• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Sapphire Nitro+ 9070 XT Steel Nomad OC Setting Performance Effects

Wow @Kirkatron good work, that must've taken you ages!

The lazy bit of me likes it as it does suggest there's not much point in me fiddling with mine, just leave it at stock. Saves a lot of time and effort (thanks again) tweaking this and tweaking that for what will probably be less than a 9% increase (given I've likely not won the silicone lottery and it's a lower tier card than the Nitro+).

I will say that Nvidia should've used Steel Nomad in their slides and stuff when they released the 5090 as it at least doubles the 4080 SUPER and everything below it.
 
GoogalyMoogaly said:
Wow @Kirkatron good work, that must've taken you ages!

The lazy bit of me likes it as it does suggest there's not much point in me fiddling with mine, just leave it at stock. Saves a lot of time and effort (thanks again) tweaking this and tweaking that for what will probably be less than a 9% increase (given I've likely not won the silicone lottery and it's a lower tier card than the Nitro+).

I will say that Nvidia should've used Steel Nomad in their slides and stuff when they released the 5090 as it at least doubles the 4080 SUPER and everything below it.
Click to expand...

Double doesn't look as good as triple....
 
Last edited:
GoogalyMoogaly said:
Wow @Kirkatron good work, that must've taken you ages!

The lazy bit of me likes it as it does suggest there's not much point in me fiddling with mine, just leave it at stock. Saves a lot of time and effort (thanks again) tweaking this and tweaking that for what will probably be less than a 9% increase (given I've likely not won the silicone lottery and it's a lower tier card than the Nitro+).

I will say that Nvidia should've used Steel Nomad in their slides and stuff when they released the 5090 as it at least doubles the 4080 SUPER and everything below it.
Click to expand...
May still be worth trying. You'd be surprised how well the lower tier undervolt.
I managed to get 79fps by running -95mV with same settings as -90mV with 50% fan but that's the limit of my card's stability.
 
Last edited:
humbug said:
Double doesn't look as good as triple....
Click to expand...
I mean it's triple the 4070 SUPER and everything below it...

Kirkatron said:
May still be worth trying. You'd be surprised how well the lower tier undervolt.
I managed to get 79fps by running -95mV with same settings as -90mV with 50% fan but that's the limit of my card's stability.
Click to expand...
I might give it a go, although I don't have much patience for such things. Is 3DMark any good for testing stability in games? I might buy it if it is, but if you have to test games then it'll take too long and get too annoying when it crashes.
 
GoogalyMoogaly said:
I mean it's triple the 4070 SUPER and everything below it...


I might give it a go, although I don't have much patience for such things. Is 3DMark any good for testing stability in games? I might buy it if it is, but if you have to test games then it'll take too long and get too annoying when it crashes.
Click to expand...
Unfortunately not in my experience.
 
BS Dave said:
Benching is fun I don't care what anyone says. Getting that extra 10%, finding out what your card can do. Love it
Click to expand...
I'm sure it can be fun for some people, not for me though.

eeii said:
Unfortunately not in my experience.
Click to expand...
That's a pity, although I guess at the end of the day it is a benchmark not some stability tester so it's fair enough.
 
GoogalyMoogaly said:
I mean it's triple the 4070 SUPER and everything below it...


I might give it a go, although I don't have much patience for such things. Is 3DMark any good for testing stability in games? I might buy it if it is, but if you have to test games then it'll take too long and get too annoying when it crashes.
Click to expand...
No, those settings I used are stable in all the games im currently playing though.
I can go down to -175mV in some of the 3dmark tests.
You do have to test in games. I find it fun getting every last drop of performance out of the card. I'm hoping they do a voltage/frequency tool so I can map out a perfectly stable voltage/frequency graph.
It's certainly not for everyone, I respect that.
 
Kirkatron said:
No, those settings I used are stable in all the games im currently playing though.
I can go down to -175mV in some of the 3dmark tests.
You do have to test in games. I find it fun getting every last drop of performance out of the card. I'm hoping they do a voltage/frequency tool so I can map out a perfectly stable voltage/frequency graph.
It's certainly not for everyone, I respect that.
Click to expand...
What settings are you using for games?
 
Good work Kirkatron.

For reference my 9070 XT Pulse goes from 69 FPS stock to 76 FPS in steel nomad. So the extra power consumption of the Nitro at both stock and overclocked gives ~3% headroom.

I get the joy people get from seeing how much they can overclock and the edge of the cards ability. I also find it fun to find how efficient I can make the card run without losing any performance.

For me after a few weeks of use now, -45mV, no core offset, - 18% power limit, VRAM 2700 fast timing. It sits about 247w on average for similar to stock performance.

If you want your new profile to apply on every start up, disable CSM and fast boot in BIOS. Also make sure your PCI slot is not affected by any Power Saving settings.

Or you could set each game you play to use a specific performance profile.
 
Last edited:
I've not gone for overclock on Pulse, since it didn't seem to provide much benefit in Steel Nomad.

I've got it at -20% power limit and -75mv, sits at 144fps on the games I've tested it on, Le Mans Ultimate since it's main game now has no issues. Could probably go lower on mv for it.
Though my GPU clocks lower in Steel Nomad than in games, Steel Nomad, I was lucky to see 2800mhz even with +10% power and -90mv, on my game settings it'll hit just over 3000mhz

Haven't sat and played with it loads like OP though, it's drawing less power and doing better than out the box, so I'm happy.
 
Chopasaurus said:
I've not gone for overclock on Pulse, since it didn't seem to provide much benefit in Steel Nomad.

I've got it at -20% power limit and -75mv, sits at 144fps on the games I've tested it on, Le Mans Ultimate since it's main game now has no issues. Could probably go lower on mv for it.
Though my GPU clocks lower in Steel Nomad than in games, Steel Nomad, I was lucky to see 2800mhz even with +10% power and -90mv, on my game settings it'll hit just over 3000mhz

Haven't sat and played with it loads like OP though, it's drawing less power and doing better than out the box, so I'm happy.
Click to expand...

You will find that anyway, the way i have my 7800 XT setup it runs about 2900Mhz in Death Stranding but 2600Mhz is Star Citizen, stock its 2500Mhz vs 2200Mhz, different games put a different load on the GPU and it responds accordingly, the "game clock is 2124Mhz with a boost clock of 2430Mhz" so it grantees 2123Mhz in games (slightly higher than that in Star Citizen in my case) and a boost of 2430Mhz, this is "up to" as in Death Stranding.
What its doing is giving you a minimum guaranteed clock speed but will always try for higher where it can.

Good job on the undervolting.

The lowest i managed to get mind was 170 Watts at the same performance, this from an out of the box of 245 watts, but this turned out not to be 100% stable, i can get it to 190 watts with a less aggressive profile and that is stable, i tend to run mine about 15 watts higher, so 260 watts but for a 15% performance increase. Effectively a 4070 Ti.
 
Last edited:
michty_me said:
This is a grea thread. Thanks for all the hard work you put in.
Does CPU have an adverse effect on final scores on Steel Nomad?
Click to expand...
I do not think it does compared to Timespy. It's a much more level playing field, but it will still have some kind of impact. Have a look at 3DMark result browser and filter by your CPU and GPU to get an idea of where others with similar systems perform.
 
Kirkatron said:
The last -90mV entry in the table but without the negative core offset.
So -90mV, 2760 VRAM, 0 offset, +10% power. It's stable in everything I play at 4k.
Click to expand...
What does your card pull at default. I’m just curious. My xfx Mercury 9070xt pulls 340watts. So far 2760 is spot on for me aswell for the vram.
 
@Kirkatron Thanks for all the legwork here. I used your top settings on my Nitro and got similar results, fully stable. I've moved over to a 5090 now, but good lord the 9070xt is a performer at stock, a beast when tweaked. 7800+ in Steel Nomad is nearly butting up to a 5080, for goodness.
 
I added the 7952 score I got when using almost max ambient air cooling, 90% GPU fan.
I wouldn't be able to use this in game as it's way too loud.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom