*Official Benchmark Thread* - Post it here or it didn't happen :D

[jaybee83]

5 minutes ago, johnksss said:

Can you show me your stock runs for PR?

sure thing, can post later once im free from work.

[Mr. Fox]

The reason shunting works is because it deceives the GPU/firmware into believing it is pulling fewer watts than it actually is. Under full tilt load the GPU is pulling up to double what you see in GPU-Z or HWiNFO and I think you will find the WireView will confirm that. Voltage will also stay higher with the GPU pulling up twice the watts that it thinks it is using, although the voltage cap will not increase. It won't try to moderate itself because it has no idea what you're doing to it and won't attempt to decrease the voltage. Your software monitoring tools will become totally worthless with respect to GPU watts and show bogus information, but the core voltage reading should be as accurate as it was (or wasn't) prior to shunting.

[johnksss]

37 minutes ago, Mr. Fox said:

The reason shunting works is because it deceives the GPU/firmware into believing it is pulling fewer watts than it actually is. Under full tilt load the GPU is pulling up to double what you see in GPU-Z or HWiNFO and I think you will find the WireView will confirm that. Voltage will also stay higher with the GPU pulling up twice the watts that it thinks it is using, although the voltage cap will not increase. It won't try to moderate itself because it has no idea what you're doing to it and won't attempt to decrease the voltage. Your software monitoring tools will become totally worthless with respect to GPU watts and show bogus information, but the core voltage reading should be as accurate as it was (or wasn't) prior to shunting.

A guy made a shunt calculator along with the settings to change in gpz and hwinfo to get the correct watts. The problem I'm having is I used 1.5 which it does not calculate. 🤦‍♂️

 

Edit

[jaybee83]

14 minutes ago, johnksss said:

A guy made a shunt calculator along with the settings to change in gpz and hwinfo to get the correct watts. The problem I'm having is I used 1.5 which it does not calculate. 🤦‍♂️

 

Edit

 

now thats a pretty nifty tool 🙂 

[jaybee83]

5 hours ago, jaybee83 said:

sure thing, can post later once im free from work.

 

here you go:

 

This is with the PNY Epic-X vBios flashed with stock clocks, no voltage added. At stock and 32C, the V/F curve for the PNY is 45 Mhz higher than that of the Suprim.

This is just regular OS install that im running day to day, so no extensive tuning. Win11 24H2 Build 3476

RTX 5090 Suprim SOC (PNY Epic-X vBios stock).3dmark-result

[win32asmguy]

On 4/30/2025 at 1:41 PM, electrosoft said:

Ran some raids in my fixed D2D OC profile with GPU Tweak III and just smooth.....even when fps dipped into the 70s still smooth.....The problem with the WoW engine will always be when all that player data hits in certain moments, your 5090 will perform like a 4070 Super or 9070xt. Same with the 5080FE. There's just nothing you can do about it. The engine CPU physics need to be overhauled badly.

 

I am tempted to try and move my 5090 into my Intel 12900ks + z690 system this weekend and see if it does better with wow than the HEDT setup. I could even potentially try a 14900k or ks if it's supported via a bios update. That 12900ks seemed like it was pretty leaky in the raptor x15 but it might do perfectly fine even with good air cooling.

[johnksss]

2 hours ago, jaybee83 said:

 

here you go:

 

This is with the PNY Epic-X vBios flashed with stock clocks, no voltage added. At stock and 32C, the V/F curve for the PNY is 45 Mhz higher than that of the Suprim.

This is just regular OS install that im running day to day, so no extensive tuning. Win11 24H2 Build 3476

RTX 5090 Suprim SOC (PNY Epic-X vBios stock).3dmark-result 46.94 kB · 0 downloads

And here is one of mine

PNY Epic. Core 2736/1750/65C. Running on my gaming drive with everything running. lol

https://www.3dmark.com/pr/3394426

 

 

 

[jaybee83]

18 minutes ago, johnksss said:

And here is one of mine

PNY Epic. Core 2736/1750/65C. Running on my gaming drive with everything running. lol

https://www.3dmark.com/pr/3394426

 

 

 

 

well id say were spot on then 😄

 

ive dialed in new OC settings due to the changed V/F curve, seems like i can now do +20Mhz on the core offset, pretty nice. although not sure if this is vbios / driver / backplate cooling related or maybe a combination of all three. not 100% sure yet on the stability for D2D but should be pretty close. gonna do a session of Deliver Us Mars later to finish it up.

[jaybee83]

Aight boiz, heres the jist of my vBios testing.

 

Testing conditions: Alan Wake 2 Chapter "Summoning" Savegame 4K maxed out with all RT / PT maxed +HDR.

Driver: 576.28 (XtremeG Mod)

Duration: 60 min.

 

Flashing / Setup Procedure:

1) Flash vBios via NVFlash 5.867.0

2) Uninstall Drivers via DDU

3) First Reboot

4) Blackscreen upon Boot, switch to iGPU

5) Only mouse cursor visible, blindly type in password to login

6) Desktop appears, but no icons - switch to dGPU, all good

7) Install Drivers, adjust settings

8 ) Second Reboot for Nvidia Overlay functionality

9) Set 100% fan speed via MSI Afterburner, check HWInfo for readout

10) Check V/F curve offset vs. stock Suprim vBios, adjust core offset to reach 3240 Mhz at 33-34C (D2D OC profile for my card at stock vBios)

11) Run AW2 as detailed above, record via HWInfo in background. After 60 min. minimize game while still letting it run and do quick screengrab of dGPU section in HWInfo.

 

I attached the screengrabs to this post, together with the PNY Epic-X vBios ive ended up choosing, for anyone interested.

 

How did I choose which vBios to pick on TPU? Welp since most creation dates of the uploaded vbios files are the same, i just went by the most recently uploaded for highest chance to get the latest version for each card. Manually checking if its indeed the performance profile for cards that sport two vBios versions.

 

 

Quick conclusions:

 

- Asus Astral LC/OC and Gigabyte Xtreme Waterforce were disqualified pretty quickly due to lack of proper fan control (Asus) and very low max rpm (Gigabyte), respectively

 

- Stock Suprim SOC and Zotac Amp Extreme vBioses demonstrated the overall lowest average and peak wattage numbers, resulting in lower core clocks. So they dropped out of the race, as well

 

- Gigabyte Master and Master ICE vBios versions are basically identical, only difference here being the impact of differing ambient temps (+2C on the ICE), thus some impact on the clocks.

 

- In the end it was a neck-and-neck race between the Gigabyte Master and PNY Epic-X vBios. I ended up choosing PNY due to the following:

1) Highest avg. and peak wattage numbers

2) Highest core clocks for avg. and effective

3) Highest max. RPM for the fans (+200 above stock and the other fan functional vBioses)

4) Highest average voltage, even when accounting for differing ambient temps.

 

There u go guys, enjoy! Any questions or comments, lettem rip 🙂 

 

edit: as an afterthought, some additional things i could have included: higher measurement resolution in HWInfo to better get a grip on wattage peaks / shorter, burst-style loads in terms of standard benches & benchmark scores.

 

but hey, only got so much time to go around i guess 😄 

 

MSI RTX 5090 Suprim SOC vBios Testing.zip

PNY Epic-X OC 10-Apr-2025 (275445).rom

[electrosoft]

Looks like 5090 pricing really is starting to find that supply/demand price point as 5090s are staying in stock everywhere now much longer including Amazon with the $3400 shipped and sold from Amazon Epic-X ARGB.

 

My local MC has 9 Gigabyte 5090 OC's in stock for $2919. They would have been gone by 10am before.

 

Newegg has a slew of Gigabyte and MSI "minimal" combos in stock (IE, w/ a PSU OR MB only).

 

Pre-builts everywhere.

 

Pricing is basically back to MSRP now which is pretty high.

 

Now comes phase 2, the slow trickle of 5090s because they can't get any cheaper in the US and might get even pricier with the full effects of the tariffs not even being felt yet. AIBs will ship 5090s to other countries where it is cost effective and makes better sense to sell there than here.

 

5080's on down are priced low enough (relatively speaking even with AIB pricing), that they will continue to flow in along with 9070xt's and that is where we will see the real pricing war if applicable.

 

 

9 hours ago, tps3443 said:

Yeah, I’ve made an offer on a few things on eBay on a whim. And it always makes you attach payment method first. Wake up next day and the money is gone lol. Anyways, you keeping the Astral? I’d be tempted to keep both 5080FE and Astral if it were me. The 5080FE is amazing for the price. 

 

Yeah, I haven't bid on items in quite some time just outright purchased them so it was new to me. As a seller who just turned their store back on last night to unload some more hardware, it was a pleasant surprise. Nothing worse than leaving things up for bidding then have a winner take forever to pay or request a cancellation or accepting an offer and the buyer then goes awol.

 

I was tempted to keep both, but I have to be realistic with my tech budget constraints. As I slowly start to list stuff on my eBay store, I'll probably put it up for $1500-$1600. I'll also list my NH55 here and eBay and other small stuff.

 

I am going to start monitoring drops again. VPA has picked back up and BB has had two drops now of 5090 FEs.  If a 5090 FE drops on BB or VPA, I'll probably pick it up for comparison. I did the same with watching the 4090 FE and snagged one when it dropped many moons later with my 4090 Liquid for comparison. Liquid ran roughshod over it badly, so I returned the 4090 FE but with such a large pricing gap, I can settle for not so great silicon/cooling/D2D OC as long as it doesn't have coil whine.

 

Pricing differentials are in play, so if I snag a 5090FE that has no coil whine for that price, I would sell the Astral or if it happens before early June, just return it.

 

 

3 hours ago, win32asmguy said:

 

I am tempted to try and move my 5090 into my Intel 12900ks + z690 system this weekend and see if it does better with wow than the HEDT setup. I could even potentially try a 14900k or ks if it's supported via a bios update. That 12900ks seemed like it was pretty leaky in the raptor x15 but it might do perfectly fine even with good air cooling.

 

What z690 board and memory is it? I would suspect a decently tuned 12900ks system (5.5 all core, turn off e-cores if needed to achieve) and some halfway tuned DDR4 or DDR5 will outperform your HEDT. Even pure stock it might too.

 

As a baseline, do the Oribos and Hallowfall tests I posted at stock 5090 settings to see where they go with GPU utilization. Oribos might cap out even on HEDT at 4k Ultra max settings across the board RT high. Hallowfall will not but let's see how high it goes.

 

I know with my 9800X3D rig, flying around even had my 9070xt and 5080fe capping out routinely at 99% over Dornogal too. Oribos caps out at 99% ~186fps stock, ~210fps OC'd. Hallowfall would sit at ~92% stock, ~76% OC'd on the 5090 (capped out on everything else including 4090 OC'd).

 

27 minutes ago, jaybee83 said:

Aight boiz, heres the jist of my vBios testing.

 

Testing conditions: Alan Wake 2 Chapter "Summoning" Savegame 4K maxed out with all RT / PT maxed +HDR.

Driver: 576.28 (XtremeG Mod)

Duration: 60 min.

 

Flashing / Setup Procedure:

1) Flash vBios via NVFlash 5.867.0

2) Uninstall Drivers via DDU

3) First Reboot

4) Blackscreen upon Boot, switch to iGPU

5) Only mouse cursor visible, blindly type in password to login

6) Desktop appear, but no icons - switch to dGPU, all good

7) Install Drivers

😎 Second Reboot for Nvidia Overlay functionality

9) Set 100% fan speed via MSI Afterburner, check HWInfo for readout

10) Check V/F curve offset vs. stock Suprim vBios, adjust core offset to reach 3240 Mhz at 33-34C (D2D OC profile for my card)

11) Run AW2 as detailed above, record via HWInfo in background. After 60 min. minimize game while still letting it run and do quick screengrab of dGPU section in HWInfo.

 

I attached the screengrabs to this post, together with the PNY Epic-X vBios ive ended up choosing, for anyone interested.

 

How did I choose which vBios to pick on TPU? Welp since most creation dates of the uploaded vbios files are the same, i just went by the most recently uploaded for highest chance to get the latest version for each card. Manually checking if its indeed the performance profile for cards that sport two vBios versions.

 

 

Quick conclusions:

 

- Asus Astral LC/OC and Gigabyte Xtreme Waterforce were disqualified pretty quickly due to lack of proper fan control (Asus) and very low max rpm (Gigabyte), respectively

 

- Stock Suprim SOC and Zotac Amp Extreme vBioses demonstrated the overall lowest average and peak wattage numbers, resulting in lower core clocks. So they dropped out of the race, as well

 

- Gigabyte Master and Master ICE vBios versions are basically identical, only difference here being the impact of differing ambient temps (+2C on the ICE), thus some impact on the clocks.

 

- In the end it was a neck-and-neck race between the GIgabyte Master and PNY Epic-X vbios. I ended up choosing PNY due to the following:

1) Highest avg. and peak wattage numbers

2) Highers core clocks for avg. and effective

3) Highest max. RPM for the fans (+200 above stock and the other fan functional vbioses)

4) Highest average voltage, even when accounting for differing ambient temps.

 

There u go guys, enjoy! Any questions or comments, lettem rip 🙂 

 

MSI RTX 5090 Suprim SOC vBios Testing.zip 2.89 MB · 0 downloads

PNY Epic-X OC 10-Apr-2025 (275445).rom 1.87 MB · 0 downloads

 

All roads lead to Rome with the PNY doing it while giving you the best overall clocks. In a game of inches it seems, it wins. One inch more is one inch more (insert joke here).

 

Nice work and write up!

 

Did you try for an overall higher OC as you were testing the vBIOS's or did it come down to your card's silicon wall so trying to find the best vBIOS that ran within it to produce the best results?

[jaybee83]

18 minutes ago, electrosoft said:

All roads lead to Rome with the PNY doing it while giving you the best overall clocks. In a game of inches it seems, it wins. One inch more is one inch more (insert joke here).

 

Nice work and write up!

 

Did you try for an overall higher OC as you were testing the vBIOS's or did it come down to your card's silicon wall so trying to find the best vBIOS that ran within it to produce the best results?

 

spot on, the differences are miniscule. ofc when doing benches those differences should be larger. and yup, the second one was it. i figured if a vbios allowed for a bit moar powah / higher voltage i could also squeeze a bit more out of it, so best to compare oranges to oranges and set all vbios testing to the same clocks at idle 33-34C temps.

 

naturally, one might argue that slight differences in the overall V/F curve shapes might account for slightly higher / lower clocks depending on the vBios. but since im basically only interested in the highest V/F points on the curve, the range "should" be minimized.

granted, if one would be interested in undervolting and going further down the curve, those variabilities would likely be larger. but not interested in UV 😄

 

all in all, 5090 beggars cant be choosers 😛 being stuck with the 600W wall means well have to find other ways to squeeze a bit more outta these chips haha