Info Oblivion FPS Test

[BoredErica]

The Fallout 4 Draw Call thread is successful so I was wondering if anyone was interested in an Oblivion FPS test. The reason why I want an Oblivion test is because the engine in Oblivion is very different from FO4. Not sure if anyone is interested but I figure it was worth a shot at least. Here is the procedure:

1. Ensure latest version of Oblivion is not running any mods, including OBSE or DXVK. (DLCs are encouraged.)
2. Replace Oblivion.ini with ini found in this link, and move the save file included to save directory. Keep the read-only flag on the ini. https://mega.nz/file/C1pynYKL#PcJqg3xUrNdaJWKwfpwl59HHiNGmjwKp29D4ldRagNw
3. Get FPS value 3-4 seconds after loading in. Try not to wait too long after that or take FPS before 3 seconds.
4. With just 1 mod (which has no dependencies), ImpeReal City - Unique Districts, FPS can be pushed down to 38FPS already on my CPU. DL is only 150kb. Get it here. (First or second DL, doesn't matter.)

Pictures are appreciated. If you've manually tune all subtimings on ram, let me know (binary, either you did or you didn't).

If Oblivion Launcher starts and says it is changing graphical settings, it should attempt to do so and fail silently since the ini is read-only. If you dunno where Oblivion ini goes because you don't use a mod manager, by default it belongs in My Documents/My Games/Oblivion.

 

[KompuKare]

Hi,
had some problems with using your ini.
Whenever I tried to use yours, the game's launcher detected changes and would re-detect everything.
So I left it at the default it had detected (medium), changed to 1280x720, and checked all three items under distances.
That got me these results:

and with the mod (I installed the all-in-one esp version):

Also had Process Explorer running on the second screen:

And it looks like Oblivion runs everything on one thread! At least FO4 put something onto other threads even if one was a clear bottleneck.

Anyway, a 3600 should not beat a 5600X, I decided to re-run at Ultra too, but stick to the 1280x720 which was in your ini. (This goes a bit against the spirit of what the FO4 benches are trying to do, where the idea is to put all the stress on the CPU.)
Without the mod:

and with the mod:

Process Explorer looked the same too:

To recap:

Medium settings - 1280x720
without mod: 81
with mod: 60

Ultra settings - 1280x720:
without mod: 50
with mod: 37

My system: Ryzen 5 3600, 2x32GB DDR4 3200, GeForce 3050, Win10 a64

So it looks like changing the details does become CPU bound. GPU fan never spun fast during this, and when I looked in Process Explorer it wasn't been utilised much. OpenHardwareMonitor saw a max load of 61% on Core #2, and a max GPU load of 55% during the Ultra without mod run.

There is a bottleneck, but hard to see where it is. Lack of threading is an obvious problem but I suspect that there's more going on.

 

[BoredErica]

Hi,
had some problems with using your ini.
Whenever I tried to use yours, the game's launcher detected changes and would re-detect everything.
So I left it at the default it had detected (medium), changed to 1280x720, and checked all three items under distances.

To recap:

Medium settings - 1280x720
without mod: 81
with mod: 60

Ultra settings - 1280x720:
without mod: 50
with mod: 37

My system: Ryzen 5 3600, 2x32GB DDR4 3200, GeForce 3050, Win10 a64

So it looks like changing the details does become CPU bound. GPU fan never spun fast during this, and when I looked in Process Explorer it wasn't been utilised much. OpenHardwareMonitor saw a max load of 61% on Core #2, and a max GPU load of 55% during the Ultra without mod run.

There is a bottleneck, but hard to see where it is. Lack of threading is an obvious problem but I suspect that there's more going on.

Hi, very interesting.
I almost never use Oblivion launcher because it often changes settings to their presets which always gave me headaches. The ini I used is more intense on CPU than ultra high which is why your FPS was higher than mine. Using launcher, setting to ultra, resolution back down to 720p, I got 130fps. In playing modded Oblivion so much, I forgot what it's like opening Oblivion for someone without Mod Organizer, where I can just start Oblivion without going through the launcher.

Compared to Ultra High, my ini has:
fDecalLifetime=100.0000 (10 default)
fShadowLOD1=0 (400) (Shadow distance. 0 means no lod fade out for shadows)
fShadowLod2=0 (200)
fLightLOD2=40000.0000 (1500) (How far until lights fade)
fLightLOD1=35000.0000 (1000)
iActorShadowIntMax=250 (10) (Max # of shadows from NPCs)
iActorShadowExtMax=100 (10)
iShadowMapResolution=512 (256)
bAllow30Shaders=1 (0) (Probably does nothing)
fGrassStartFadeDistance=30000 (6000)
fGrassEndDistance=30000 (7000)
bUseWaterReflectionsMisc=1 (0)
bUseWaterReflectionsStatics=1 (0)
bUseWaterReflectionsTrees=1 (0)
bUseWaterReflectionsActors=1 (0)

I don't think resolution/quality settings actually matter here. The game is so monumentally CPU limited in this scene, even at 8k my FPS doesn't get any higher vs 960x540. So I would say in both of your runs, you were CPU limited. (It's a Bethesda game from 2006.) However, I suppose it can't hurt to decrease resolution from 720p down to 540p for the bench. I've updated the ini in OP to lower resolution down to 540p. UI is still very legible even at 4k desktop resolution so it's still convenient for people to use.

The reason why Oblivion launcher changes to presets on launcher is because your GPU differs from mine, so your uVideoDeviceIdentifierPart1/2/3/4 values differ from mine. The launcher starts, reads the ID and sees your GPU being different from mine, and thinks it's time to adjust graphics settings in accordance with your new GPU. Two ways to get around this:

1. Start launcher, letting it reset settings. Close launcher. Check ini file, and at end of Display should be the uVideoDeviceIdentifier ini settings. Just copy those 4 settings, and replace with mine in the ini file. This is kind of annoying.

2. Easier is to simply make the ini file read-only. Launcher will attempt to reset graphics settings, fail, and the game will load my settings.

The instructions now only feature 3 steps, as ini and save file are both included in 1 file. Ini file should now come with read-only checked. What confuses me is how you managed to get the exact same FPS I got in the second set of tests... Would you mind redoing it but with read-only ini instead? Thanks.

edit:
Some extra misc stuff. I tried to measure FPS differences in SSE/Oblivion/FO4 with DDR4 3600 c14 1T vs 3000. The older the Bethesda game, the less ram seemed to matter. This is concerning to me. A OCN member tried to test Oblivion w/ 5800x3d with another version of this test and we found 5800x3d gave almost no benefit at all. But I'd like to reserve judgement until somebody tries this test.

 

[KompuKare]

What confuses me is how you managed to get the exact same FPS I got in the second set of tests... Would you mind redoing it but with read-only ini instead? Thanks.

Done:
Without mods: 38

With the mod: 33

I did try launching without the launcher but Steam seemed to not like it as it wasn't from my C drive. Don't recall seeing that with Steam for other games. I do actually have both Morrowind and Oblivion on GOG too as I though I might want to try modding them again without Steam getting in the way, but Steam was easier to install quickly.

I guess it is possible that the earlier games don't really care too much about memory or cache. FO4 was supposed to be one where it mattered the most - not that anyone really benched modded Skyrim.

All the Process Explorer bit showed me is that unlike FO4 there really is only one main thread. Yes, it also showed CPU time against the Nvidia drivers but for FO4 there were two heavy threads - just one was heavier than the others. Speculation might be that the heaviest one in FO4 does the drawcall and render pipeling, while the less heavy does things like AI. Oblivion looks like it does things far stricter serially.

If the faster IMC or the huge L3 on the X3D doesn't matter for Oblivion, then the next question is: how does Alder Lake cope with that game?

 

[BoredErica]

Done:
Without mods: 38
With the mod: 33

I guess it is possible that the earlier games don't really care too much about memory or cache. FO4 was supposed to be one where it mattered the most - not that anyone really benched modded Skyrim.

All the Process Explorer bit showed me is that unlike FO4 there really is only one main thread. Yes, it also showed CPU time against the Nvidia drivers but for FO4 there were two heavy threads - just one was heavier than the others. Speculation might be that the heaviest one in FO4 does the drawcall and render pipeling, while the less heavy does things like AI. Oblivion looks like it does things far stricter serially.

If the faster IMC or the huge L3 on the X3D doesn't matter for Oblivion, then the next question is: how does Alder Lake cope with that game?

It's not surprising to me Oblivion only sits on 1 thread since it's from 2006. CS GO is also (I think?) a very old first person game and it doesn't benefit from more cache at all. Probably a bad comparison though. Problem is, if cache/faster ram doesn't improve performance much in Oblivion, Fallout 4 performance will grow more quickly in FPS over generations I think. If the player uses Unique Landscapes and such for Oblivion, in general I find Oblivion far harder to run smoothly compared to FO4. Especially now, with 5800x3d and Alder Lake. Oblivion frame rate can be low, but the worst of it are the .5-1s long stutters when moving around outside cities.

Alder Lake would be very interesting to compare! Hope someone else shows up with results. I find it interesting you don't seem to lose as much FPS percentage wise as I do when benching with the mod on.

edit:
In my normal playthrough I have ambient occlusion effects on via OR, and at 8k the effects scale kindda poorly. So at 8k I can actually be GPU limited in few select scenarios outside. It's not too common and when it happens frame rate's already significantly above 60. Improving FPS in those scenarios don't matter that much when I move my character at 100 athletics and the game stutters another .5-1s a short while later, or when the FPS varies wildly as I pan around the scene quickly. And fully modded, my own personal Oblivion CPU test has my framerate going down to 26.5, and that's with DXVK. Maybe by the time I get 120fps there, I can run around world without stutters.

 

[BoredErica]

A friend of mine has run the benchmark as well.

The chart really needs someone with an Alder Lake CPU, so if someone would take this streamlined test (just doing unmodded is ok) that would be awesome, thank you.

 

[BoredErica]

The 12600k isn't overclocked, for whatever that's worth. Very interesting nonetheless.

Probably overclocked Alder Lake still beats 5600x here, but at a smaller margin than I would've assumed. More Alder Lake or 5800x3d results are always welcome. Remember, the benchtakes <60s to set up if you're familiar with Oblivion modding and have it already set up!

 

[BoredErica]

Here we go. Got a 11900k. Friend says Andrea might have a bloated system and get derpy FPS. *shrug*

You can see video of Audley's submission here:

https://media.discordapp.net/attachments/897296277329563738/988723885950255134/explorer_qzhCPL6Dhd.gif

 

[BoredErica]

Two more submissions.



The 12600k result is sus. It is an AMD GPU too and I dunno if that has any noticeable impact.

 

[KompuKare]

The 12600k result is sus. It is an AMD GPU too and I dunno if that has any noticeable impact.

For FO4, AMD GPUs do show a higher or far higher overhead. Polaris is pretty old too but unsure whether that matters too.

 

[BoredErica]

I wish more people did the test with the 1 mod. I couldn't think of a good way to load the game down more without doing weird stuff, or making the tester DL many mods. In my current modded game, I'm getting 26.5 rather than 37FPS. Still a long road to 120FPS. The real improvement though I think will be reductions in stuttering and I dunno how that improves over time.

At this rate with 12900k, I'd still only get ~39 FPS in my load order for Oblivion.
73FPS FO4 in Fanuel Hall w/ ultra shadows + full mod load order.
20.35FPS in my giant settlement.

Apparently I'm not even allowed to DM people on here so I can't ask people to help.

 

[BoredErica]

For FO4, AMD GPUs do show a higher or far higher overhead. Polaris is pretty old too but unsure whether that matters too.

Yeah. I didn't know if it applied to DX9 games too though.


I benched my slow laptop. I dunno what it's running ram wise. I think it's 1x4GB 2400. Software can't read it.

A person attempted to bench their 5800x3d with an earlier version of my test. They were using AMD GPU though and scored WORSE than my 5600x. There is a thread on Reddit of a person claiming 12% FPS increase moving from 5800x to 5800x3d. If anyone has 5800x3d + Nvidia GPU, that'd be great. In either case, I think Alder Lake is the best option for Oblivion. Even if 50FPS x 1.15, that's still 57.5FPS and nowhere near 68-73FPS of Alder Lake I think. My plan currently is to buy 13600kf + a used mobo, sell my mobo + 5600x. A cheap upgrade that should provide a massive boost in performance for my use case. I'm hoping 13600KF can match 12900k in perf while being significantly cheaper than used 12900kf.

 

[BoredErica]

Got 5800x3d, 9750H, 7700, 2400G, a10-8700p.

As suspected, extra l3 doesn't improve perf as much as Fo4. IMO: If 11990k is hitting 54fps, it's conceivable for 12900k to hit 73fps. We are comparing highly clocked 12900k to a stock 11900k that can't entirely hit the highest part of the boost.

If I get another good Alder Lake result, I think I'm good until next gen arrives. Note that Audley's 11900k is beating my 5600x slightly overclocked by 8%, which compared to avg results in things like 720p testing from Techpowerup, is atypical.

 

[BoredErica]

I got hold of another laptop.

 

[BoredErica]

7950x bench here. 2 things to keep in mind:
1. Their CPU was clocked 5.41ghz which is very low for 7950x. An OCed 7600x should beat it.
2. The 12900k submission was using ddr4 ram (which won't make a huge difference I assume since L3 hit rates are good in Oblivion).

 

[BoredErica]

Looking for Raptor Lake test for Oblivion and Zen4 for FO4!
Please help contribute.

Also, averages are now geomeans.

 

[BoredErica]

New update courtesy of DryBones12.

Raptor Lake results coming in coming weeks.

 

[Hotrod2go]

I have Oblivion game of the year edition deluxe via Steam, will this affect the scores in that chart if I follow the instructions for the benchmark?
Also about half way through a modded playthrough atm. So shifting that data & swapping it for this single mod in order to benchmark could cause problems?

 

[BoredErica]

I have Oblivion game of the year edition deluxe via Steam, will this affect the scores in that chart if I follow the instructions for the benchmark?
Also about half way through a modded playthrough atm. So shifting that data & swapping it for this single mod in order to benchmark could cause problems?

No problem to the GOTY edition, I assume that's the version everyone runs these days. As long you ensure you're actually using my ini file and there are no other mods enabled, it should work fine. Most people use a mod manager like Mod Organizer, which has profiles you can set. So you could set a new profile, just enable that one mod, set it to use my ini, and you should be good to go. (OBSE shouldn't be used, and more importantly, no DXVK for the test.)

It should be obvious if the game is using my ini because the game will be very low resolution and windowed.

So it can totally be done, it just has to be done correctly.

 

[BoredErica]

Anandtech's forum is so much better than OCUK's, lol. Here's an update. That 5950x scored very high for non3d zen3 part. My OC is not ready yet so the numbers for 13600k are preliminary. Having said that, I'm surprised 13600k is barely beating 12900k. Sure, their ram is DR but I'm going to 5.7ghz, and ram doesn't matter that much for Oblivion. So I'm not entirely sure what's going on there.

Before my part came I estimated my fps in Oblivion and FO4. I can say so far FO4/SSE performed better than I thought and Oblivion worse. Overall mixed bag but I'm still happy, since this is the biggest and most price efficient upgrade for me in almost a decade. More to come in the future. I am busy tuning all my freakin' ram timings, which will take a while.

 

[BoredErica]

New update. My ram OC is getting close to being done...
Dang that 9900k is scoring very highly as well.

 

[BoredErica]

2 more updates.

 

[BoredErica]

13600kf OC is stabilized. This is my daily driver settings. I can push for 5.8/5.9ghz but I rather play it safer.

Some extra benchmarking data:

As expected, ram speed matters lot more in FO4 vs Oblivion.

An aspect of perf I think is underrated, loading times. Especially with Bethesda games crashing, I want to get back in the game ASAP.

I've repeated this many times to people I talk to: CPU performance often makes larger difference than SSD speed for game load times, AND improves FPS. A drive can be as fast as it wants, there is still overhead for IO requests and CPU still has to do something with the data it gets. While Skyrim runs very smoothly, starting the game up takes a looong time. Starting near vanilla game quickly matters because I intend to do Skyrim texture testing which requires loading the game tens of thousands of times with a script. The loading time decrease here is so much it looks wrong, but seems correct from my testing. It just takes way less time to load in Skyrim. This is why I lean towards RPL over 5800x3d: RPL is faster in Oblivion because large L3 is far less important than in FO4, and the single thread perf outside of L3 is far lower. 13600kf has like 42% lead in that department. So until I get evidence to contrary, I believe RPL will be superior to 5800x3d in game loading time. OFC, 7800x3d will not be behind by much, so it looks a lot more promising. Time will tell. Do I go for Arrow Lake or Zen5 3d? Too early to tell.



Let's look at FPS figures. Massive gains in performance. DXVK on my 5600x increased avg fps but increased stutters. For some reason that problem might just be gone on my 13600kf. I didn't change Oblivion or anything, just the CPU. So... I dunno what's going on. Followup tests must be done to ensure there is no performance degradation.


It ends up being like this, in terms of perf increase. Note that the FO4 figures is higher than my score in FO4 benchmarking thread. I have my own tests I prefer to use in evaluating perf for my own personal game. Massive gains in perf, especially for an upgrade that ended up costing me <$300.

I'm starting to think about 240hz SSE gaming, but FO4/Oblivion are still way behind.

 

[Det0x]

@BoredErica
I will make a entry
On what forum are most of the results from ?

 

[BoredErica]

@BoredErica
I will make a entry
On what forum are most of the results from ?

Hi,
Thanks for the interest. Most of the results so far are from Discord channels in modding channels around Bethesda games (mostly from Nexus Mod Author channel). So if you play a lot of Bethesda games modded you might recognize some of the names.

eg:
halgari is now an employee of Nexus Mods,
Hazakthemad has Skyrim mods,
I made a few FNV/SSE mods,
Audley is in Fallout New Vegas community,
Wall_Sogb is in FNV as well, same for VishVadeva,
etc