Andrew_WOT
ML120 should dim at lower voltage, do you observe that.

Andrew_WOT

petmic10

Andrew_WOT
ML120 should dim at lower voltage, do you observe that.

I notice no dimming with my fans.

With voltage regulated fan speed? That can only mean one thing, the voltage is constant.

smackey
i just got my 1080 Ti FTW3 hybrid today. I'm also changing the stock fan into corsair HD 120 but i noticed that it's using a 3 pin fan compared to my corsair fan which is 4 pin. How do i connect it to the gpu? is there an adapter? Right now the fan is connected to commander pro which controls the fan speed. what do i lose for not connecting the fan straight to the gpu? I'm also on the slave bios right off the bat and i want to overclock on it's max potential. What is the typical boost clock range for 1080 ti FTW3 hybrid? 

petmic10

smackey
this is what i ordered 
I'm only getting+50 right now without even touching the memory and the boost clock is fluctuating between 2025-2037 depending on the temps.
It's already in the slave bios with 127 power target and voltage is all the way up. do i also need to update the gpu bios? what i am doing wrong? or do i have faulty gpu?

It doesn't sound faulty. Remember these gpu's are advertised at 1683 boost clock and you are getting over 2000. Are you stress testing or just gaming? Try adding +75 and run a stress test like Unigine Superposition. Keep adding a little at a time until it crashes. When it crashes back it down 10 or 15 MHz and run the test again. Once you get your gpu clock set move to memory but you can probably do memory in +100 MHz increments. 
 
Superposition is free if you bought a gpu from EVGA after March 1st.
 

 

smackey
this is what i ordered https://www.amazon.com/gp..._s00?ie=UTF8&psc=1
I'm only getting+50 right now without even touching the memory and the boost clock is fluctuating between 2025-2037 depending on the temps.
It's already in the slave bios with 127 power target and voltage is all the way up. do i also need to update the gpu bios? what i am doing wrong? or do i have faulty gpu?

SecretTI
 
Hey Smackey, I'm using the exact same setup as you. Corsair CPU cooler, same 570x case, now the same GPU, but I'm sticking to the stock SP fans. A recommendation, I suggest you not use the fan power from the GPU, but to get a Corsair Commander Pro to control all of them, Corsair Link is a beautiful tool. Also, I noticed you curled your tubes underneath to the right. I curled them to the right from the top, but still sticking the tubes to the bottom of the rad configuration like yours. I'm unsure if either of those ways are causing stress on the tubes. I'd hate to have liquid spill all over.
 
To the rest of the community, I do want to bring up some concerns and ideas. I've got my card today, so things may be different now.
 
I've hard wired checked the voltage reading on the radiator fan powered by the GPU. voltmeter remains constant at load and idle. I will be switching it out for a different fan (consistent with my case, my spare SP120s). This will be connected to my Corsair fan controller, Commander Pro. I know that it sucks that this is one of the most expensive Ti's out there and EVGA failed to pulsate the power on it or allowed for a 4 pin. But this shouldn't be a reason for a return, just fan control it elsewhere, otherwise you're missing out on a decently cooled card.
 
Next, I wanted to mention that XOC isn't really updated for this card. For me at least, only one fan can be controlled from it, and its not memory. It's the power fan option out of the three on the right. The power fan controls the blower for me. 
 
The other concern I wanted to bring up was that the LEDs for the GPU temperature and the memory temperature are switched. If GPU on xoc is green, the M on the GPU turns green, not the G. XOC needs to be reprogram the LEDs for this.
 
Also, analyzing my ICX sensors, I noticed a extremely variable temperatures. My memory 2 sensor is often significantly more than the other memory sensors under stress. Same case for my PWR 4 sensor.
 
I'm a big fan of rainmeter, I hope there can be a plugin to retrieve the ICX temperature data instead of just the main GPU one. 
 
 
Some OC stats:
 
setup
pwr target: 115%
Temp target: 75 degrees
voltage: +75%
 
stable results
GPU clock: +75 basic curve increase ~2050 mhz
Mem offset: +200 (bit lazy on this one)

SecretTI

smackey

Hey Smackey, I'm using the exact same setup as you. Corsair CPU cooler, same 570x case, now the same GPU, but I'm sticking to the stock SP fans. A recommendation, I suggest you not use the fan power from the GPU, but to get a Corsair Commander Pro to control all of them, Corsair Link is a beautiful tool. Also, I noticed you curled your tubes underneath to the right. I curled them to the right from the top, but still sticking the tubes to the bottom of the rad configuration like yours. I'm unsure if either of those ways are causing stress on the tubes. I'd hate to have liquid spill all over.
 
To the rest of the community, I do want to bring up some concerns and ideas. I've got my card today, so things may be different now.
 
I've hard wired checked the voltage reading on the radiator fan powered by the GPU. voltmeter remains constant at load and idle. I will be switching it out for a different fan (consistent with my case, my spare SP120s). This will be connected to my Corsair fan controller, Commander Pro. I know that it sucks that this is one of the most expensive Ti's out there and EVGA failed to pulsate the power on it or allowed for a 4 pin. But this shouldn't be a reason for a return, just fan control it elsewhere, otherwise you're missing out on a decently cooled card.
 
Next, I wanted to mention that XOC isn't really updated for this card. For me at least, only one fan can be controlled from it, and its not memory. It's the power fan option out of the three on the right. The power fan controls the blower for me. 
 
The other concern I wanted to bring up was that the LEDs for the GPU temperature and the memory temperature are switched. If GPU on xoc is green, the M on the GPU turns green, not the G. XOC needs to be reprogram the LEDs for this.
 
Also, analyzing my ICX sensors, I noticed a extremely variable temperatures. My memory 2 sensor is often significantly more than the other memory sensors under stress. Same case for my PWR 4 sensor.
 
I'm a big fan of rainmeter, I hope there can be a plugin to retrieve the ICX temperature data instead of just the main GPU one. 
 
 
Some OC stats:
 
setup
pwr target: 115%
Temp target: 75 degrees
voltage: +75%
 
stable results
GPU clock: +75 basic curve increase ~2050 mhz
Mem offset: +200 (bit lazy on this one)

timothy613
cant wait to get this together 
 

AdamInk
Try to switch to slave bios, it has 127% power target :)

AdamInk

petmic10

AdamInk
Try to switch to slave bios, it has 127% power target :)

The 127% power target doesn't seem to be working as advertised. The gpu hits the power limit way before 127% and needs to be fixed by EVGA.
 
https://forums.evga.com/1...over-117-m2693145.aspx

Yeah, i red this post sadly. Not sure its evga's fault, nvidia has his/her hands in it i think for the limitation. Gpu boost 3.0 is still there which is very strange for me to.

EVGA Double BIOS

A secondary BIOS allows you to experiment with custom BIOSes without risk of breaking the card, and allows easy BIOS recovery with just the flip of a switch. Also, on select cards the secondary BIOS may unlock additional overclocking potential and/or improved fanspeed control!
 

smackey
Hi secretti, im also using commander pro to control my fans. The hd 120 is not compatible with the fan connector on the ftw3 hybrid as it is 4 pin. The max rpm that i can set the hd120 using commander pro is around 1,800+ rpm. Im about to buy a 4 pin to 3 pin connector and connect the hd 120 directly back to the gpu. Maybe it will increase the rpm of the fans more and make me overclock better. Can you show me the pic on how you mounted the radiator and the tubes on your 570x?

Attached Image(s)