Topic on Talk:Grand Theft Auto IV

The source for this entry in the article says that we can use a value as high as "819200". Is this true?

Also... Does it matter if the player uses the latest version of the game (no downgrade)?

Considering this is tagged as an "essential improvement", we need to double-check.

GB, not GiB. The game couldn't use more than 4 GB of RAM even if it recognized values higher than 409600 in that file, since it is a 32-bit application. It does not matter if the game is up to date; I have checked the vanilla stream.ini file in the Complete Edition version of the game on Steam and the file comes with the values set to 204800.

It does come with such value. I wonder whether it really matters if players don't mod their game... 🤔

Yeah, think Bowi is right here.

Can't be too sure about it, maybe it didn't matter when the game released but it's actually useful nowadays when running the game maxed out and at very high resolutions. No real reason to remove or change that infobox anyway and it's been removed from the essential improvements section for a while now (which was the original topic of discussion here).

I don't remember moving it from "essential", but I wouldn't be surprise if I had done it. :P

Uh... Do you know whether the game does use more than 2GB (2 147 483 648 bytes 😆) when using the 409600 value? If not, I might test tomorrow to see if it appears in the Task Manager or another overlay... I only have GPU Tweak III which doesn't display SDRAM, but maybe the Game Bar does... 🤷‍♂️

If it really work, no matter the true benefit, maybe we could consider bringing it back up to "essentials". 🤷‍♂️

I did almost 6 hours test of various stream.ini settings on RGL version.
The most interesting results gave me:
virtual 102400
physical 409600
virtual_optimised 153600
physical_optimised 614400


Above, I went further with going with the scheme of better results when virtual value was set to be smaller than virtual_optimised, when it gave better results. A similar approach also seemed beneficial for physical memory.
Doing the opposite (smaller optimised value for virtual) gave worse results before.

It's pretty hard to get completely matching results in one pass in testing the same driving route, like I was doing, because traffic is pretty much random, and only weather remains more or less the same using the same save all the time. Benchmark is useless - average FPS is stupidly high and incomparable with results achieved in the heavy traffic. 2nd pass in benchmark always increases average FPS by ~1FPS.
So I also share settings I got good and better than stock results (204800 everywhere) out of many more tested if someone wants to experiment. Usually more than 1 pass for all settings used for tests.
virtual 102400, virtual_optimised, 153600, 614400 for both physical
614400 everywhere
409600 everywhere
default (204800 everywhere)
It still requires further testing since I couldn't use DSR/VSR this time to max out resolution, and I was stuck to FHD and max details. Used DXVK, ENB, Reshade, W10, 4 cores. Too low virtual memory set (e.g. 512 or even 512000) causes object disappearing (but 0 not, although decreases performance) hence it needs further testing for settings with virtual memory decreased. 102400 for virtual memory was a minimum value for FHD and maxed out details when there was no texture disappearing.

As for an influence on memory of these settings - it is minor. Depending on settings, RAM allocation difference is not bigger than 200MB.
The behavior is generally the same for all these settings (if only objects always appear).
The game can allocate up to 4GB of RAM, but when it's above 3900, it usually drops memory usage to around 3000MB, rarely with a freeze.
I was driving the wrong way on the highway for a long time to get faster memory allocation across the gameplay (in GTASA it was faster to fly by a Hydra to trigger memory related crash on big texture mods, but here I got only 3500MB max). When you press Print Screen, ENB+Reshade screenshots are being taken at the same time, and the memory usage usually jumps by 100MB, and then usually drop to 3K happens shortly afterwards, but the game itself also does it by itself, but rarely. Usually you won't get much above 3800MB (1.4GB used by the OS with the closed game, nothing really memory-consuming opened, 1.3GB used by the game process when minimised, but that value doesn't include pagefile). I never exceeded 9700MB for RAM and pagefile combined (memory compression on).
I have more than 4GB RAM, so it is not the system which limits the game from using more than 4GB in that case. It's 32 bit app limitation, and as someone checked headers in Visual Studio before, also patching the binary with Large Address Aware patch returns the same checksums as before, so it is already built-in with Large Address Aware in mind in vanilla game (RGL version was said to be the same as the others at some point)
Out of curiosity, I updated DXVK to 2.0 after getting the results above. Didn't do it before to avoid stuttering under specific circumstances described in 2.0 changelog, so no new cache wasn't needed to be rebuilt in such worst case. Old is almost 600KB, and new after one testing pass is 66KB. Not much.
Anyway, clean cache with 2.0 version provided the highest FPS in a certain place when I previously had maximum 47 (45-47) FPS. Few more. First time so much in that place.
Compared it with async version with async turned on in config, and it didn't get such an FPS neither with old nor new cache. Haven't experimented with dxvk.numasyncthreads yet. Tried out only maxFrameLatency = 1 which turned out to be beneficial, while VSync and triple buffering set in dxvk.conf wasn't, but actually prevented from one stutter, but I didn't perform additional pass to ensure it wasn't random.
Till the next time.

You are waaaay more patient than I am. I would have never experimented beyond "409600". 😅

By the way... I think we should do more test with the Rockstar Games Launcher version... Not everyone want to deal with third party tools, and even less will want to downgrade to an older version. Starting with myself. 😅

I just want to "play the damn game". 😅

Its not totally bad, but 1.0.4.0 is said to perform better. Once some YT link benchmarking this and newer version was posted, and I think someone deleted the reference or even whole paragraph saying it's not a real scenario, configuration or drivers shown are old etc. But boy, if even sh*y benchmark like here shows up differences, though it is much less demanding than the rest of the game, then maybe it's not that stupid? But yes, newer versions are more demanding due to new shadows, and I pretty much like them. Also 1.0.4.0 doesn't support episodes, but you can make episodes and main game standalone (probably downgrader irc, or at least installation of retail versions). But yeah, I'm also lazy these days and use RGL version.

I think that was indeed removed for having lots of caveats, eventually making the claim stated very unreliable. I think due to its unreliability it was deemed to speculative and thus removed. Also, if the decreased performance was caused by improved shadows being implemented, then maybe it is also not a huge negative? Would be a real negative in my opinion if the performance was noticeably decreased with no visual increase. I honestly think that for most people downgrading and having to separate the base game with the expansion is really too much of a hassle for most to only get back a little bit of performance maybe, while missing out on all the later bugfixes, removed GFWL etc. Still, downgrade info is still on the PCGW, so... should be ok right?

"But boy, if even sh*y benchmark like here shows up differences, though it is much less demanding than the rest of the game, then maybe it's not that stupid?" Btw, honestly don't know what you really mean here haha. But if the benchmark was at the time seen as too unreliable and there is no other similar more reliable proof backing up the made performance claim, then it of course shouldn't be mentioned in my opinion on the PCGW page as we could then be possibly aiding in spreading false information...

I played the game a bit today and did not manage to reach more than ~1700 MB of RAM. I decided to edit the stream.ini anyway and set everything to 409600 (I don't understand why "00" at the end, but that another topic. 😅)

Performances are OK, but I play at locked 30 FPS by capping with the NVIDIA Control Panel. 😅 4K with maxed out settings, except for the two details sliders which I left to default... I also disable the "Definition" option because I prefer the blurry DoF filter. Finally, Car Density up to 100 is not very hardware intensive...

Even with my current settings, trying to maxed out the details sliders is a very bad idea. I honestly think that the only "remaster" this game need is a patch that would make "maxed out" settings playable on 4-core CPUs. 😅 The game is already quite beautiful, IMHO.

But if you would unlock FPS you would likely be going over 2 GB of RAM right? Wouldn't surprise me that the two zeros at the end of the command would indeed be wrong and was just never corrected or noticed by anyone from the community haha as of course honestly many people seem to enjoy just throwing random commands against game like this without any knowledge about what it does haha.

I did my tests with unlocked FPS.

I now see this is also already stated in the note on PCGW: https://www.pcgamingwiki.com/wiki/Grand_Theft_Auto_IV#Use_over_2_GB_of_RAM "The game will only use up to 2 GB of RAM by default" So, likely only needed when using wrappers or extensive mods.

@Deton24 When looking here (https://www.pcgamingwiki.com/wiki/Grand_Theft_Auto_IV#Downgrading). You can see the claim is actually still there. I think you are confusing this situation with the note that was removed about the differences with the console version as the comparison video used to back those claims up were deemed to unreliable.

@Mrtnptrs "if even sh*y benchmark like here shows up differences, though it is much less demanding than the rest of the game, then maybe it's not that stupid"
I meant that benchmark in GTA IV usually shows much bigger framerate than actual game. I'm trying to say here that it might be much less sensitive to any detail changes, and possibly also patches (at least CPU-wise due to no heavy traffic) but iirc it showed some differences.
But knowing that shadows can be completely disabled, the comparison between various patches might be the most correct using that setting, but it still shows something about game performance if you know what are the differences between shadows and what fits to you. Because even lower shadows settings since 1.0.8.0 are altered vs 1.0.4.0, so it wouldn't be an accurate comparison, but it's also up to someone's taste if they like old shadows (and possibly better performance), but also lowered shadows in new patch can look better.
Here is decent comparison, but I think possibly better place would be near first safe house, when I really don't like how the old shadows looks, at least on certain level of detail. https://gtaforums.com/topic/975079-gta-iv-shadows-pc-version-1040-vs-consoles/?do=findComment&comment=1071629015

@Bowl 409600 "I don't understand why "00" at the end"
Because it is value in KB. Default is 204800.
If you set 512 instead of 512000 in at least virtual fields, it will cause object disappearing (even with 51200, but 76800 is fine but causes FPS drops, so it's still not enough - 102400 was the bare minimum for FHD maxed out to not cause object disappearing - also lowers values slightly lowers RAM consumption) And probably I get bigger memory consumption due to DXVK (that's normal behavior I noticed in the past with DX11 games too).
And yeah, thanks for clarification about the note, wasn't sure here what actually happened...