Slashdot Mirror

Yeah, 3ghz doesn't come close to the light speed barrier. i think the issue is more from heat dissipation and electron bleed...

moving the gpu on-die will fix the latency associated with the pci-e bus, but it's not because of the reasons you seem to believe

Want to bet?

At 3 GHz, light moves just 7.2 cm, given a typical upper range for the velocity factor of copper of 0.72. Silicon and fibre optics are usually worse, with a VF between 0.4 and 0.6, or between 4 and 6cm per clock. That's barely enough to traverse a CPU die, let alone the motherboard. Moving parts physically closer together has a lot to do with the speed of light!

Looks like a photon travels 10 cm in a clock pulse at 3GHz.

we just have to respectfully disagree.

the mandelbrot set is a red herring because it is fractal.

on the other hand, even if the genetic code is fractal, we know its granularity: its final granularity is not on the level of atoms. Nowhere even close.

Its final granularity is on the level of cells.

We know, for a fact, that the genetic code describes a way to produce the cells in the human body.

What we do not know, as you point out, is the atomic interactions that produce proteins and so on. But those atomic interactions are simply not important.

You know, we are getting to where we can just arbitrarily change the source code and see what happens (ie genetically modified organism).

I don't care what you say, when for the next ten years we have teams of researchers working with a compiler and 600 megabytes of binary code, and they can move anything and get the result (ie they can breed a new mouse with whatever changes to its genetic code they want) you cannot pretend that we won't have things figured out.

Going back to reality, as opposed to your red herring of the mandelbrot set, the number of transistors in a quad-core Nahalem Intel i7 processor is 731M and costs $287. When you compare the physical atoms of the brain to the physical atoms of the i7, you will not find as large a difference between the number that constitute an atomic computing element as you would have us believe.

I want you to imagine the brain squished down into a 2D matrix that is 1 neuron tall. How much area does it take up? You are pretending that the answer is "a larger area than the Solar System".

But it obviously isn't a larger area than a solar system. If you take 2-3 pounds of stuff (the human brain) and use it to put a 1-neuron thick coat on as large an area as you can, you're still not talking about many square kilometers.

How many neurons are in the human brain? 100 billion.

What is the density of neurons in the human brain? 1200 cubic centimeters.

You divide the two, you get 83,333,333 neurons per cubic centimeter. The cube root of that gives you the fact that there are 436 neurons per centimeter in each dimension. So, how much larger is it if you need the same 83 million neurons but you are working with with square metal? The square root of 83 million is 9,128. So each cubic centimeter of the brain is 9128 square centimeters at the same density. Therefore, the 1200 cubic centimeters of the whole brain in three dimensions represents 25,123 square centimeters at the same density.

ie we are talking about 158.5 centimeters by 158.5 centimeters of silicon (sqrt of above number). Remember my i7 link? It says the die size is 263 square mm. So, if we just divide 25122 square centimeters by 263 square millimeters, we get the fact that 9552 processors, at a cost of $287 each, will represent the same area as the human brain, squashed.
Call it $400 per processing unit (along with motherboard, interconnects, but also bulk pricing on the chip itself), and you get 9552 * $400 = $3,820,800 of hardware.

I guarantee you an i7 can simulate 436 neurons by 437 neurons in a square centimeter of metal. But let's say it can't. Let's say it takes 100 square centimeters of metal just to simulate the 437 by 437 neurons found in a "square centimeter worth" of human brain. You multiply the figure of 3.8 million by a hundred, you get 380 million.

Honestly, simulating two to three pounds of stuff, when you have the source code, can play with it until you discover how it works, is not that hard.

I would be downright surprised if within my lifetime this was not done (and for well under fi

Yes, because we know that it's not possible for all different kinds of complexity to arise from a simple system.

(For those of you that can't be bothered to click the links, they depict different kinds of complexity arising from the simplest possible system.)

Yes, because we know that it's not possible for all different kinds of complexity to arise from a simple system.

(For those of you that can't be bothered to click the links, they depict different kinds of complexity arising from the simplest possible system.)

Yes, because we know that it's not possible for all different kinds of complexity to arise from a simple system.

(For those of you that can't be bothered to click the links, they depict different kinds of complexity arising from the simplest possible system.)

Yes, because we know that it's not possible for all different kinds of complexity to arise from a simple system.

(For those of you that can't be bothered to click the links, they depict different kinds of complexity arising from the simplest possible system.)

> did problem 2a on the board one afternoon(s). After 3 hours of derivation, we called it quits.

What was problem 2a? I'm just curious on what Wolfram Alpha would do with it: http://www.wolframalpha.com/

e.g.: http://preview.tinyurl.com/36gppwf

(click on "show steps").

So, a very dumbed-down and unrealistic example:

Suppose that for each digit of pi that is calculated, there is a 1 in 1 trillion chance that the hardware was faulty and produced an incorrect digit.

Probability of 100% accuracy: ( .999999999999 ) ^ (3000000000000)

Crunching those numbers gives about a 5% chance of 100% accuracy, in other words, according to the dumbed-down (but not so unbelievable) premises, there's a 95% chance that one of the digits is wrong.

is man actually capable of changing the properties of something as huge as planet Earth?

man? no.

But men - yes. Your intuition fails at the huge dimensions involved here, because it evolved to deal with the small immediate surroundings of you and your tribe on the plains of africa.

We are talking about 7 billion people, eating, shitting in the woods, making fire to cook their food, and - increasingly many of them - driving cars, flying planes, burning fuel to generate electricity and so on. Wolfram Alpha computes we use 86 million barrels of oil every day.

Unless you can create a picture in your mind of 86 million barrels a day, every day, build that up to a year, and then to a decade or five, I strongly suggest you stop relying on intuition and common sense and start relying on science and data.

Or if you ignore the sea
http://www.wolframalpha.com/input/?i=land+area+%2F+earth+population
=0.0219 km^2
Or three football pitches

I'm not sure how you get that speed from my numbers, but you seem to be treating acceleration as delta-acceleration. Wolphram Alpha confirms my speed number: 39.92m/s after 20 weeks.

Yea why the hell isn't the actual press release linked, instead of the shallow 3rd party blurb that filters out everything interesting and introduces its own egregious errors?

If the press release had been linked, maybe we would have been discussing the graphs instead of heretofore unknown super photons.

From the JAXA press release:

The thrust by solar light pressure is 1.12 mili-Newton (*2,) which is the expected value.

and

1.12 mN is equivalent to about 0.114 g

For some interesting perspective on this acceleration courtesy of Wolfram Alpha:

0.114g is:

* 1.118 m/s^2
* 4.025 km/hr/s
* 2.501 mi/hr/s

So every hour the velocity of Ikaros will increase by about 9,000 mi/h ( 14,490 km/h ). After a day it's speed will have increased by about 216,086 mi/h ( 347,760 km/h ).

This is assuming the acceleration continues linearly for a day. I have no idea if that's a reasonable assumption.

There are a few other interesting calculations on Wolfram: Wolfram Alpha

A 3000 square foot sail is about 16 metres across.

Am I missing some attempted nuance here?

3000 feet is 914 meters.

On the off chance you're talking about the length of one edge, the sqrt of 914 is 30 meters. So, still, nothing matches 16.

3000 square feet

Incidentally, outright genocide just made a little dent. The anomaly between 1990 and 2000 is visible; but in terms of long-term growth rates, turning the entire country into an abattoir had about the same effect as a slight bump in the condom supply would have.

I couldn't follow your calculations, but your answer is ok; http://www.wolframalpha.com/input/?i=(12*10^6+light+years%29+%2F+%281000000+km%2Fh+*+%2824*365+hours%2Fyear%29%29 Using units in the calculations makes it easier.

So what you say is that this will happen on Thursday the 7'th of April 2011?
http://www.wolframalpha.com/input/?i=15%2F20+years

This results in roughly 80 farads.

Which says nothing about the power stored without the voltage across the planes!

        W = C V^2 / 2

The energy stored goes as the square of the voltage. It drops precipitously as the voltage approaches 0, as it does in this case.

Even if you somehow managed to get 100mV of potential difference*, that's still only 800 mJ of stored energy, or about 2kJ/kg specific energy. Compare that to just burning the graphite at 32.8 MJ/kg (not counting the weight of the oxygen), which is about 16,000x greater!

See: Specific energy computation

You'll find that the energy stored in the internal stress of concrete is similarly low. Chemical energy density is surprisingly high, only nuclear power sources beat it.

Try it, look up the typical stress in a block of pre-tensioned concrete, and work out the J/kg and J/$!

*) Not likely! At that charge, the electric field strength between the layers is 700MV/m, which next to nothing can resist. See: Field Strength computation. Air breaks down far below that, so the stack would short out towards the sides where it is exposed, at the very least.

PS: All of these computations may be off by a few orders of magnitude, it's nearly 1:30 in the morning here...

This results in roughly 80 farads.

Which says nothing about the power stored without the voltage across the planes!

        W = C V^2 / 2

The energy stored goes as the square of the voltage. It drops precipitously as the voltage approaches 0, as it does in this case.

Even if you somehow managed to get 100mV of potential difference*, that's still only 800 mJ of stored energy, or about 2kJ/kg specific energy. Compare that to just burning the graphite at 32.8 MJ/kg (not counting the weight of the oxygen), which is about 16,000x greater!

See: Specific energy computation

You'll find that the energy stored in the internal stress of concrete is similarly low. Chemical energy density is surprisingly high, only nuclear power sources beat it.

Try it, look up the typical stress in a block of pre-tensioned concrete, and work out the J/kg and J/$!

*) Not likely! At that charge, the electric field strength between the layers is 700MV/m, which next to nothing can resist. See: Field Strength computation. Air breaks down far below that, so the stack would short out towards the sides where it is exposed, at the very least.

PS: All of these computations may be off by a few orders of magnitude, it's nearly 1:30 in the morning here...

Even if right now it's at it's maximum distance and it were skimming the surface of the star on the short end, it would still end up being a 1985 year trip. Calculation.

Formula for maximum resolution of a lens:
sin(theta) = 1.22 (wavelength / diameter of lens aperture)

Moving things around we get:
1.22 * wavelength / sin(theta) = diameter

1.22 * wavelength / sin(tan^-1(resolution length / distance to object))

With a wavelength of 500 nanometers (visible light), a resolution of a kilometer (might be enough to see that there's something interesting there), at a distance of 20 light years (the nearest known extrasolar planet is 10.4 light years, so let's be generous and assume that there's some sort of life within twice that distance).

The lens on the telescope would have to be over 115000 kilometers across, almost twice the radius of Saturn. Wolfram Alpha calculation. Up the distance and lower the desired resolution for a more realistic picture of what it would take (100 meters resolution at 200 light years means a lens diameter 17 times the radius of the sun. We'll have finished the Dyson Sphere long before we build that.)

I have always wondered this myself, but I guesstimate that it would require a lens the size of the Earth. Or the Sun. Or something impossible like that.

The problem is that you can't get details from long exposures. These far-off objects require exposures that are hours long. Imagine taking a 5-hour long exposure of a soccer game: all the players would be blurred. Now imagine that the players are running at the speed of a planet: upwards of 65,000 miles per hour. That is going to be one heck of a blurry picture.

So here is my back-of-the-napkin calculation:
This article says they photographed an exoplanet using an 8 meter telescope, with a 4 hour exposure. To get a good picture of a moving object, you need about 1/30th of a second. So we need a lot more light, which means we need a larger telescope. That 8 meter telescope (pretending it is just a circle) is 8*8*3.14 ~= 200 square meters. To get the same amount of light in a 1 second exposure would require 14,400 times more area. (14,400 seconds in 4 hours). Add another factor of 30 to get 1/30th of a second exposure, so the telescope is now 3,456,000 times bigger than the original 200 meter telescope. The exoplanet pictures we have are only a few pixels. So let's say we wanted a 1 megapixel image of the planet, so we need about 1 million times more light, so lets increase the surface area of the telescope another million times. So now the lens is 3,450,000,000,000 * 200 meters = 690,000,000,000,000 meters, which is slightly more surface area than the entire surface of the earth.

And that just gets you a 1 megapixel image of the planet. That won't show you a city. And I bet a 1/30th exposure would be too slow at the speed a planet moves - it would still be too blurry. So I think my estimate of something the size of the Sun was pretty close!

"This difference, however, will be "very small," said the study's co-author Marten van Kerkwijk of the University of Toronto, since the fastest possible orbital period is more than one thousand years.

If the period of rotation for two bodies is T = 2 * pi * (((length of semi-major axis)^3)/(G * (M1+M2))), then the time works out to be 5615 years and change. Anyone know why they're low balling the estimate so much?

You should tattoo pi * 1337%

Rather that a full up Navier-Stokes solver (I was going to say openFOAM too), since it's a short section of a course, maybe just have them play with the NACA airfoil potential flow solver; that's pretty neat. You can explain the simplifications to the governing equations between "real" CFD and potential flow, and show them that the simple models can still be useful in certain situations.

• 5 versions of the .Net Framework and Visual Studio
• 2 generations and 6+ versions of the Xbox gaming system
• 4 generations of the Zune music player
• 2 major desktop, 2 major server, and 4 Mobile/Embedded Operating Systems Updates

Not to mention large investments in online search, software as a service, and cloud computing. With the exception of their Online Services Division (MSN, Bing, Hotmail, advertising) Microsoft makes significant income from each of their product divisions and has more than twice the income that Apple does. Many of their business products are doing very well, and Sharepoint recently became their latest billion dollar sales product.

I will admit that Apple's products are more popular than Microsofts, but that is because they are tailored to the consumer market. Most business uses Microsoft because it costs less and makes users more productive. I personally think that the Zune HD and Windows 7 are great consumer products, and the Windows Phone 7 is designed to compete with the iPhone as opposed to the Palm OS for Windows Mobile, so it will be interesting to see how the next 10 years progresses.

there is unfortunately insufficient data to estimate the velocity of an African swallow (even if you specified which of the 47 species of swallow found in Africa you meant)

Wind energy this, Solar energy that. It's all fantasy dreamed up by hippies. It may or may not be able to meet a high percentage of our energy needs at some point in the future.

A Nice Dream: http://www.wolframalpha.com/input/?i=portugal+energy+wind

Nuclear power is here now. We know it works. We know it's safe, if done right. Sure, it's expensive, but if we'd invested a few trillion in nuclear power over the last 30 years ago we'd have ended up saving a shitload on foreign wars, cost to the environment from oil spills and pollution, etc...

And have more Three Mile Island and Chernobyl spills. No, I think I prefer an oil spill...

At the rate we're going now, nothing will have changed 20 years from now. Instead, we need to start building nuclear plants and investing in research on portable power like fuel cells so we can use that nuclear power outside of the main power grid.

Portable radiation sources! The solution to overpopulation, guaranteed...

Oh, I'm well aware indeed. The lifestyle you describe is for us few remaining middle class folks.

Our poor people buy these things. Open your eyes.

262 million cell phones subscribers in the united states.

227 million internet users in the united states

The median income of the united states is enough to afford plenty of shit we dont need.

Dont be a jealous ass. We are rich. The whole "upper class" vs "middle class" vs "lower class" shit is exactly that.. shit. That doesnt tell you how rich you are. How rich you are is irrespective of how much more other people have. If I had a 100 trillion dollars, would that instantly make everyone else poor? This tactic of class separation is tired old shit pulled by politicians that want to sell you a story and get your vote. They prey on your jealousy and you have fallen for it hook, line, and sinker.

TFA fails on basic mathematics. Let's take as given that there's 2117 cubic meters of water per capita per annum for each Chinese citizen.

The average person needs about 2 liters of water a day. Let's suppose they get all of that by drinking bottled water and soda from Coke's bottling plants. That amounts to less than 1 of those cubic meters per year.

Let's suppose the average person buys one new computer chip per week. Probably most people go weeks or months between purchases, but each device has many chips, so 1/week is about right. From this press release, it takes 10 gallons of water to make 1 computer chip. Oh gosh! That's two cubic meters per year!

To a rough approximation, all fresh water is used for farming. Water use for all other purposes is quite literally a drop in the bucket. Yes, wars have been fought over water, and they may be fought again in the future. But we're talking about agricultural irrigation here: everything else is negligible.

Now, in certain areas, water availability can be orders of magnitude less than the 2000 m^3/year average in the article, so water conservation there is a serious issue. But you don't grow crops in those areas ... and you don't build a chip fab plant there either.

TFA fails on basic mathematics. Let's take as given that there's 2117 cubic meters of water per capita per annum for each Chinese citizen.

The average person needs about 2 liters of water a day. Let's suppose they get all of that by drinking bottled water and soda from Coke's bottling plants. That amounts to less than 1 of those cubic meters per year.

Let's suppose the average person buys one new computer chip per week. Probably most people go weeks or months between purchases, but each device has many chips, so 1/week is about right. From this press release, it takes 10 gallons of water to make 1 computer chip. Oh gosh! That's two cubic meters per year!

To a rough approximation, all fresh water is used for farming. Water use for all other purposes is quite literally a drop in the bucket. Yes, wars have been fought over water, and they may be fought again in the future. But we're talking about agricultural irrigation here: everything else is negligible.

Now, in certain areas, water availability can be orders of magnitude less than the 2000 m^3/year average in the article, so water conservation there is a serious issue. But you don't grow crops in those areas ... and you don't build a chip fab plant there either.

WolframAlpha recently added space weather, including sunspot activity. Seems to me that the next cycle is just a bit of a late riser.

WolframAlpha recently added space weather, including sunspot activity. Seems to me that the next cycle is just a bit of a late riser.

I didn't believe this bit Microsoft's Project Natal to cost around $149. 'The figure for the standalone unit is significantly higher than a previous sub-£50 estimate, then I looked up the history of UK currency. Gee the pound has gone down. Look at the history since late 2007.

Wolfram|Alpha is great.

According to that excellent tool, Apple was valued higher than Microsoft through the '80s, as high as 3.2x as much as Microsoft. Then, right around the turn of the decade to 1990, Microsoft pulled ahead.

By 1998, Microsoft was worth 100x Apple.

Now, they're back up to even.

There are lots of Catholic schools in America (Catholics too, obviously) and they all teach that the Church has accepted the notion that man came about by the process of evolution, albeit a process conceived of and initiated by God. Also, I would guess that the vast majority of Christian schools in the country are Catholic, even though Catholics only make up 30 percent of US Christians.

...a cellphone left on continuously could represent as much as an oil tanker worth of energy usage. It also broadcasts EM waves everywhere -- if everybody in the world used their cellphones at the same time, imagine the impact.

Are you seriously that stupid? Or just trolling? Well, alright, I suppose I'll throw some nice troll food. Don't say I never did anything for you.

By some off the cuff calculations (with a little help from Wolfram Alpha), my 7.2 V, 3 Ah battery would equate to 21.6 Wh to fully charge, see here. Generally, leaving my cell phone on all day results in the battery draining about 20% to halfway, depending on how much I use it, by the time I go to bed. However, I'll be generous to your presumption here, and we'll presume for the sake of argument that when my phone goes on the charger before bed, I've totally exhausted the battery and it needs a full recharge.

So, let's presume I'm putting that full 21.6 watt hours into the phone every night. A barrel of oil contains 1.7 million watt hours. At that rate, my phone would have to be fully drained and recharged daily for a little over 78703 days, or 215.6 years, to consume one barrel of oil. Given that cellular phones were not quite invented yet in 1895, I kind of doubt that anyone's been charging their phone quite that much.

Isn't it nice to actually do the fucking math, instead of just parrot crap that you heard somewhere?

...a cellphone left on continuously could represent as much as an oil tanker worth of energy usage. It also broadcasts EM waves everywhere -- if everybody in the world used their cellphones at the same time, imagine the impact.

Are you seriously that stupid? Or just trolling? Well, alright, I suppose I'll throw some nice troll food. Don't say I never did anything for you.

By some off the cuff calculations (with a little help from Wolfram Alpha), my 7.2 V, 3 Ah battery would equate to 21.6 Wh to fully charge, see here. Generally, leaving my cell phone on all day results in the battery draining about 20% to halfway, depending on how much I use it, by the time I go to bed. However, I'll be generous to your presumption here, and we'll presume for the sake of argument that when my phone goes on the charger before bed, I've totally exhausted the battery and it needs a full recharge.

So, let's presume I'm putting that full 21.6 watt hours into the phone every night. A barrel of oil contains 1.7 million watt hours. At that rate, my phone would have to be fully drained and recharged daily for a little over 78703 days, or 215.6 years, to consume one barrel of oil. Given that cellular phones were not quite invented yet in 1895, I kind of doubt that anyone's been charging their phone quite that much.

Isn't it nice to actually do the fucking math, instead of just parrot crap that you heard somewhere?

Ugh, slashdot destroyed my link. Try this one.

Some conceptual errors in this, plus a massive units error. What do you mean by "average atomic weight of 9 a.u."? I'm going to assume you mean the slick is 1 molecule and 9 angstroms thick. Which isn't the way oil slicks work, but let's go with it.

Anyway, total units fail:

http://www.wolframalpha.com/input/?i=(10000+miles^2+*+1+angstrom%29+in+m^3

6000 gallons. Which is wrong because of course the slick *isn't* one molecule thick, but your calculation is meaningless.

Actually, not so fast. Roughly 6% of people who ever lived are alive today...

http://www.wolframalpha.com/input/?i=how+many+people+ever+lived+on+earth

Leaving 94% dead!

Not to mention that a single trillion barrels of oil equates to about 159 CUBIC KILOMETERS of oil, not to mention multiple trillions.

Citation

They just get the answers from Wolfram Alpha.

How much will it cost us later when we have to clean up all of this damn space debris and avoid collisions

I think you are vastly underestimating the vastness of space.

I think you are vastly overestimating the vastness of geosynchronous/geostationary orbit. We have a belt that's "only" 264,924 km in circumference, miniscule compared the the space we have in the variety of low-earth orbits. Most importantly, this is the only region of space that can be used for geostationary satellites. We can't go somewhere else if we clutter it up.

"Not sure, but wherever she is, it isn't here."

I was hoping Google still had that easter egg, but it doesn't seem to work anymore.

Isn't this what wolframalpha does? I wonder what those guys make of google muscling in on their turf (tho I suspect google's implementation won't have all the specialist maths and visualisation functions etc.)

Or maybe Wolfram:Alpha...

A reusable space vehicle which could be launched to retrieve or repair the satellite...

For the record, the space shuttle gets just over 250 miles altitude. The satellite in question is at an altitude of about 22,236 miles.

The amount of fuel needed to transit the space shuttle between these orbits is prohibitive. It was never designed for general purpose satellite repair; it was just a demo.

Is this a joke?

The rate of evaporation from the oceans is about 400,000 cubic kilometers per year.

To increase that by just one percent would mean pumping 4,000 km^3 of water.

Just raising that much water to 3,000 feet would take approximately, oh let's see, carry the 0x100,
about 1,651,445,966.51 horsepower. One Point Six BILLION horsepower.

Haha, I know you're joking, but I just worked this out: it would take just under 2 10^19 watts of continuous power, or about 20 billion gigawatts. For reference, a really big nuclear power plant is 1 gigawatt!

Assuming my maths is correct!

Of all those, I suspect Wikipedia is correct. If you plug the formula from wikipedia in wolfram alpha you get this. If you look at the AU page, that distance has been disputed. Also the approximation that the tan() can be ignored is wrong around the 8th significant digit.