NotebookForums.com › Forums › General Notebook Discussions › Notebook Guides › Unlocked Nvidia-Dell Video BIOS! Overclock your 17inch notebook Videocards!
New Posts  All Forums:Forum Nav:

Unlocked Nvidia-Dell Video BIOS! Overclock your 17inch notebook Videocards! - Page 17

post #321 of 1404
Quote:
Originally Posted by ZX81
Your wrong about clockspeed

Flash your card to higher clocks (not voltage or memory timings)
Run ATI tool and monitor max temp

Higher clocks, higher temp. Flash it high enough and your card will reach its throttle limit. The temp difference between stock clocks & overclock is notable (and so is the comparable 3dmark benchmarks)

I know on my 7800gtx I had to back down the clocks a bit. With higher clocks I wasnt getting artifacts on ATI tool test runs, but after a long time running FEAR the temp would creep up to the throttling limit and the card downshifts and lower FPS

These clocks I max @ 101degc which is in range for an 7800gtx and its been running quite happily there for over a year.

My card just gets to do a clock flash so I dont have the experience of messing with timings that 7900/7950 users have.
Well, like I said overclock can increase temperatures slightly. But, from my experience, it's never very much. I did test my card at stock clocks in ATI tool and then tested it OC'd and I saw a 4-5 degree increase in temperature. Is it using more power now because of this? Hell no.
post #322 of 1404
Well i was finally able to flash my GS to GTX. Not sure if i was doing something wrong. Now i have another problem. I have no experience with coolbits. Can someone walk me through on how to install and use it? And also is there a way to check the voltage of your GPU? I tried modding some of the bios to lower voltages and am unsure if it is working


Thanx
post #323 of 1404
Mmm dont want to be pedantic or start something because I really dont care for that, but your wrong again

There will be more power drawn at higher clocks, although the difference between whatever standard is eg 400/1000 (I cant remember) and 450/1100 wont be huge

Flipping to 2d gets much longer battery life, but I think the "table" in nibtor also has a lower voltage with the lower clocks (in this case 100/658)

The card wont even run at full clocks with the regular 90w psu.
You cant get something for nothing. Read through K6's old thread on as5 and trying to cool the overclocked cards; theres some interesting posts there

Edit>Have fun, life too short
post #324 of 1404
Dis anyone tried with success flashing an TOSHIBA P100 model PSPA6E (I believe it's an european model) which has one 7900GS??

Do you think is safe??
post #325 of 1404
Quote:
Originally Posted by psychoLEMONADE
Dis anyone tried with success flashing an TOSHIBA P100 model PSPA6E (I believe it's an european model) which has one 7900GS??

Do you think is safe??

I honestly would not risk it psycho. We are only using this bios because dell locked our cards but many manufacturers choose not to lock the cards so you should test to see if your card is locked before even thinking about doing something so bold.


On topic though i can defiently vouge that high clock speeds = higher temps

at base 375 / 500 @ 1v I top out at 82c

at base 450 / 700 @ 1v I top out at 89c

as base 600 / 800 @ 1.24v I top out at 94c this having throttle kick in and pwnin my graphics.

gonna try working in between voltages to ge ta sweet spot since my card tops out at 450/700 at base voltage.

Also Ive discovered a better method of stress testing the card. If any of you own the game AgeOfEmpires3 then your in luck. For those of you that dont then grab the demo! Basically set the options to max and stay in the main menu of the game. The games menu is drawn using your gfx cards shaders so having higher settings = prettier and thus stressful menus. within 2 minutes I can hit top temps on my card much quick than ATI tool or any other program can and it makes it VERY obvious when you have defects to do a starving card or a overheating one.
post #326 of 1404
That's true! It's unlocked... However, I've heard the GTX BIOS was more stable and I think I can hit higher clocks. The maximum I can get is 450/650 with the original GS BIOS without artifacts which gets 80ยบ at full load:
post #327 of 1404
Quote:
Originally Posted by Disillusion
What? Why not? Overclocking a video card doesn't add voltage.
If you don't change the voltage then this is correct. However, running faster will draw more current, thus increasing the power.
post #328 of 1404
Quote:
Originally Posted by Disillusion
But the whole argument that overclocking increases power usage is just completely untrue.

If that were the case, then why would the GPU have a 'low power 3D' mode and a '2D mode' that are underclocked to save power? Why would the GPU downclock on battery?
post #329 of 1404
Indeed, higher clocks do equal higher temperatures
Decreasing my core clock speed from 600 to 580 reduces temperatures by 2 degrees instantly.
Another game that should stress your card out fast would be the new Rainbow Six Vegas.
My card tops out at 92 deg. celcius on that even with 1.05v voltage...
Although the undervolt from 1.05v to 1.24v did yield somewhat cooler temperatures for me...I'm starting to see that lowering clockspeeds have an even greater effect on the temps.
post #330 of 1404
On the same note...has anyone here who's using 3 vid's (as shown by Krayziepop) have their vid stuck at 07 all the time? Even @ idle and during ATiTool, my VID never seems to change from 07 to 04 (which is during max load) or from 07 to 03 (idle state).
post #331 of 1404
Quote:
Originally Posted by ChrisLilley
If that were the case, then why would the GPU have a 'low power 3D' mode and a '2D mode' that are underclocked to save power? Why would the GPU downclock on battery?
I don't think you completely understand. Overclocking is just overclocking. You are increasing the clock speeds of a given component ONLY. You are not increasing voltage. When you run on battery, your GPU underclocks AND undervolts. They are two totally different things.
post #332 of 1404
Ach, your fixed on voltages. Its not worth arguing over this
Yep 2d does reduce both

I'm an engineer, for my many sins

chill out
post #333 of 1404
Quote:
Originally Posted by Disillusion
I don't think you completely understand.
I think I understand Power = Volts times Amps pretty well, thanks.
I also understand that if the heat emissions are up, that means its using more power. You know, conservation of energy and all that.

Quote:
Originally Posted by Disillusion
Overclocking is just overclocking. You are increasing the clock speeds of a given component ONLY. You are not increasing voltage. When you run on battery, your GPU underclocks AND undervolts. They are two totally different things.

Well said. Now, can you explain why it undervolts and underclocks? According to your alternative physics, just an undervolt would have the exact same effect, no?
post #334 of 1404
Quote:
Originally Posted by ChrisLilley
Well said. Now, can you explain why it undervolts and underclocks? According to your alternative physics, just an undervolt would have the exact same effect, no?
Because if you undervolt it without underclocking it, it will not be stable and you will get artifacts all over your screen. Have you ever overclocked a CPU? When your system becomes unstable, what do you do? Add more voltage. So if you take voltage away, you must underclock to keep the system stable.
post #335 of 1404
Quote:
Originally Posted by ChrisLilley
Now, can you explain why it undervolts and underclocks?

The powermizer of 7900gs (stock) underclocks only, all the modes (2D/thr/3D) operate at 1v.

So yes, underclocking alone actually saves battery.
post #336 of 1404
Quote:
Originally Posted by Disillusion
Because if you undervolt it without underclocking it, it will not be stable and you will get artifacts all over your screen.

If that were true, why would the 2D mode on a 2500M run at 100/100 @1v and the low-power 3D mode run at 200/300 also @ 1v ? On your theory, that means stock 2D mode is all full of artefacts?
post #337 of 1404
Quote:
Originally Posted by ChrisLilley
If that were true, why would the 2D mode on a 2500M run at 100/100 @1v and the low-power 3D mode run at 200/300 also @ 1v ? On your theory, that means stock 2D mode is all full of artefacts?

this is because 1v is enough for 200/300, if you want to go higher in clocks, more juice is needed
post #338 of 1404
Quote:
Originally Posted by kasaweb
this is because 1v is enough for 200/300, if you want to go higher in clocks, more juice is needed

I understand. I was replying to the poster who seemed to argue that if 1v at 200/300 was stable then 100/100 would need a lower voltage to be stable. Which is clearly not the case.
post #339 of 1404
Quote:
Originally Posted by ChrisLilley
I understand. I was replying to the poster who seemed to argue that if 1v at 200/300 was stable then 100/100 would need a lower voltage to be stable. Which is clearly not the case.
That isn't what I said. Higher clocks need more voltage to be stable. Lower clocks CAN run lower voltages and still be stable, therefore saving battery power and making the device run cooler. This is what I said, please read it again:
Quote:
Because if you undervolt it without underclocking it, it will not be stable and you will get artifacts all over your screen. Have you ever overclocked a CPU? When your system becomes unstable, what do you do? Add more voltage. So if you take voltage away, you must underclock to keep the system stable.
What I meant in that last sentence was that if you undervolt and your system becomes unstable, you have to lower the clocks to where it is stable. In this case, 1v is enough for 200/300, so obviously it's also going to be enough for 100/100 as well.
post #340 of 1404
Okay... Listen Disillusioned.

If you take a 7900GS, which is clocked at 375/507 stock and proceed to overclock the core to say.. 475. You are trying to tell me that a 100Mhz clock increase in the core will not increase power consumption? This is very wrong sir, because as you increase the core clock, the amount of wattage drawn increases with it. This has nothing to do with voltage at all. The voltage is the passageway through which amperage flows. Volts = passage and Amps = Volume of electrons moving through the passage.

If you use a utility called Notebook Hardware Control then you can easily prove or disprove your theory, so please take the time to do so. In NHC there is a convenient little section that shows a handy piece of information to us. Its called DISCHARGE RATE, and this will show you the wattage your computer is drawing from the battery while the AC current is offline (unplugged). Get the reading while it is unplugged and idling first. Once you know this value, overclock the core by like 100Mhz without adding voltage and then see what the new value is. I will guarantee you that it will show a higher draw while overclocked. Try it before you try to disprove it again...

Electronics need current and voltage both to do their job. To make a CPU faster, you don't always need voltage but by raising clockspeeds you are inherently increasing current draw. There is no way to avoid it really. By adding more voltage, you are also allowing the current to flow faster to its destination, thus allowing a higher current flow/volume by nature. It's really hard to disprove a scientific fact. If you wish to try, please show some kind of proof that you are right rather than just saying over and over again that the majority of us are wrong. The burden of proof lies in your hands buddy. Kthxbye!
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Notebook Guides
NotebookForums.com › Forums › General Notebook Discussions › Notebook Guides › Unlocked Nvidia-Dell Video BIOS! Overclock your 17inch notebook Videocards!