Undervolting a 1660Ti

[BoomAM]

Hi,
I just got a 1660Ti for a secondary PC, sat in a Phantec's Evolv Shift.
Doesn't run too hot, but i want to make it a little cooler to reduce fan noise, hence, undervolting.

I'm trying to follow some guides that are posted on LTT or Reddit in regards to the process, but it makes no sense whatsoever.
Using MSI Afterburner, i can adjust the pretty little graph that shows, BUT most of the guides list dragging the line DOWN to adjust the voltage. The up/down axis is frequency however?

Am i missing something obvious here?

Thanks in advance!

 

[BoomAM]

Vertical is frequency, horizontal is voltage, just make the curve stop at the voltage/frequency you want then flatline it. Then test if it's stable, adjust as needed.

Yes, thats what i said.
All the videos/guides online show people adjust the frequency to lower the voltage!?

Interestingly, i'm playing around with increasing core frequency on the main MSI menu, whilst limiting the max power.
Running a bench in Heaven benchmark and recording max temps/score.
Stock = 2245 points, max temp of 75.
+100Mhz core, -10% power = 2285, max temp of 72
+150Mhz core, -20% power, = 2319, max temp of 71

Makes no sense at all other than 'because nVidia boost', but ok, i'll take it.

 

[BoomAM]

It's because frequency & voltage are interlinked.

In what way? As the results im seeing seem to be a bit like a 'yo-yo'.

Just replaced the stock Zotac paste with some MX4. Lets see if that also helps.

 

[BoomAM]

In the sense of frequency being linked with voltage. So if you wanted a higher voltage, let's say, for a particular frequency, then you couldn't do that. There's a particular voltage & frequency pairing and you can't modify them, you can only select them. So if you want to choose a particular voltage, you can (for the most part), do that. But you can't then also choose which frequency to go with that voltage. And vice-versa. So you can choose voltage, or you can choose frequency, but not both at the same time. Whichever you choose, the second is given to you.

ok, so how does that help me undervolt? What is the correlation and how would this apply in practice within MSI afterburner?

 

[BoomAM]

...then you'd just drag every point after 975mV down to the frequency linked with that voltage...

That's kinda the reason I've posted to ask this question.
Dragging down is the frequency, not the voltage. You and everyone else are seemingly referring to the wrong axis. This makes no sense to 'drag everything after 975mV down'.
So either its badly explained as a whole and the 'community' around this aspect has just taken a bad description as the accepted explanation, or theres some strange axis swapping going on with my system here.
(Note, not having a go at you. )

I guess you didnt care about the warranty ?

Not really.
Plus, there were no 'void if broken' stickers to be seen

Wont reducing power reduce performance as well?

You'd think, but from the sketchy assumptions im seeing online regarding the nVidia boost tech on these, sometimes it actually helps. Case in point, what i'm seeing.

I'm seeing some good results with MSI afterburner set to a max power rating of 80%, +150 on the core.
After letting heaven benchmark run for about 15-20mins in a 24*c room (75*f) the max temp seen is about 68*c.
Gonna run it for a bit longer, but if this method appears to work (higher perf, lower temp) then the faffing with the graph could ultimately be pointless.

 

[BoomAM]

Dont use heaven for testing overclocks etc i hear its no good for that as you need something more precise or something like firestrike.

Noted, but im not testing for out and out performance, just to check performance regressions...

Does anyone know how the 'set clocks on startup' function in MSI afterburner works?
Do i need to manually open the app on each windows boot? As i cant see a service running to set it on Windows boot...

 

[BoomAM]

Yes vertical is frequency, all your telling the card is don't boost any higher than X frequency at Y voltage, so every boost data point past the one you want is irrelevant.

Im just gonna give up on this option.
Whilst i understand the concept, the process of doing it makes no common sense for how a program is laid out to achieve it.

Doesnt matter Zotac know when its been removed from experience. They know you've undone the screws.

I'll take the risk.

Tick the startup option in settings so it starts with Windows.

I know this.
I was asking how it set it. Usually things like this set with an app or service on boot. Neither of these are present on the system.

show the results here then we can see if they have made any difference.

...above, Post #3.

 

[BoomAM]

Actually you are incorrect usually startup actions are set in the registry startup key or the windows startup folder.

Thats not accurate - in the case of physical hardware related settings, usually these are set via an app that accesses the hardware's address bus in-lieu.
This is because Windows lacks the correct components/APIs to be able to talk to custom hardware buses, a registry entry would not allow that as its just a variable for an application to store/check a setting when its running.
If it has been set in the registry, i would have guessed that its just a call function for the service/app that runs in the background that does know how to access the hardware directly, with said key telling the app what to do.
The only way this is achieved without an accompanying app, is if the nVidia driver is providing that link to the hardware and recognizes the registry setting as something it needs to set on its own to the physical hardware on boot.
....which is what i was asking, how its achieved at software level.

You were advised to use a different benchmark other than Heaven hence post the results...

Whilst i see what you are getting at here, the performance regression testing was duplicatable each time. Using a more stressful test would not have improved that outcome.
By most testing methodologies, that's conclusive evidence.

It just shows the 1660ti you have is a good overclocker if you can reduce power by 20% and +150 core

Yes, that does appear to be one of the conclusive results. Which is surprising considering how new the 1660Ti silicon is to the market. I either hit the silicon lottery or the yields are efficient already.
IIRC, boost clocks were sustaining above 2Ghz, which is noticeably above 'stock' boost clocks.

It's just a case of dragging the dots up and down....

I've not read all of your post yet, i will give it a good read when i get in tonight, but to pre-post as it were, thank you for providing, it looks useful.
I will read & test when i get in.

 

[BoomAM]

Still not read HeX's post, but to add to this, my other half (who owns this 1660Ti) has decided she wants to WC the thing. So we're going out to get a G12/an AIO that'll fit in her SFF. So this whole thread could be practically pointless, even if productive in theory.

 

[BoomAM]

Water cooling a 1660Ti, an already very cool running card that'll hit the limits of the silicon on air, is madness!

Tbh, its the case thats the issue.
Phantec Evolve Shift (non-X) - its airflow starved, so its a miracle i can get it running at 70*c with that small tweak, but the fans hit 70-75% and is loud.
We're literally gonna replace the H80i v2 for the CPU, with 2x H55's instead as they should both fit in. With the aim of sacrificing some CPU *C for an overall quieter system.

 

[BoomAM]

Turns out the local store didn't have the H55 in stock....so before we order online, i played around a little more...

• 70% power limit.
• CPU fan locked at 50%
• CPU core boost of +175Mhz

FurMark ran for 30mins

• Max temp seen: 73*c
• Avg temp seen: 70*c
• Max core speed seen: 1945 Mhz
• Avg core speed seen: 1775 Mhz

Tbh, i think that's pretty good considering FurMark is not indicative of average gaming workloads, and average and max are comfortably above boost clocks.
Other observations, a locked fan speed of 40% shows the highest temp to be 77*c.

I've now set a custom fan curve that locks then fan at 40% upto 70*c, then ramps up to 50% at 75*c and 100% at 80*c.

That should tide things over for a while, whilst t'other half starts getting into gaming proper. If it ends up being an annoyance, we always have the H55 option (and i'll take the 'old' CPU AIO and whack it on my 1080Ti )

 

[BoomAM]

It's just a case of dragging the dots up and down.

Just revisiting this.
Is this just a case of your 'first' dot, adjusted as the max clock speed at a given voltage on the horizontal axis, becoming the new max voltage? Assuming that the following 'dots' in the graph are on the same level on the vertical axis?

So if i set for example, 1,000Mhz to be 0.5v, and everything after that is set to be at the same horizontal 'level' as that, then the max voltage would be 0.5v at any given speed above 1,000Mhz? With the speed maxing out at the best it can achieve at that voltage?

If so, then that is understandable, but a completely messed up way for MSI Afterburner to represent that.

 

[BoomAM]

So i tried dropping the voltages at 2Ghz down to 0.950 and 0.900, in both scenarios i got more heat output?
(done by finding the voltage, moving the dot up to the frequency, then flattening everything after that to the same level).

Tried combining adjusting the voltages charts with the max power limit and instead saw a 200-300Mhz drop in clock speed, and still, more heat.
Went back to my original 'basic' undervolt (40% locked fan speed, +175Mhz core, 70% power limit) and i got back to good temps again and higher/more consistent 1,950-2,050Mhz clocks again.

No idea what is going on there! Especially considering the voltages with the basic undervolt were a little higher.

 

[BoomAM]

Is there a trick to flatten curves in Afterburner? Like a shortcut?

 

[BoomAM]

That's a shame, as its as PITA to do manually.

Anyway latest result seems good:
0.800v max voltage
1,800Mhz frequency base
Custom stepped fan curve.

Had 'other half play the GPU killer itself, the title that puts Furmark, Heaven, SuperPosition & Vally to shame when it comes to to heating the 1660Ti up....Snake Pass.

Max temp we saw was 65*c, and the fans mainly ran at 45%, which from 7-8ft away on the couch isnt really noticeable. This was after 30-40mins of playing.

I'm reasonably happy with that...

 

[BoomAM]

Doesn't have to be perfectly flat, as long as the values are the same/lower than the undervolt you want you're fine.

its not the overall flatness that's the issue, but when every point past 0.800v is higher, i have to adjust each one.

With the results i'm getting out of tuning the 1660Ti, i'm half tempted to give the same undervolting a go with my 1080Ti.