Well now that the Resizable Bar is out, gave it a spin. Gaming its amazing 20+ FPS boost in Cyberpunk 2077 alone combined with DLSS and now this extra 20 plus FPS on top of it on my 3070 is a FTW just wow on the gaming experience. 60-80 FPS range, depending on inside outside, congested area to out in the middle of nowhere. Really impressed on the gaming aspect of it alone. Average 64-66 FPS. 1440p Ultra,Ultra-Psycho Reflections:DLSS Quality or Balanced.


Now while bone stock cpu, gpu the works I saw about a 400 points increase in 3DMark Timespy. Was all excited so did some overclocking, was not that impressed, I gained only 6 points. I also noticed with the resizable bar bios it would let me gpu go past the crashing clocks but would upper cut my total score at least on stock voltages. 

So my over a decade long made statement on these forums still holds true today, 3DMark is a benching application. In no way does it really reflect real world gaming, it will help you stress your hardware though to maybe trigger an issue, or give you something fun to do with your time playing with your hardware. After the Resizable bar with no real benefits from it while overclocked, its pretty safe to say that. At least while stock the resizable bar improves performance in games so much its really a game changer. 3DMark though cannot say that unless you are running a stock cpu/gpu/ram then you get a nice boost in your score about 400 points from your last bench without the resizable bar. Once OCed at least in my case on my 3070 Ultra it was not really anything special or way to really get an edge for a higher score.

donta1979Now while bone stock cpu, gpu the works I saw about a 400 points increase in 3DMark Timespy. Was all excited so did some overclocking, was not that impressed, I gained only 6 points. I also noticed with the resizable bar bios it would let me gpu go past the crashing clocks but would upper cut my total score at least on stock voltages. 

So my over a decade long made statement on these forums still holds true today, 3DMark is a benching application. In no way does it really reflect real world gaming, it will help you stress your hardware though to maybe trigger an issue, or give you something fun to do with your time playing with your hardware. After the Resizable bar with no real benefits from it while overclocked, its pretty safe to say that. 

 
Or maybe Rebar isn't doing anything in 3DMark because it's not whitelisted in the NVidia drivers yet?

FWIW, I have compared before and after ReBAR for Borderlands 3 canned benchmark and 3DMark Timespy with my 3070 Ultra.  Neither showed significant improvements.  BL3 +1fps, Timespy less than a couple dozen points.  I know everything is properly enabled...
 
Gaming on BL3 seems like I'm getting 10-20 fps gain depending upon map.  1080p badass preset.  I don't think the bench inside BL3 is representative of actual gameplay.  Neither is 3DMark.
 

Sorry, donta1979 you are not getting a 20+ FPS boost in average fps in Cyberpunk 2077. Just not happening. Resize BAR is not magic and not capable of producing results like that. I have Resize BAR enabled on my RTX 3090 and have only seen at best a few fps bump in performance in Cyberpunk 2077. While notable, it's not exactly Earth shattering as you are claiming. I've tested in 1080p, 1440p, 4k and ultrawide 3840x1080. The biggest gain was in 1080p at about 5-6fps and it tapering off quite a bit as resolution increased with 4K maybe gaining 1 fps. 
 
If you are seeing 20+fps you are not testing in the same scene, with the same patch (1.2 patch to be exact) or not using the same settings between runs. 
 
EDIT: Just did another confirmation test at 3840x1080 (which is close to the pixel count of 1440p for reference) using the same scene, same settings. The difference is 4fps from 62 to 66 with resize BAR disabled to enabled respectively. 
 
resize BAR disabled

 
resize BAR enabled

@ donta no, no you're not. It's just not possible unless you know or are doing something that literally no one else on the planet knows. 20+ fps is fantasy land dude or do you really think your RTX 3070 is performing miles better than anyone else? C'mon man, be realistic here. I know you're obsessed with that game but what you are trying to claim is pure nonsense. I've also tested like I said 1080p, 1440p and 4k in addition to my ultra wide resolution. 

Donta,
do you have any special patches installed?🤔 I know you like to tweak the games you like to play. Share the settings or tweaks you have😁

atfrico
Donta,
do you have any special patches installed?🤔 I know you like to tweak the games you like to play. Share the settings or tweaks you have😁

Nope fresh install of windows and games. Got a Samsung 980 Gaming m.2 as my main drive so kind of had to. No EZ Config mod on new pc. All developer options disabled. Just installed it with GOG a week or so ago.

Here are my settings, I only disabled film grain because it makes textures look pixelated, then motion blur to help the game look sharper its one of those "enhancements" that just does not add to much visually while making the rest of the game look fuzzy at times at a nasty performance hit. Nvidia overlay is disabled, nor did I let that thing touch the game. So why am I doing so well? The hell if I know. I do know I got 20 more FPS total from the Resizable Bar update on my board and gpu. Before this it was good in the 40-50 FPS 41-45/46ish FPS average range pre patch, after patch about 45-48 FPS but nothing like this after the bar updates performance exploded.

donta1979
I am very much getting 20 FPS more.

 
They also had a giant patch just the other day for Cyberpunk. I tried the patch before installing the new Rebar BIOS, and gained 10 or so FPS. That might be part of the equation.
 

Robs03gts
 
Ignore that troll..  he's been doing this for years telling people he knows more about their own systems performance than they do lol..  username says it all and he's just bitter because he isn't doing something right.

I have known Crap for years and we have butt heads a lot over the past decade and a half haha. He does knows his stuff he just does not know his stuff like I know my stuff on the Developer end of things just a different view on the same subject. I normally go to him once I have been absent for a long time and ask him what's up. I have had magical cards in the past like my 280 FTW's he called magical. Yet they really were magical and broke some world records first by me then by Q56Monster that I sold them to he wanted them bad. Premium silicon I lucked out. I would not call crap a troll I would say hes very strong in his knowledge to his opinion on his point of view. There has to be something crazy/lucky or strange going on if I am exploding in performance, its not the Geforce Experience I do not let that thing touch my games with a ten foot pole on settings. just a new 3.0 ssd, fresh install of windows 10, fresh install of the game, patch, then bar support added to my motherboard and gpu. For whatever reason my performance exploded, even with the dev options off.... I frankly do not get it.

CraptacularOne
 
 
 
@donta1979 I got your PM requesting assistance investigating this and for the sake of transparency lets continue our discussion here in the open so that we may better get to what's really going on here and maybe help some others along the way. 
 
I would first do some ground verification tests by changing nothing in your "in game" settings and in a stationary location. Go into your BIOS and disable Above 4G decoding to deactivate resizable BAR and launch a save file. Let the game run for a few seconds to make sure any and all assets are loaded. Don't move around and just record your fps. Then go back into the BIOS and re enable above 4G decoding and repeat the above and note the fps. 

Both Disabled 4g crypto and bar in bios I am getting about 41-58 FPS range Average 44-51. Do not see 70-80 FPS on occasion anymore I do sometimes see 60-65 briefly. First time playing the game since the Bar update with the Bar and the 4g setting in bios disabled.

atfrico

donta1979

Both Disabled 4g crypto and bar in bios I am getting about 41-58 FPS range Average 44-51. Do not see 70-80 FPS on occasion anymore I do sometimes see 60-65 briefly. First time playing the game since the Bar update with the Bar and the 4g setting in bios disabled.

Can the bar' performance in bios differ in the coding from board to board?🤔

Maybe, and I don’t have a bios on can flip back and forth on. It’s possible. But I did lose a good chunk of performance disabling the two things in the bios. But it’s still a little better than before the the bar flash.

One thing I have noticed since the bar flash, clocks that would stop the 3DMark test no longer do they just take punches out of my score. So something on the card with that flash changed.

Depending on how many NPC’s walk in to how many cars high 48-58 FPS fluctuations as each new asset gets on the screen. Camera did its automatic looked to the sky getting 62 FPS randomly. While pointed at the back of the motorcycle tire and looking up. Loaded a save didn’t touch a thing.

donta1979
Depending on how many NPC’s walk in to how many cars high 48-58 FPS fluctuations as each new asset gets on the screen. Camera did its automatic looked to the sky getting 62 FPS randomly. While pointed at the back of the motorcycle tire and looking up. Loaded a save didn’t touch a thing.

Shouldn't fluctuate 10fps while sitting stationary should be a relatively consistent number. Regardless can you give us numbers with resize BAR on and off in the same place while not moving. 

donta1979
Think it’s due to the NPC’s, NPC in vehicles coming in the constantly. Along with the 4.8-5.3ghz of my cpu, gpu downclocking/up clocking. 4g and bar back on in bios 64-68, Same place/time and save. Camera finally did it’s own random afk position to the sky thing FPS shot up to 76-80. depending on how many sky transports passed randomly.

So now you're saying if I'm reading this correctly, it goes from mid 48-58 disabled to 64-68fps enabled correct? That is still rather on the high side if we take your maximum values of 58-68 as that's a 17% gain. Still very high, but not completely impossible given the right situation or area. I suspect you aren't using a frame counting tool to average your fps in a given and repeatable sequence to see what your averages actually are? If you are just "spot checking" and haven't really run the game since the 1.2 patch and have resizeBAR enabled I guess it isn't entirely unreasonable to think you are getting more out of something than you really were. If you were to run a canned repeatable area and averaged out the gains I suspect you'll see just like the rest of the world that it's only about 5-7% gain overall. There may be some areas that respond a little better or a little worse but the average gain ISN'T going to be a 20fps gain. That's quite literally impossible in this already GPU bound game. There is only so much that giving the GPU faster and more effective access to it's VRAM buffer is going to help in GPU saturated situations. 

CraptacularOne
 
So now you're saying if I'm reading this correctly, it goes from mid 48-58 disabled to 64-68fps enabled correct? That is still rather on the high side if we take your maximum values of 58-68 as that's a 17% gain. Still very high, but not completely impossible given the right situation or area. I suspect you aren't using a frame counting tool to average your fps in a given and repeatable sequence to see what your averages actually are? If you are just "spot checking" and haven't really run the game since the 1.2 patch and have resizeBAR enabled I guess it isn't entirely unreasonable to think you are getting more out of something than you really were. If you were to run a canned repeatable area and averaged out the gains I suspect you'll see just like the rest of the world that it's only about 5-7% gain overall. There may be some areas that respond a little better or a little worse but the average gain ISN'T going to be a 20fps gain. That's quite literally impossible in this already GPU bound game. There is only so much that giving the GPU faster and more effective access to it's VRAM buffer is going to help in GPU saturated situations. 

Correct spot checking I have nothing but time these days, the only extra thing I run in the background is MSI Afterburner, GOG, Magician, Blue Tooth Devices, Nvidia control panel/Geforce Experience"I should remove that at some point unless I want to use the sharpen tool and give up more fps". No I didn't run a frame counter for this, have that running in the background unless I am really using it/it's important.

I picked one of the more demanding areas of the game. depending on where I am in the game it is 20 FPS gained, along with whats going on what assets get on screen. I get more jumps in FPS while driving around/running around on the map to different areas, NPC's, Car's, Time of day/Night, all play a factor.

Prior to the patch I was at 180hrs, since patch I put in an additional 40+ I want to say..., in that time 20+ have been with bar support, and it seems with bar support off whatever happened to my card something did go on if I can overclock it past speeds that would make 3dmkark crash/stop the test and now its not doing it and finishing the test after the bar flash. So have seen the performance differences though the phases from prior to patch, the patch, the patch with bar. In the game what assets are on the screen will change your FPS. As I observed since the game launching, first big awesome push was DLSS and getting to use it. I also know Bar support you get less of a gain if you are overclocked/higher clocked way past your stock specs. Thus my 400 points in 3dmark at stock vs my 6 little points gained with bar support while overclocked. I am guessing that's the same here, I only have a 3070 less speed vs a 3080 or 3090, so the bar maybe is doing it's thing like the benchmark with the 400 points gained while stock in 3DMark. Overclocked/running faster your gains are less. Also what assets hit your screen and have to be swapped out/called on will affect FPS. Even just sitting there when the area starts to populate/depopulate as time goes by as NPC's go off and fade into the distance into nothingness mix that with your cpu randomly clocking, gpu fluctuating between clocks. Then picking one spot in the game, things are not being swapped out as much, so your gains will be less vs more. Its why benchmarks tend to constantly be moving the environment unless your in a cpu only test. Gains in one area will be higher than another if you are constantly going around a game level or map with what is in them. Overall I did get a 20 FPS boost while playing the game. Why I still do not know. Maybe my theory of stock vs overclocked, a faster card vs a slower card and the gains gotten on what the bar does is a thing maybe its not? I have been reading on reddit, and discords those with 3080's and 3090's get less of a boost from bar its minimal as I see the post/replies complaints/questions "is there something wrong with my card I am not getting the gains on my 3080 to 3090 as some are reporting on their 3070's to slower models of my own card".  It is honestly the only thing I can think of at this point as I sit down and think about it.