.

Probably the most comprehensive idea we came up with was to use computers to imitate the brain, to see if this would help us to understand it better.


It is possible, in my opinion, that however our brain was made (whether it be by Evolution or God), it was formed intentially so that we should never understand how it works. If we did know everything about the brain, then we would also be able to manipulate it in many ways. This may be
seen as a good thing or a bad thing, depending on how you look at it.

It could have serious moral implications, as we
could reach a stage where peoples' private thoughts were no longer private, where thoughts could be manipulated, memories could be erased, the possibilities are endless. Maybe humans
should stop being so arrogant that they think they can do anything.


On the other hand, sometimes psychologists are arrogant in wanting to know exactly how the brain works, so we look to computers to help us.

One similarity is that if you give a computer a command that it has never encountered before and
is not programmed to understand, it will not do it. In the same way, if a human brain encounters something unknown, like words spoken in a foreign language, it will not understand.

There is a discrepancy in this however. That is that if you continue to give this computer the same command over and over, you will not make any progress. On the other hand, if foreign language is encountered enough by the human brain, it
will eventually pick it up and learn what it has heard. This is because there are other factors involved. The human brain can recognise facial expressions, gestures and even percieve
emotion. In this way, it will eventually learn what is being said, even though the initial words have remained the same.

Computers cannot recognise emotion or facial expression, they do not have a body language all of their own, so they do not have the same capacity for learning that we do.

That brings question to the Brain and the Mind??

Do we also have the added advantage over the computer using our Minds?

Is there something called the mind which is non-physical,or everything including our mind is just a set of bounded physical set of objects ?

If the latter is true, then everything at any given instant of time in future is theroretically
predictable if we can model the whole universe as a huge state matrix at this instant of time.

However if we postulate that there is some unpredictability in future if only due to the unpredictability of our own behaviours,
then it is attributable to a non-bounded, non-physical model of mind.

Then the other interesting question arises :-
If mind is not matter, then how can it control the brain which is constiting of matter ? In our model of today's laws, only physical entities (mass, energy, momentum) can cause a change in a physical object.

IF THE BRAIN CAN BE INFLUENCED BY NON-MATTER (WHICH IS OBVOIUSLY NON-BOUNDED IN SPACE), CAN "MY MIND" INFLUENCE "YOUR BRAIN" TOO ?


Rationalism is a desired trait, but trying to explain everything away with the limited set of models that exist today will actually stifle the creation of new models.

Back to Networks Forum
25 total posts (Page 1 of 3)   01 | 02 | 03   Next

Related Discussions

Related Forums