*** NVIDIA 470/480 Series Thread ***

Quote:

---

Originally Posted by p2501

Oooh well, it can't be helped then, ty. I'm just asking because my first card is running 850/1700/2100 @1138 while the second one only reaches 823 at the same volts, on air mind you. I thought I might get an idea if going water would fix that or if I've gotten a lemon.

---

My maximum overclock on air is 850/1700/2100 @1.138v, I can't push 20MHz more on Core, Shader and Memory without errors and artifacts! I've heard of someone in my clock range (852MHz on air) getting 900MHz Core at the same voltage switching to water cooling, read 3DGuru's 16 watts less power under water!

>> Right here.

Quote:

---

Well, the end result is that we are consuming LESS power on water-cooling. Two things are responsible for that, the colder GPU (better conducting) and thus that reference cooler.

Check it out:

Setup IDLE WATTS FULL WATTS
GeForce GTX 480 203 428
GTX 480 Liquid Cooled 201 412
GTX 480 Liquid Cooled 900c|1800s|4400m 215 506

Now granted it's not a lot. But yeah... we are using 16 Watts less all of a sudden. Overclocked obviously changes that and the card starts to consume much more current quickly, but that's also because we overclocked rather insanely and added the voltage tweak.

---

- Systemlord

Thanks for your input! :)

That sounds like my first card will propably make it to 900 on water, I just don't know about the second one. I'm thinking about sending it back and getting another one. Maybe MSI this time. :shrug:

When I got the second card there were visible marks on the PCIe connector, even though the card looked new it was definitely put in a system in the past. Also, one screw hole close to the edge has marks that look like copper corrosion. And the PCB is running ~10°C hotter than the first card. :shakes:

Quote:

---

Originally Posted by p2501

Thanks for your input! :)


That sounds like my first card will propably make it to 900 on water, I just don't know about the second one. I'm thinking about sending it back and getting another one. Maybe MSI this time. :shrug:

When I got the second card there were visible marks on the PCIe connector, even though the card looked new it was definitely put in a system in the past. Also, one screw hole close to the edge has marks that look like copper corrosion. And the PCB is running ~10°C hotter than the first card. :shakes:

---

You are welcome! :)

I would send that card back immediately, 10C difference in PCB makes me think of a bad mounting of the cooler, also PCB temps happen a little slower because the heat is absorbed through the 8 layers. A GPU's temp fluctuation is almost instant, that card was put together by a well train monkey! :rofl:.

- Systemlord

Quote:

---

Originally Posted by Systemlord

You are welcome! :)

I would send that card back immediately, 10C difference in PCB makes me think of a bad mounting of the cooler, also PCB temps happen a little slower because the heat is absorbed through the 8 layers. A GPU's temp fluctuation is almost instant, that card was put together by a well train monkey! :rofl:.

- Systemlord

---

Well, then I guess Mr. Monkey will be able to play a bit more with this particular card because I'm sending it back. :up:

Already ordered an MSI card, hopefully that one will be better. And hopefully the store I'm sending the card back to isn't slow on the money return.

Any BIOS newer then 70.00.21.00.02/82?

Quote:

---

Originally Posted by donmarkoni

Any BIOS newer then 70.00.21.00.02/82?

---

Not according to this site. Although, you could check your brand's supportsite.

Quote:

---

Originally Posted by tentp0le

Not according to this site. Although, you could check your brand's supportsite.

---

Thanks. Nothing on mvktech also.
I was hoping for some leaked beta BIOS, or something to test...

I was curious what higher CPU OC will give 3-sli.
Previous benching done with 4.43, this time I pushed to 4.67 and results were good.

All 480s @875, 1950mem, 1.088 volts.

45363 Vantage PhysX OFF

http://rapidshare.com/files/40144206...8_CPU_4.65.jpg

and run for fun, PhysX ON:

52318 Vantage PhysX ON

http://rapidshare.com/files/40144206...hysX_52318.jpg

Alright luke what overclocking tool do you use exclusively? Everyone seems to be getting higher overclocks on less voltage and I think it has to do with the PWM voltage tweak that keeps voltage droop from dropping, I'm not exactly sure which program performs this voltage tweak! Also what are your cards reporting temp wise for all three?

- Systemlord

Quote:

---

Originally Posted by Systemlord

Alright luke what overclocking tool do you use exclusively? Everyone seems to be getting higher overclocks on less voltage and I think it has to do with the PWM voltage tweak that keeps voltage droop from dropping, I'm not exactly sure which program performs this voltage tweak! Also what are your cards reporting temp wise for all three?

- Systemlord

---

NvidiaInspector only and obviously nibitor for bios cooking.

Highest through benches ~52C, that's with rig under the desk.
Much less with 24/7 volts (1.025) pulled out :D

Quote:

---

Originally Posted by luke997

NvidiaInspector only and obviously nibitor for bios cooking.

Highest through benches ~52C, that's with rig under the desk.
Much less with 24/7 volts (1.025) pulled out :D

---

How do you get the fan to stay at say 70%, it seems like NvidiaInspector will not apply the fan setting after reboot or after a shutdown...? Is there a setting other than the Vcore voltage that helps your overclock?

- Systemlord

Quote:

---

Originally Posted by Systemlord

How do you get the fan to stay at say 70%, it seems like NvidiaInspector will not apply the fan setting after reboot or after a shutdown...? Is there a setting other than the Vcore voltage that helps your overclock?

- Systemlord

---

I'm under water so I only used Inspector for benching sessions and for settings clock voltages just BIOS mod - no need for fans for me :)

Quote:

---

Originally Posted by luke997

I'm under water so I only used Inspector for benching sessions and for settings clock voltages just BIOS mod - no need for fans for me :)

---

It seems the newer version fixes the fan setting bug.

Jebus, MSI card arrived and there are also traces on the PCIe connector.... but let's see what she can do first.

Quote:

---

Originally Posted by p2501

Jebus, MSI card arrived and there are also traces on the PCIe connector.... but let's see what she can do first.

---

Traces of what?

Quote:

---

Originally Posted by Systemlord

Traces of what?

---

Traces that it was already installed in a system. Scratches from the contacts. It was also missing the protective film pieces. :shrug:

It's installed now and it seems to do the job. It has an 1050mV VID, which isn't the best but I think the first card is the same. PCB is also running 10° hotter so I guess I have to blame the fact that there's a soundcard in the slot directly above which impedes airflow to some extend. Core is running 6°C hotter even with a single slot bracket, so it'll receive the TIM change treatment soon, IC Diamond is waiting. :D

What do I need to get a GTX 470 reference card over 1.082v that afterburner maxes out on? BIOS and another afterburner? Using latest afterburner.

just mod the BIOS should be able to get 1.213V with MSI AB then

Quote:

---

Originally Posted by justin.kerr

just mod the BIOS should be able to get 1.213V with MSI AB then

---

Thanks will try.
Thanks to Donmarkoni for helping me out through PMs as well. :up:
Will report air clocks with max voltage as soon as its fixed up.
On 1.082v the max this PNY 470 will do it ~850/1050 3dmarks. :)

Quote:

---

Originally Posted by justin.kerr

just mod the BIOS should be able to get 1.213V with MSI AB then

---

Has there been an acceptable safe limit in voltages established for the 470/480 using a water cooling solution? We already know that 1.138v is safe since most programs support it, just trying to findout the limit for voltages regardless of how low the temperatures are under water cooling. Also is there a visual guide for us noobs to mod a BIOS safely, I have found the instructions a little lacking.

-Systemlord

Quote:

---

Originally Posted by Systemlord

Has there been an acceptable safe limit in voltages established for the 470/480 using a water cooling solution? We already know that 1.138v is safe since most programs support it, just trying to findout the limit for voltages regardless of how low the temperatures are under water cooling. :)

-Systemlord

---

I know my GTX 480 scales great up to 1.1v.... on water 1.3v seems fine and get a little better clock but not much. More than that might not be of any benefit so try testing different settings up to 1.3v to start and see what gains can be had with each setting. :up:

my GTX 470 on air will do Vantage at 872Mhz, with 1.13V's but adding more voltage, just gets hotter, and crashes, so the added voltage does not help me on air, but I bet it would on water. :D or outside in the winter air.

Quote:

---

Originally Posted by justin.kerr

my GTX 470 on air will do Vantage at 872Mhz, with 1.13V's but adding more voltage, just gets hotter, and crashes, so the added voltage does not help me on air, but I bet it would on water. :D or outside in the winter air.

---

Wow 872Mhz is an amazing overclock for any 470, I can't even get past 852MHz on 1.138 volts! When your chip starts to get hotter in a way that rises quickly you are going beyond the limit of that chip, kind of like a car going into the redline in RPMs! Also when you see little increases in MHz and you are dumping voltages into it, better start backing off!

- Systemlord

Here's my GTX 480 result on air... Runs good with 1.1v but very hot. :D
http://www.xtremesystems.org/forums/...1&d=1274242473

Quote:

---

Originally Posted by Linuxfan

Here's my GTX 480 result on air... Runs good with 1.1v but very hot. :D
http://www.xtremesystems.org/forums/...1&d=1274242473

---

Water cool that beast if you want long life out of your card!

- Systemlord