Question Replacement for inno3d iChill 1080ti with waterblock.

Boraes

Honorable
Apr 24, 2014
28
0
10,530
0
Hiya guys and girls! I was hoping for some advice to help me out in my dilemma.

I had the above card for 2 years and it was awesome. A few months ago I upgraded to a custom loop and added a waterblock to the GPU and reaped the benefits as it really made a noticeable difference.

I’ve used MSI Afterburner off and on and thought to try the OC Scanner out. Very bad idea as whilst out smoking leaving it doing its thing it had fun killing my GPU + one of my RAM modules (2 weeks outside of GPU warranty DAMMNIT).

So I’ve got £600 available right now and a few choices I’m considering and would love to hear some opinions on the below.

  • 1080ti then mod and install my waterblock
  • 1080ti with AIO
  • Go for a GTX 20** gen GPU (if so which one?)
I’d really appreciate people weighting in
 

Boraes

Honorable
Apr 24, 2014
28
0
10,530
0
How have you determined that your GPU + RAM have been affected? OC Scanner shouldn't do anything irreparable.....

A fair question. Ok, well firstly the classic ‘burnt’ smell wafting strongly from inside the case and my PC wouldn’t wouldn’t power up upon my return from a smoke; this is using either the case switch or the button on the board.

Performed a couple of power washes and got it to power up. No sign of life from gpu and no output from any of the ports and rapid continuous beeping from the chassis speaker.

Switched GPU power leads - nada. Same result as before. Next I unplugged GPU and attached to board VGA. Got a sign of signal but still the beeping.

After testing each of my dimms in each slot I identified the offending knave and posted into the bios with the message ‘.... recovered from unstable overclocking settings’ Chose ‘load defaults and restart’. When back in bios no GPU detected.

Got into the GUI (GPU installed to board at this time with one monitor hooked to that and the board VGA atm still only life from VGA)
No gpu detected in device manager. Uninstalled everything Nvidia related (even down to leftover registry items) and restarted.
Got into the GUI (GPU not installed to board at this time. Uninstalled intel adapter before restarting, clearing bios and powering up with GPU installed. Nothing.
When trying to run the manual install of the nVidia drivers it failed as no Nvidia product was detected.

Will of course run a few more tests once I have the replacement products; as you can appreciate constantly draining and dismantling a custom loop isn’t really fun on the 10th time so waiting till then. However I do have some isolation fittings on order to make it easier in future (Fingers crossed I only need to use them for maintenance and upgrading).

Surely though the MSI afterburner OC scanner could cause damage as it plays about with all the settings of the GPU power/voltage and speeds?

Believe me my life is filled with dark ironic slapstick such as this; A 3D printer missing a capacitor blew my cpu, mobo and aio pump beginning of May, hence why I went for a custom loop instead which has (Had been 😢) only fully up and running about a month. Thank God for alcohol 😊
 

Phaaze88

Honorable
Herald
Unless you hacked your gpu's vbios, OC Scanner can't kill your gpu. The generous power/temp/voltage limits placed on these cards won't let you.

I've played around with OC Scanner, and found it to not be very good. It's more aggressive on the core OCs that it needs to be - Pascal and Turing don't care for it. +100 core is about all one should aim for.
Each time I ran the program, it never told me anything beyond '90% confidence'. I'd run Test, and it'd pass, then go play a few games... it would eventually crash; sometimes taking a day or two.
For me, it was pushing core clocks of 130, 140, and even 164! None of those proved to be stable over time.

The memory is what should be tweaked, yet OC Scanner doesn't touch this at all, and neither does Gpu Boost.

Are you sure the waterblock you added was compatible with your gpu? Did you mod anything?
Because it sounds like a vrm or memory chip wasn't making contact. Have you taken a look under the WB to check where the damage is located?
 

Boraes

Honorable
Apr 24, 2014
28
0
10,530
0
Unless you hacked your gpu's vbios, OC Scanner can't kill your gpu. The generous power/temp/voltage limits placed on these cards won't let you.

I've played around with OC Scanner, and found it to not be very good. It's more aggressive on the core OCs that it needs to be - Pascal and Turing don't care for it. +100 core is about all one should aim for.
Each time I ran the program, it never told me anything beyond '90% confidence'. I'd run Test, and it'd pass, then go play a few games... it would eventually crash; sometimes taking a day or two.
For me, it was pushing core clocks of 130, 140, and even 164! None of those proved to be stable over time.

The memory is what should be tweaked, yet OC Scanner doesn't touch this at all, and neither does Gpu Boost.

Are you sure the waterblock you added was compatible with your gpu? Did you mod anything?
Because it sounds like a vrm or memory chip wasn't making contact. Have you taken a look under the WB to check where the damage is located?
That’s what I thought. I was under the impression the whole point of the OC Scanner, though maybe not the best option would at least be safe. Rather ironic that I could manually OC it fine - that’ll teach me to be lazy 😆

The waterblock is the correct one (barrow RGB Titan/1080ti/70 block checked with their ‘compatibly checker before purchase), and other than installing the waterblock on it no other modding to the GPU.

I admit I haven’t remounted the waterblock on the card, mainly due to running out of patience after spending most nights testing. I’m pretty certain of where any damage to the card will be thanks to the smell. I will attempt a remount once everything has arrived over the next 4/5 days. Coincidentally I happened to have next week off work already a couple of months ago so will better placed to do it then rather than piecemealing over a couple of nights.

I would of thought that even if the cooling wasn’t up to scratch surely the GPU would still initially power on, but just heat up damn fast?

N.B - I really appreciate the advice.
 

Boraes

Honorable
Apr 24, 2014
28
0
10,530
0
Ok I've been looking around and doing the math - I'm 99% sure my best option given the reviews, specs and prices I'm going to get

GIGABYTE GeForce RTX 2080 XTREME WATERFORCE WB 8GB GDDR6 Graphics Card (GV-N2080AORUSX-WB-8GC)

Block already installed on this and already OC'd, just reassures me that I've not caused any issue; as well as some time - plus it comes with a 4yr warranty. I'm able to get my current waterblock refunded I've found it for £722, so if it's £500 ish for a good 1080ti, GPU WB cost me £90 and I can make up the difference
 

ASK THE COMMUNITY

TRENDING THREADS