what is best graphics card I can use

[AzN]

Originally posted by: brblx
you're wrong.

more ohm's law than joule's law, though. i guess either one arrives at the other. V(voltage) = I(current) x R(resistance). and P(wattage) = I x V.

increasing voltage does not magically double anything. it's linear.

When I used ^ that is power of.

So all these calculators are wrong on the internet? It's not simple as I x V.

 

[brblx]

sorry, i'll use my time machine to go back and tell that ohm guy he's wrong and that all modern electrical theory has been disproven by you.

 

[vj8usa]

yh125d and brblx, you're both assuming that current is remaining constant. Power = Current * Voltage, but Current = Voltage / Resistance. Thus, Power = Voltage^2 / Resistance. If Resistance is held constant, Power will indeed increase as the function of Voltage squared.

I'm not sure how current and resistance change in a CPU, but I'm inclined to think that resistance won't change if you don't change the CPU/FSB clock frequencies.

 

[AzN]

Originally posted by: brblx
sorry, i'll use my time machine to go back and tell that ohm guy he's wrong and that all modern electrical theory has been disproven by you.

Well that Ohm guy wasn't born in the modern day of CPU overclocking with hundreds of CPU calculators that say 20% for 10% voltage increase. Unless you are an expert mathematician with an electrical engineering I suggest it's probably best you stay out of these type of threads to preach to someone.

 

[vj8usa]

Originally posted by: Azn
Well that Ohm guy wasn't born in the modern day of CPU overclocking with hundreds of CPU calculators that say 20% for 10% voltage increase. Unless you are an expert mathematician with an electrical engineering I suggest it's probably best you stay out of these type of threads to preach to someone.

Ohm's Laws and Joule's Laws are certainly applicable today (which is why they're "laws"). P = I * V still holds true, but I is a function of V (see my above post) which could explain why the calculators you're referring to show P increasing as a function of V^2.

 

[AzN]

Originally posted by: vj8usa

Originally posted by: Azn
Well that Ohm guy wasn't born in the modern day of CPU overclocking with hundreds of CPU calculators that say 20% for 10% voltage increase. Unless you are an expert mathematician with an electrical engineering I suggest it's probably best you stay out of these type of threads to preach to someone.

Ohm's Laws and Joule's Laws are certainly applicable today (which is why they're "laws"). P = I * V still holds true, but I is a function of V (see my above post) which could explain why the calculators you're referring to show P increasing as a function of V^2.

I never said it didn't imply but I know there's something wrong with brblx equation and not as simple as I x V. Thank you for that explanation though.

 

[brblx]

Originally posted by: vj8usa
yh125d and brblx, you're both assuming that current is remaining constant. Power = Current * Voltage, but Current = Voltage / Resistance. Thus, Power = Voltage^2 / Resistance. If Resistance is held constant, Power will indeed increase as the function of Voltage squared.

I'm not sure how current and resistance change in a CPU, but I'm inclined to think that resistance won't change if you don't change the CPU/FSB clock frequencies.

whoops. you're right. yeah, assuming resistance is a constant, you've gotta recalculate your current draw with a voltage increase.

Originally posted by: Azn
Unless you are an expert mathematician with an electrical engineering I suggest it's probably best you stay out of these type of threads to preach to someone.

unless you're an expert at the difference between your ass and a hole in the ground, i'd suggest you stay out of ALL types of threads. there's a difference between being right by accident and actually having any clue about what you're discussing.

edit: yes, it IS as simple as P = I x V, you just have to recalculate your numbers in the other equation (V (or E) = I X R). again, if you think that these equations are simply 'wrong,' you're retarded.

 

[dclive]

Originally posted by: yh125d

Again, the only load present is GPU load. CPU is barely being used. If you read Anandtechs disclaimer (I assume you're referring to this article http://www.anandtech.com/video...wdoc.aspx?i=3520&p=11) they noted that the CPU and memory were not being taxed, and to expect an extra 50-100w when gaming for example

I wasn't referring to that article (I was referring to the article in the link I posted) but that's still a good article. Thanks for posting.

 

[dclive]

Originally posted by: yh125d

Originally posted by: dclive

Originally posted by: happy medium

Originally posted by: dclive

Originally posted by: Azn


So tell me do you use a 300/350 watt power supply to power a GTX 280? If you don't why do you suggest running such power supply with such video cards when you haven't had any experience with it other than just reading someone else's testing which doesn't tell much other than that they are running a much higher rated power supply to test for wattage?

Someone suggested that a GTX 280 used 236W.

I showed two tests - from two good sources - that showed it doesn't. Both tests show the entire system run using under 300W.

Are Anandtech and Tom's Hardware unreliable?

No they are not unreliable but do give the imppression to novice users that you can use a cheap 300/350 watt psu to run a high end gaming system just because it only draws 275 watts from the wall. It depends on 12v+ line amperage , psu efficiency and the ambient temperature which these components operate in.

The op has a 300watt (el chepo) psu and should not even consider running anything more then a 4770 if that even.

Why does PSU efficiency have anything to do with this? It seems like you're now saying it's _nothing_ to do with wattage (which I agree with - but PSU efficiency is usually tied to wattage) and everything to do with the 12V rail power - is that right?

Efficiency always matters. Especially when you're judging based on AC draw. For example...

System A and B both require 250w DC at full load. System A has an 85% efficient PSU at that load. System B has a 75% efficient PSU at that load. System A will pull 250w/.85 = 294w AC from the wall. System B will pull 250w/.75 = 333w AC from the wall. By only changing efficiency, system B appears to be using 40w more than system A, when its really using the same power.


Also, an efficient PSU will run cooler, as more AC power is being converted into DC power rather than thermal power. Cooler running PSU's deliver cleaner, stable, quieter power and operation

Agreed with all you wrote above, but what does that have to do with PSUs and which to buy? B is less efficient, and so worth a little bit less $$, but aside from that, ??

 

[AzN]

Originally posted by: brblx
unless you're an expert at the difference between your ass and a hole in the ground, i'd suggest you stay out of ALL types of threads. there's a difference between being right by accident and actually having any clue about what you're discussing.

edit: yes, it IS as simple as P = I x V, you just have to recalculate your numbers in the other equation (V (or E) = I X R). again, if you think that these equations are simply 'wrong,' you're retarded.

ROFL.. IS that the best you can do to personal attack? A guy implied I was a retard and an asshole because I was right all along and he was wrong. lol

I suggest you GTFO out of this thread if you don't have anything constructive to say.

Edit: I just wonder how I got my math? A calculator does not teach you but I knew the equation before hand which means I know what the fuck I was talking about but couldn't quite explain the equation while you ONLY cited ohm's law to tell people they are wrong without a full explanation.

 

[yh125d]

Isaac Newton was around way before the internet, and I read on 4chan once that gravity didn't exist, so obviously Newtons law of universal gravitation is false.

vj8, resistance will not be constant. As the voltage increases, so will the temperature. Higher temperatures = higher resistances. higher resistance coupled with p=i^2/r results in a linear power increase. Ohm's law rearranged to R=V/I shows that as voltage increases, so will resistance. As current increases, resistance will decrease. Only when voltage and current increase equally will resistance be constant


Azn, I feel the need to reiterate my point. In a given situation, changing only a 10% voltage increase will result in 10% power increase. Always. Forever. Every time. There is no refuting that. There is no disproving that. This is not my opinion, this is part of the fundamental laws of electricity.

 

[yh125d]

Originally posted by: Azn

Originally posted by: brblx
unless you're an expert at the difference between your ass and a hole in the ground, i'd suggest you stay out of ALL types of threads. there's a difference between being right by accident and actually having any clue about what you're discussing.

edit: yes, it IS as simple as P = I x V, you just have to recalculate your numbers in the other equation (V (or E) = I X R). again, if you think that these equations are simply 'wrong,' you're retarded.

ROFL.. IS that the best you can do to personal attack? A guy called me a retard because I was right all along and he was wrong. lol

I suggest you GTFO out of this thread if you don't have anything constructive to say.

Edit: I just wonder how I got my math? A calculator does not teach you but I knew the equation before hand which means I know what the fuck I was talking about but couldn't quite explain the equation.

If you knew the equation before hand, and you saw that a calculator said 20% increase when it obviously should have been 10%, you should have said to yourself "wtf?" and never used that calculator again

----------------

Keep it on topic guys, personal attacks do nothing but put you at the risk of being banned, warranted or not

 

[AzN]

Originally posted by: yh125d
If you knew the equation before hand, and you saw that a calculator said 20% increase when it obviously should have been 10%, you should have said to yourself "wtf?" and never used that calculator again

I read an article explaining all of this but I had forgotten where that number doubled. I knew it was 20% and this is the reason why I said 20% increase in watts with 10% increase in voltage.

I never used that calculator in my life but randomly picked one from the net to show you how I got 20%.

 

[yh125d]

Originally posted by: Azn

Originally posted by: yh125d
If you knew the equation before hand, and you saw that a calculator said 20% increase when it obviously should have been 10%, you should have said to yourself "wtf?" and never used that calculator again

I read an article explaining all of this but I had forgotten where that number doubled. I knew it was 20% and this is the reason why I said 20% increase in watts with 10% increase in voltage.

I never used that calculator in my life but randomly picked one from the net to show you how I got 20%.

Again. FUNDAMENTAL LAWS OF ELECTRICITY DONT CARE WHAT YOU READ IN AN ARTICLE


P=IE. E increases by 10%, and so does P. That is the end of it

 

[brblx]

minor changes in resistance aside, i think you're wrong, duder. the only way for a totally linear increase in power is for resistance to increase equally with voltage, which it will not. if you have a 4ohm resistor and double the voltage to it, it's not going to turn into an 8ohm resistor in order to keep current constant. the heat may result in a slight increase in resistance, but not enough to completely nullify any current increase (and any current increase would disrupt a linear relationship). P = I x E only works when your I value is correctly derived from E = I x R.

and azn, just go the hell away. it doesn't matter if you're 'right' when you have no comprehension of the topic being discussed. go find another thread to troll in with your regurgitative google 'knowledge'.

 

[yh125d]

Originally posted by: brblx
minor changes in resistance aside, i think you're wrong, duder. the only way for a totally linear increase in power is for resistance to increase equally with voltage, which it will not. if you have a 4ohm resistor and double the voltage to it, it's not going to turn into an 8ohm resistor in order to keep current constant. the heat may result in a slight increase in resistance, but not enough to completely nullify any current increase (and any current increase would disrupt a linear relationship). P = I x E only works when your I value is correctly derived from E = I x R.

and azn, just go the hell away. it doesn't matter if you're 'right' when you have no comprehension of the topic being discussed. go find another thread to troll in with your regurgitative google 'knowledge'.

That's not what I'm arguing. Of course resistance won't increase linearly, but that's not what I was trying to get across to azn.

Power is determined by E, I, and R all together. But when you single one out like azn did, the power increase *is* linear

 

[brblx]

say you have 2v and a 2ohm load. 2v = 2ohm x 1a. and 1a x 2v = 2w.

double the voltage to 4v. now, using E x I = P, you get a doubling of output to 4w. BUT, the other equation you used to determine your current flow is now invalidated. 4v does not equal 2ohm x 1a. voltage and resistance determine current, which would now increase to 2a. so your power equation is actually 2a x 4v = 8w, making power increase by a factor of four even though the voltage has only doubled.

 

[yh125d]

While true, it's not so much that voltage and resistance *determine* current, so much as they're all dependent on each other

 

[AzN]

Originally posted by: yh125d

Originally posted by: Azn

Originally posted by: yh125d
If you knew the equation before hand, and you saw that a calculator said 20% increase when it obviously should have been 10%, you should have said to yourself "wtf?" and never used that calculator again

I read an article explaining all of this but I had forgotten where that number doubled. I knew it was 20% and this is the reason why I said 20% increase in watts with 10% increase in voltage.

I never used that calculator in my life but randomly picked one from the net to show you how I got 20%.

Again. FUNDAMENTAL LAWS OF ELECTRICITY DONT CARE WHAT YOU READ IN AN ARTICLE


P=IE. E increases by 10%, and so does P. That is the end of it

Never said it did but increase isn't as simple as P=I*E but ^2 / resistance.

and brblx. If you explained it a little better instead of trolling and calling people names I would have listened but in your case my regurgitative google 'knowledge was correct. :roll:

 

[Silicon Spear]

Originally posted by: Azn

Originally posted by: dclive

Originally posted by: Azn

Originally posted by: vj8usa
Anyways, your earlier post implied that you thought a 300W PSU could power a system with a GTX280. That's what I really wanted to address, as that's almost certainly wrong. I'm not aware of any 300W PSU that could reliably power a full system with a GTX280 in it (and there's a good chance you'd be pushing your luck even with a 4850/GTS250).

Definitely. I don't see how people think they can run a GPU rated to 236watts on a 300 watt power supply when they aren't experts and just looking at consumption wattage. They don't factor in things like 12v wattage, aging capacitors, and the PS ability to run it in operating temperatures.

I have a hard time running a GTX 260 with a 450watt power supply with 2 12v rails.

Which GPU is rated to 236 watts?

How do you have a hard time? Does it sometimes not start, or ?

GTX 280 is rated at 236watts.

If I raise voltage on my CPU I get constant restarts when I'm playing a game.

It's probably that the CPU can't handle that much overvolting. You can only overvolt CPUs so far before they become unstable, no matter how much of a power supply you have. If it's supposed to overvolt that far, then it could be slightly defective. =P Really, the way I see it, all CPUs have different amounts they can be overvolted. The companies are supposed to give the CPUs a little headroom, but if they don't, then I'm not sure you can complain, since they do work fine at stock voltage.

 

[Keysplayr]

Hehe. Google warriors.

 

[yh125d]

Originally posted by: Keysplayr
Hehe. Google warriors.

No google here, 2 years Engineering/Digital Electronics

 

[brblx]

i works on teh cars, which luckily don't use electricity. ;P

either way that azn dude is a troll in every thread i see him. i feel all fuzzy inside that he thinks i'm one, too. my problem is that i just don't shut the hell up as easily as i should, i guess. 'least i actually know when i've made a mistake and why.

 

[Fox5]

Originally posted by: yh125d

Originally posted by: Azn

Originally posted by: dclive

Originally posted by: Azn


Why don't you explain to me why I get ramdom reboots when I raise the voltage on my CPU with a 450 watt power supply with 2 12v @ max 30A? I didn't get random reboots when I had G92 8800gts in the same system when I raised voltage BTW nor I get random reboots at default voltage.

A GTX280 doesn't need 236W by itself, which is the only point I'm rebutting. As one can see by looking at the Anandtech and Tom's Hardware review, the GTX 280 / 285 likely only need a fraction of that 236W power because an entire system only needs about 260W of power - with a pretty modern system being powered.

If your system needs more power due to overclocking a CPU or whatnot, go for it. That has nothing to do with the GTX 280/285's power requirements.

Raising voltage 10% raises 20% power usage on the CPU itself

False. Voltage and power are linearly related, a 10% increase in voltage is a 10% increase in power when current is constant

P = V^2/R. Multiply by time to get the amount of heat energy. (Q = V^2/R * t)
From what I recall (and I could be wrong), for a cpu, heat is square with voltage, linear with current (and thus roughly linear with clockspeed). However, as leakage occurs at higher speeds (well, some relation between speed, temperature, and voltage), the relation becomes closer to exponential.

 

[AzN]

Originally posted by: brblx
i works on teh cars, which luckily don't use electricity. ;P

either way that azn dude is a troll in every thread i see him. i feel all fuzzy inside that he thinks i'm one, too. my problem is that i just don't shut the hell up as easily as i should, i guess. 'least i actually know when i've made a mistake and why.

Thanks for pointing that out I was wrong without a clue of knowledge and citing ohm's law without even research. When that didn't work you called me names. You really need to look at your actions before talking shit about others.