Has anyone measured the speed of human brain (in FLOPS)?

Deepak

B3D Yoddha
Veteran
I just remember reading an article in newspaper reporting about some research IIRC by a BBC future tech group (something like that), and the lead scientist claimed that in few decades we could match human brain's computing power. And he gave the example of Playstation 3 :smile: as a TFLOPS monstor.

My question is has anyone ever measured a human brain's FLOPS?
 
http://en.wikipedia.org/wiki/FLOPS

Humans are even worse floating-point processors. If it takes a person a quarter of an hour to carry out a pencil-and-paper long division with 10 significant digits, that person would be calculating in the milliFLOPS range. Bear in mind, however, that a purely mathematical test may not truly measure a human's FLOPS, as a human is likely also processing smells, sounds, touch, sight and motor coordination. This takes an average human's FLOPS up to an estimated 10 quadrillion FLOPS (roughly 10 PFLOPS ) according to.
 
PFLOPS means PetaFLOPS? When will be able to have that kind of processing power at our desktop? And does anyone know how many Bytes of memory does a human brain have?
 
I would of thought sight must take the most processing of all the senses though. its not just processing a flat image like a camera might.. its also processing distance colour shadow size depth perception.. its a complex thing..
 
Deepak said:
PFLOPS means PetaFLOPS?

Yes.

When will be able to have that kind of processing power at our desktop?

Not soon. Supercomputers aren't even close.

And does anyone know how many Bytes of memory does a human brain have?

Estimates vary from a few hundred megabytes to thousands of terabytes. I think they believe there is a lot of redundancy so I suppose it depends on how you count that.
 
well simple way to look at it might be.. how much power would it take to just be able to get a system to percieve everything we do naturaly. then add ontop out computational skills.

The system to be level with us would have to do this..

be able to control all the muscles of the body. i think thats 264?

be able to process with the same visual skills we have.

be able to feel tell texture heat cold etc. (dont forget feeling includes every nerve ending in the body)

be able to taste

smell

hear..

cross reference all of the above with each other simultainously

and then also be able to control all the automated functions of the body.


figure out how much precessing power a computer would need to do that.
 
Here's a quote from a NewScientist article I read awhile ago (24th September2005)

Supercomputers are already at 100 trillion (10^14) calculations per second (cps), and will hit 10^16 cps around the end of this decade, which is the level I estimate is required to functionally simulate the human brain.
 
I'm willing to bet we'll never know, the brain can grow and gain knowledge and therefore increase processing power simply through learning, also we become more efficent. To actually simulate a human body, and brain I think is impossible, at least impossible in any current type of computer, even super computers.
 
I read on one of IBM's articles that in a decade or two, the current CMOS process will reach dead end, and then we will have to find new types of computing. What will be next big thing? Quantum computing?

I think computing power to match brain will be there soon, but the difficult task to is desing a software to emulate brain.
 
CMOS growth in 2D is dead within in the next 50 years.

So we got really 2 options either faster clock rates ( either using another technology or a different substrate that is CMOS compatible ) or 3d chips.
 
I forget where i read it but it seems that the general attribution that humans on average use only 10% of their brains calculated against both storage and processing power!

  • So if over say 100 years you store at max 10% of your brains capacity you have a 1000 years worth of storage capacity inlcuding sights smells sounds pattern recognition and just as importantly vivid memories.
  • its interesting that those that are considered geniuses either process faster (using >10% actively) or remember more and more accurately than the average person.
 
blakjedi said:
its interesting that those that are considered geniuses either process faster (using >10% actively) or remember more and more accurately than the average person.

So, in layman's lan. a genious has P4/AMD 64 and an average person 386, a genious has DDR3 and an average person SDRAM. :LOL:
 
actually a moron has a first gen P4 (will something) and a genius has a 4 way dual core opteron.
One can not lump in the shitty marketing P4 and the awesome A64:!:
 
blakjedi said:
I forget where i read it but it seems that the general attribution that humans on average use only 10% of their brains calculated against both storage and processing power!
That number most likely originates from some form of popular science source, and likely has very little connection with reality.

As for the guy who claimed 10^16 flops or whatever would be enough to simulate a human mind, ehhh, no, I don't think so. A brain functions in such a fundamentally different manner compared to a digital computer it might not be practical/possible to simulate our brains at all, considering we have neurons with thousands of interconnects in some cases. The processing requirement would likely be so large that the physical layout of the mainframes needed to run the brain would introduce weird delays in the simulated neural interconnects...
 
We dont use above 10% of it at any given time I believe is the correct quote, not "we only use 10% of our brains".
 
Back
Top