#### garylau

Sorry

i wonder if i download a lot of files/documents from the websites into the computer,then
how much weight did the computer gains in terms of kg or g
by the way
how did i convert mb/gb into kg ?thank you

Related Computing and Technology News on Phys.org

#### Ibix

Relativity theory does say that a change in the internal energy of something does change its mass. I'm not sure to what extent downloading a file changes the internal energy of your hard disk. Not a lot, I suspect.

That last is why you don't notice any effect. The change in energy (assuming there is one - I haven't thought about this in any detail) is tiny; the change in mass (in SI units) is 17 orders of magnitude smaller. Basically, a bacterium landing on your computer will have a lot more effect - and you'd never notice that.

#### Mark44

Mentor
Sorry

i wonder if i download a lot of files/documents from the websites into the computer,then
how much weight did the computer gains in terms of kg or g
None -- no weight gain. This is like asking whether your weight increases after you learn something new.
When you download a file, all that happens is that the states of the memory domains on the hard drive change.
garylau said:
by the way
how did i convert mb/gb into kg ?thank you

#### Quotidian

It's no heavier because information is immaterial. When you download a file, all you're doing is re-arranging the material that's already there; just a different arrangement of 1's and 0's.

#### Fervent Freyja

Gold Member
I reckon you would see an increase in the computers energy consumption before any kind of weight gain...

#### Quotidian

writing a file to a hard disk consumes energy, but I don't know about writing a file to a solid-state drive, as many drives now are. But without knowing, I would guess that the increase in average energy usage for writing a file to a solid-state drive would be extremely small. (Happy to be corrected by anyone who actually knows.)

#### rkolter

Energy is mass. Mass is energy. Information is stored in a pattern in your computer - the sequence of transistors flipped between 1 and 0 for example, for a solid state device. So your question is fundamentally, is the final state of the media imprinted with the data in a higher or lower energy state than it was before being imprinted?

The correct answer is that the answer is indeterminate. While there would be a change, that change would be so small as to be negligible and the final result could be either positive or negative depending on the exact initial state of the media, how the media stores its data, and the data you want to save.

#### Baluncore

There is a confounding variable here.
As you download more data, the weight of the computer increases due to dust accumulation.

#### cosmik debris

Storing information takes energy, if all energy taken from the power supply is returned as heat then no mass change will take place. However, storing information changes the state of the storage system to a more organised state, this reduces entropy, energy has been put into the storage system increasing its mass. All this good but unfortunately you will run out of storage space so you will have to delete stuff and that will increase entropy, heat will be given off and your mass will go down.

None of this mass change is practically measurable.

#### Baluncore

so you will have to delete stuff and that will increase entropy,
But, if you overwrite the deleted data with all zeros, then will entropy not decrease?

Does it really matter if the information stored is all zeros or some really useful data. If the bits stored at all address locations are defined, and do not take random shuffles with other addresses, then surely the two situations should have the same entropy.

#### Greg Bernhardt

I swear we've had a discussion on this in the past. Anyone remember?

#### rbelli1

Gold Member
The best I can find is that one bit represents on the order of 1000 electrons. A 1 TB hard drive has approximately 10^13 bits so 10^16 electrons. One electron is about 10^-34 grams. The maximum mass change from the two extremes of storage on that drive is 10^-18 grams. That is about 21 orders of magnitude smaller than your laptop.

The mass change from the "New Car Smell" coming off of it when you first turned it on is way bigger than any possible change to the storage.

BoB

I swear we've had a discussion on this in the past. Anyone remember?
When I first read this thread I wondered why my post got deleted. Then I found it on the one cosmik linked.

#### Ibix

But you don't add electrons to a hard disk to change its contents, surely? You just flip magnetic domains from north-is-up to north-is-down (or whatever). So any mass change comes from the difference in energy needed to create co-aligned versus contra-aligned magnetic domains. I suspect that's orders of magnitude smaller still, although I haven't calculated it. And it would be inversely related to run length, rather than a bit-wise cost.

#### rbelli1

Gold Member
But you don't add electrons to a hard disk to change its contents, surely? You just flip magnetic domains...
That is precisely how a hard disk drive (spinning rust) works. But a flash based "hard disk" (no disks involved) uses a change in charge via quantum tunneling to store information.

BoB

#### Guidestone

Hey guys, really interesting topic. I just wanted to say that if we are talking about solid state devices then logic gates are the ones who play the binary role. Logic gates made of transistors whose outputs reflect either a X volt with respect to ground or just ground and these are the states that reflect 1 or 0. Very tiny currents are involved in the process and the fact that there's a current means charges are not getting stored anywhere so as to cause weight to increase if we don't count parasitic capacitances.
In the case of hard drives, an electromagnet strong enough to change the direction of the magnetic field of each "grain" in the disk and each of those two directions is a 1 or 0.
Both situations of data storage don't seem to accumulate mass to me. Please correct me if I'm wrong. :)

#### rbelli1

Gold Member
charges are not getting stored anywhere
You need to change or store something to have a "data storage device". Look up how flash storage works.

change the direction of the magnetic field of each "grain" in the disk
Take two magnets near each other and move or turn them in relation to each other. What did you feel?

You have changed the amount of energy embodied in the arrangement. Energy is equivalent to mass.

Now we are talking infinitesimally (and currently un-measureably at the scale of an actual storage device) small amounts. Don't go to your local dollar store and buy a scale expecting the dial to move as you write to the storage in your laptop.

BoB

#### rootone

The thing is you can use lots of different technologies to digitally represent a 1 or a 0.
You could make two identical systems in which the representation of 0 and 1 are reversed, it would work the same.
A 1 bit does not have an intrinsic mass that is different to a 0 bit.

#### Logical Dog

the signalling convention just switches voltages between the ranges to represent 1s and zero, so yes..energy is increased. Now by how much or whether this causes an increase in mass that can be measured?
Both situations of data storage don't seem to accumulate mass to me. Please correct me if I'm wrong. :)
AFAIK the signalling convention most commonly used is of course that of electrical voltage ranges.

0 Volts____________________ V-low ------forbidden range----- V-high _________________Vmax

[Vhigh - Vmax] = 1. Voltages that fall between this range are a ONE.
(0 - Vlow] = 0. Voltages that fall between this range are a ZERO.

and the reason we use binary signalling convention is because the noise is less and its not as complicated to manufacture as using other signalling conventions.

You can look up the voltages for this signalling convention and use some physics to find the increase in mass. I THINK. I gues an increase in voltage is an increase in potential energy and that can be used to add to the total energy and somehow that energy can be converted to mass ( i suspect on very famous equation)

Last edited:

#### Guidestone

But doesn't the energy/mass equivalence only make sense at speeds approaching light speed?

#### Guidestone

You can look up the voltages for this signalling convention and use some physics to find the increase in mass. I THINK. I gues an increase in voltage is an increase in potential energy and that can be used to add to the total energy and somehow that energy can be converted to mass ( i suspect on very famous equation)
Isn't potential energy relative? As I know only differences in potential energy are useful.

#### Logical Dog

Isn't potential energy relative? As I know only differences in potential energy are useful.
I have no answer for these questions.

Yes PE is relative but an increase in voltage with some ground reference implies an increase in PE with the same reference yes? computer architecture is quite nasty but overall the digital signalling convention converts analog electrical voltages to digits. an increase in voltage means an increase in electrical potential energy between two points.

Last edited:

#### Ibix

But doesn't the energy/mass equivalence only make sense at speeds approaching light speed?
No. Note that nuclear reactors function by releasing some of the mass of their fuel as energy. This actually mostly comes from the binding energy of the nucleons - a potential energy that previously counted as mass.

It is true that (nuclear reactions aside) every day energies contribute negligible mass to a system because m=E/c2 provides a large conversion factor.

### Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving