is underCLOCKing GPUs or anything any use towards battery life and heat development?
Arent the voltages the only important factor? Because my T8300 runs with x11 multiplier with the lowest voltage, ergo uses as much power as if the CPU would downclock itself to x6.
If all that is the case, overclocking a gpu wouldnt use up more battery because the voltages stay the same, right?
That would lead to the question, htf's someone sposed to be able to find the max clock for 2D and low power 3D?
While Im at it, just to be sure, is there any way to change the voltage of nvidia gpus expect for modding the bios? Like doing the same thing the powermizer does, but just user controlled instead.
-
-
volts = amount of energy carried by each electron.
sort of off topic lol -
underclocking gpu's is different than underclocking a cpu. First of all, their are no multipliers on a graphics card processor. Therefore the core has basically a set number of speeds which it can run before it looses stability.
You can run your graphics card at say a 200Mhz core speed rather than a 500Mhz core speed, and the power used by the chip will be the same. This is because the graphics processor cannot run at a different level. The core will always use the same amount of power regardless of whether it is running at standard clocks or undervolted. If graphics manufacturers had a speed step or cool and quiet design for their gpu core's than yes you would notice a significant amount of power savings. However, this cannot be implemented. Instead new systems will have a hybrid graphics technology, which uses integrated graphics on battery or low graphics intensive programs, and then if a game is loaded a more powerful gpu will be turned on.
You can try underclocking your gpu, but you will not see any difference.
K-TRON -
You will gain one thing though : less heat.
-
-
-
No he's right
Amps = flow (speed) of electrons
volts = amount of energy per charge
he's right...................... -
Its great that ur all being so enthusiastically theoretical...
but id just like to practically clarify if ive gotten this right:
under-/overclocking a GPU has no effect on its power drain/battery life
Right? -
That's what I've found, same with heat. I find that it's the voltage that determines the temperatures since my CPU is the exact same. It runs at min and max multiplier at the same temp when the voltages are the same.
-
I am not an electrical engineer but I disagree, "Amps = flow (speed) of electrons"
And on further reading I am correct in disagreeing that Amp in anyway has anything to do with speed. The speed at which the electrons of 1amp will travel will vary depending on say for example a wire, size and material. A larger wire has a larger "charge-sea" so a wire twice as large as another of the same material will have the electrons moving at half the speed of the smaller. Some materials have a denser "charge-sea" so all else being equal 1amp will have electrons moving slower than a material with a less dense "charge-sea".
Anyway I know nothing about electricity. So you get what you pay for. -
Well..I know that voltage is a measurement of electromotive force, or how much energy is carried per coulomb, which is the measurement for the amount of an electrical charge. Current (measured in amperes) is the measurement of how many coulombs pass a given point in a circuit in ONE second. Essentially, current is speed, in a dumb way. 1 amp = 1 coulomb per second.
Wikipedia says:
"Since a coulomb is approximately equal to 6.24150948×1018 elementary charges, one ampere is approximately equivalent to 6.24150948×1018 elementary charges, such as electrons, moving past a boundary in one second."
That is essentially speed. TWO amperes means that in one second, TWO coulombs of charge pass that point in ONE second, meaning that they would have to go DOUBLE the speed in order for BOTH coulombs to pass that point. -
The current(measured in Amps) is the flow of the electrons, and it means how many of those electrons are flowing into the designated dirrection(the sense was is something chosen by scientist,it used to be from + to - now some say it`s the other way around.
The tension(measured in V) translates the potential difference between two points, which equivalates into how many electrons can come from point A to point B.
Now,how many of those things can come depends on the resistance of the wire,or of the way through which the electrons pass.
Simplest way put :
-You have a dam , and that dam has a higher potential than the water beneath it. That would translate into tension.The water is the current. So, you can regulate how many Amps go through the dam and down beneath ,but if the valley below can only withstand say 100 cubic metres of water, unleashing them 101 cubic metres would translate into overflow.
In terms of electricity , that would mean burning the circuitry.
Contrary to what many believe, it is not the tension that is the problem,but ALWAYS the current , since most of the wires are designed to withstand only a set max current.(IE - the dam potential compared to the valley beneath can range a lot and still be safe) -
I'm not saying it IS speed, but I know that current is the flow of electric charge, and a simple way to put it would be speed. One thing for sure, it IS the flow of electric current.
-
eleron911 I am familiar with the hose and high pressure nozzle analogy. But thought it fell short. unknowntt look at my 2 lane/4 lane example. Count the cars per second? Do you see why I have an issue with the use of the term speed? Speed actually starts to infringe on what voltage is.
Edit: I figured out your flaw in logic, same as dual cores. You think if you have a one ton truck to deliver two tons in the same time frame as one ton you must double your speed (I know 133% because of return trip but any way) buy a two ton truck and you don't have to increase speed (or use two one ton trucks). Why you want to limit the method of transfer is a different equation. And yes it can be created but is not relevant to the discussion at hand.
Edit 2: Your assumption that quantity and speed need to be directly porportional is flawed. A semi moves more weight per mile in a given time frame than eleron in his 911 going 140 KPH/MPH, see? -
yes it all makes sense, but I just used speed as a very simple way of explaining it to those who don't care about the real definition. They don't need to be directly proportional, I guess I didn't word it properly.
I understand your examples and explanations completely. Well done powerpack. And since you're not an electrical engineer, +rep for you -
unknowntt, you are too kind, but thank you anyway. On a similar but unrelated subject. And I have not read about this for maybe 5+ years. The PC is getting so powerful it will replace the main frames. My brother works with main frames. PC's will never replace main frames. That is for all the reasons I stated before with eleron and his 911! If you want an orange delivered from California to NY? I would pick eleron. If you want 10,000 I pick a truck. That my friends is the reality, we call it "heavy lifting".
So back on point! Oh I forgot the point, sorry to OP. I think underclocking GPU will not substantially improve power issues. You should look into undervolting for the best results. -
Undervolt! This thread was WAY all over the place and WAY off topic hehehe. Anyways yeah, 911's a nice car!
Yay 300 posts lol, I'm so happy (little things make me happy these days =D ) -
I will send you a present soon. Congrats!
Anyway to OP ignore uk undervolting is the way you want to go!Trust me.
-
I`ve tried underclocking my GPU forcedly to the lowest possible values but for some reason...it didn`t help battery that much.Weird.
-
You're seeing the exact same thing I did, speed doesn't really matter for battery life, but rather the voltage that goes through the GPU which unfortunately we can't change.
-
Exactly. So why is there a powermizer option anyway?
I`m still reading on LukeAlexander`s thread here : http://forum.notebookreview.com/showthread.php?t=152460
I`m also a battery maximising freak, on battery I run my CPU at 600Mhz at the lowest voltage possible(0.95) . -
-
That would rly be cool, and theoretically it should be possible because powermizer cant be doing much more than would be possible via a custom programme.
eleron911: Give RMClock and undervolting a shot, i can undervolt my CPU to the lowest voltage with x11 multiplier, ergo 2100 MHZ.
underCLOCKing any good?
Discussion in 'Hardware Components and Aftermarket Upgrades' started by LinXitoW, May 24, 2008.