Cryengine Neon Noir Benchmark, Submit Your Scores !

[ENTERPRISE]

 

 

Remember my test results are run in 1920 x 1080 Don't think I'd be able to get near ya in 4k

 

That is a fair point. I should try in 1080P and see what I end up with.

[schuck6566]

As an example, here's my test results again, but run in 1366 x 768. lol, Looks like I have a 2nd generation super 2080ti instead of the 1st gen. 2070.

 

[ENTERPRISE]

I will have to run these benches at 1080P is a moment for direct comparison, I just ran Superposition at 4K below is the score.

 

[ENTERPRISE]

1080P

 

Oddly enough, Now when I try and run Noir, I get like 20FPS regardless of what res I am running. Will check BIOS as that is all that I have changed as of late.

[mllrkllr88]

I am getting setup to play with this bench again. Looking forward to helping out with the testing and seeing if we can figure out the how it scales.

 

My biggest complaint is that this bench cannot be run in offline mode, and requires a huge download.

[mllrkllr88]

Bone stock run with absolutely ZERO tweaks or OS optimizations.... This is just a simple 9900k on Z170 with XMP memory and a stock 2080Ti KP. It looks like the score is higher than it should be, bummer this bench is so inconstant.

 

[ENTERPRISE]

Bone stock run with absolutely ZERO tweaks or OS optimizations.... This is just a simple 9900k on Z170 with XMP memory and a stock 2080Ti KP. It looks like the score is higher than it should be, bummer this bench is so inconstant.

 

Nice, but what RES is that at ?

 

I will re-bench, but need to disable HPET as I enabled it for bench mate the other day and ever since doing that, the NOIR bench runs at like 20FPS for some reason. HPET and a BIOS update was all I have done since the last working NOIR bench. I cannot see it being the BIOS, so I can only look at HPET. Other benchmarks and games work just fine though to add to the confusion.

[mllrkllr88]

 

Nice, but what RES is that at ?

All basic mode... 1920x1080 with 60Hz

 

The bench automatically loaded in a window, which I have no clue what size that is. I did nothing but install and run on a dirty v1909 W10 install.

 

 

[ENTERPRISE]

All basic mode... 1920x1080 with 60Hz

 

The bench automatically loaded in a window, which I have no clue what size that is. I did nothing but install and run on a dirty v1909 W10 install.

 

 

Thanks bud, this makes your score make more sense. I was like ''That better not be near the res I am benching at'' lol. Tomorrow I will tackle this again after resetting HPET/fixing the issue and I will see what I get at 1080P as a direct comparison.

[mmcdonald12]

[schuck6566]

On a separate note, Superposition benchmark is able to put a fair load on a gpu card if you want to see what kinda reults you can get. it ran my 2070@ 99% during the test.​

 

tried using afterburner for my overclock and got this for the score. previous was using evga's software.

 

[ENTERPRISE]

 

tried using afterburner for my overclock and got this for the score. previous was using evga's software.

 

Nice improvement, what clocks are you running on the 2070 ?

[schuck6566]

Here's a screen shot of afterburner & gpu-z it also seems to run a little cooler on the test since I went with a custom fan setting instead of the default 1 that evga's software was using with the OC.

 

[ENTERPRISE]

Not bad. Been a while since I owned an AMD card, but If I remember rightly the memory OC headroom is a lot tighter than Nvidia counterparts ?

[schuck6566]

Not bad. Been a while since I owned an AMD card, but If I remember rightly the memory OC headroom is a lot tighter than Nvidia counterparts ?

 

lol, the latest amd video cards I've owned are hd3450's and a hd5770 that came in my Dell & I replaced that with the GTX970. My Team red experience was slight,I based my Nvidia fandom on the fact they worked 4 me & the terrible performance of the 5770. I had a GT730 that pretty much matched the hd5770 performance wise.

I've NO idea how the new amd cards are OC'ing but if they're starting to match the nvidia's @ base clocks,they may be my next choice in a card. Edit: Just a comparison 'tween the GT730 & the hd5770 "The GeForce GT 730 v2 MSI OC 1GB Edition has 384 Shader Processing Units and the Radeon HD 5770 1024MB has 800. However, the actual shader performance of the GT 730 is 386 and the actual shader performance of the HD 5770 is 340." "The GeForce GT 730 v2 MSI OC 1GB Edition requires 49 Watts to run and the Radeon HD 5770 1024MB requires 108 Watts."

Edited May 1, 2020 by schuck6566

[ENTERPRISE]

Yeah not having owned an AMD card in a while I don't have a comparison to base anything off. The last AMD GPU I had was the 295x2. Now that was a beast of a card but as you eluded to in your post, the power draw of AMD cards back then were crazy in comparison to the green team. I think they have tightened the gap since but will have to look at a current gen AMD card versus a current gen Nvidia card. I tend to not look at the power usage of a GPU as it was rarely a concern.

[schuck6566]

So again a fairly odd score with Neon Noir I think. I even updated to the latest engine as advised. Will run super position and see what we get.

 

 

 

 

I ran across this story & thought of your problems with this test.It's slight,but maybe part of the problem? Is windows seeing this as a game maybe? https://www.techspot.com/news/85137-windows-10-game-mode-negatively-impacting-gaming-performance.html

[J7SC_Orion]

Neon Noir at 4K and 2080 TI NVlink..."AFR2" settings in pic below. Will try CFR (checkerboard / tile NVLink setting on the weekend if I can get it going). Anyhow, I really like this bench...!

 

For comparison, 2x 2080 TI at Superposition 8K, Heaven 4K and Valley 4K

 

...will try to rerun those with "CFR" as well - sometimes, CFR produces lower FPS but much better 1%, 0.1% frame times :-)

 

 

 

Edited May 9, 2020 by J7SC_Orion

[ENTERPRISE]

 

I ran across this story & thought of your problems with this test.It's slight,but maybe part of the problem? Is windows seeing this as a game maybe? https://www.techspot.com/news/85137-windows-10-game-mode-negatively-impacting-gaming-performance.html

 

Thanks for that, I did not think of that one, I will have to give that a go and see if I fair any better. Down to one 2080Ti these days, sold the other as SLI is mostly dead. However Noir was a one GPU bench anyway.

[damric]

Check it out. I stopped crying over not being able to find any new hardware, so instead I unlocked the power limit of my Vega 56 through the registry. Now I'm only limited by the silicon (the way it used to be, and should be). I re-ran all my 3DMarks for HWBOT and this pig was pulling over 400W according to the Afterburner overlay. It is now able to sustain clocks around 1775-1800MHz, and I can bump the HBM a bit more too without it adding additional power limit problems. My HBM is still pretty weak though.

 

 

[mouacyk]

On 1/21/2021 at 1:36 PM, damric said:

Check it out. I stopped crying over not being able to find any new hardware, so instead I unlocked the power limit of my Vega 56 through the registry. Now I'm only limited by the silicon (the way it used to be, and should be). I re-ran all my 3DMarks for HWBOT and this pig was pulling over 400W according to the Afterburner overlay. It is now able to sustain clocks around 1775-1800MHz, and I can bump the HBM a bit more too without it adding additional power limit problems. My HBM is still pretty weak though.

Looks like a decent jump from your stock run.

 

Throwing my 1080TI and 3080 1080p results out there.

 

1080TI / 2100MHz 12,000MHz @1080p:

 

3080 / 2070 MHz 21,000MHz @1080p: