Jump to content

Why Do Newer AAA Titles Run Better Than Older or Indie Games on my System

JPStone

I suspect it may having something to do with the API and optimization but it still boggles me. I have a Ryzen 7 1700 OCd @ 3.9GHz and a PNY GTX1060 6GB XLR8 OC2 running on an msi B350M. My steam library is on a 256GB Western Digital Black M.2 NVMe SSD. I can play Fallout 4 with most settings on high and some on Ultra with occlusion, Godrays and all that good stuff enabled at a solid 75 FPS. However, I play Kerbal Space Program a game that, don't get me wrong looks great in its own right, but is not at the level of 3D graphic that a game like Fallout 4 is. Anyways, playing Kerbal I am at like 40-50 FPS with all settings on High Fullscreen @ 1920x1080 Res. I have noticed a similar trend playing XPlane. As much as everyone says how much of a beast Xplane is and how it just consumes PC resources like a glutton my observations have shown the opposite. 

 

I always keep tabs on my temps and usages while gaming and for both games its about the same. GPU revs up to Max boost as it should and CPU 30%-40% usage and plenty of memory to spare for all the games I play. 

 

I am just curious what does it come down to. What is it that makes the Nividia GPU I'm using able to handle a very 3D intensive game like Fallout 4 with lots of Shading, Ray Tracing and all the stuff they talk up when they pitch Pascal and CUDA but struggles to pump out enough FPS to make a game playable for the other titles I mentioned. Just to be clear when I do play Kerbal on my rig it is not unplayable even at 42 FPS if I didn't have the FPS meter on I would not even guess it was bellow 60 so game does stays smooth. Same with Xplane but the frame rates are still dramatically lower than Fallout 4.

Link to comment
Share on other sites

Link to post
Share on other sites

Because those older and indie games aren't optimized as well as AAA games for your hardware.

 

Older games are designed for older hardware and technologies so it makes sense that it won't work exactly as intended on a newer PC unless the game has received developer support. Indie game developers usually don't have the resources and/or manpower to optimize and/or troubleshoot for all the different configurations and setups a user could have while AAA games generally have the support of larger, experienced teams so they can support more hardware configurations quicker and use more advanced engines.

Edited by Sombra
Link to comment
Share on other sites

Link to post
Share on other sites

Now if I went out and bought a 1080 Ti or Titan X, if I could afford one :(, would it boost the performance across the board on all games or would I still run into the same issues of optimization??

Link to comment
Share on other sites

Link to post
Share on other sites

Just now, JPStone said:

Now if I went out and bought a 1080 Ti or Titan X, if I could afford one :(, would it boost the performance across the board on all games or would I still run into the same issues of optimization??

In theory, yes. But don't expect it to scale as perfectly as a better-optimized game would. And some poorly optimized or old games can sometimes have decreased performance if you are upgrading from a more supported platform/architecture. Check what kind of hardware is optimal for those games that perform poorly.

Link to comment
Share on other sites

Link to post
Share on other sites

Found this video on LTT

Makes Sense. with the shift to DX12 and Vulcan the newer GPUs have those APIs in mind. Ill have to go keep me up a 700 series card to swap out to play my older games. LOL. Thanks for the reply :)

Link to comment
Share on other sites

Link to post
Share on other sites

Because many older titles only scale up to 4 threads, some even only 2 threads. 

 

One of the reason why quad core i5 stay relevant for many years just until recently. 

| Intel i7-3770@4.2Ghz | Asus Z77-V | Zotac 980 Ti Amp! Omega | DDR3 1800mhz 4GB x4 | 300GB Intel DC S3500 SSD | 512GB Plextor M5 Pro | 2x 1TB WD Blue HDD |
 | Enermax NAXN82+ 650W 80Plus Bronze | Fiio E07K | Grado SR80i | Cooler Master XB HAF EVO | Logitech G27 | Logitech G600 | CM Storm Quickfire TK | DualShock 4 |

Link to comment
Share on other sites

Link to post
Share on other sites

It has nothing to do with cpu. And a 2011 game shouldn't max out that card. Bad nvidia drivers for older games, simple. 

 

Dx12 shouldn't be a issue, most actual AAA games don't use it

.

Link to comment
Share on other sites

Link to post
Share on other sites

4 hours ago, asus killer said:

It has nothing to do with cpu. And a 2011 game shouldn't max out that card. Bad nvidia drivers for older games, simple. 

 

Dx12 shouldn't be a issue, most actual AAA games don't use it

I second this, I've heard lots of rumours about bad backwards compatibility with nvidia drivers.

AMD Ryzen 5 5600 | MSI B450 Tomahawk | Corsair LPX 16GB 3000MHz CL16 | XFX RX 6700 XT QICK 319 | Corsair TX 550M 80+ Gold PSU

Link to comment
Share on other sites

Link to post
Share on other sites

" However, I play Kerbal Space Program a game that, don't get me wrong looks great in its own right, but is not at the level of 3D graphic that a game like Fallout 4 is. Anyways, playing Kerbal I am at like 40-50 FPS with all settings on High Fullscreen @ 1920x1080 Res. I have noticed a similar trend playing XPlane. As much as everyone says how much of a beast Xplane is and how it just consumes PC resources like a glutton my observations have shown the opposite. "

Computers. They compute things.

 

Fallout 4 computes an enemy, a health bar going down (most games cheat the "aiming" of AI, it literally just runs your health down until it gets killed, and pretends to be shooting/aiming at you). It draws a billion (exaggeration) polygons and pixels.

KSP computes every part and position and interaction of hundreds of rocket parts, while doing extremely accurate gravitational calculations. It draws a few thousand polygons and pixels.

 

Your GPU has no problem pushing tons of polygons. So F4 runs fine. KSP though makes supercomputers grind to their knees, as it is trying to compute real gravitational equations (and other stuff).

 

The type of problems KSP does are called "NP hard", and cannot be cheated without sacrificing accuracy or giving different results. KSP is a sim, so they want to keep the results accurate. Fallout 4 does not care how stupid on loggically broken it is, it also does not matter if a pixel is slightly out of place on a plasma explosion... so it "cheats" and takes many short cuts.

Link to comment
Share on other sites

Link to post
Share on other sites

 

18 hours ago, TechyBen said:

Fallout 4 computes an enemy, a health bar going down (most games cheat the "aiming" of AI, it literally just runs your health down until it gets killed, and pretends to be shooting/aiming at you). It draws a billion (exaggeration) polygons and pixels.

KSP computes every part and position and interaction of hundreds of rocket parts, while doing extremely accurate gravitational calculations. It draws a few thousand polygons and pixels.

 

Your GPU has no problem pushing tons of polygons. So F4 runs fine. KSP though makes supercomputers grind to their knees, as it is trying to compute real gravitational equations (and other stuff).

Okay that all makes sense but you would figure my CPU would be bottle-necked then if that was the case but looking in HWMonitor while running KSP the CPU usage is far from bottle-necked. I hit maybe 20%. I have not looked really at the per core usage so maybe as mentioned before the affinity is only 2 or 4 cores max so my 8 cores and 16 threads is not really helping me and over half of my potential compute power is just sitting idle. 

 

If that is the case you would figure that they would have found a way to fix that by now unless... Its all in their playbook for planned obsolescence??? LOL Thanks for the reply. 

Link to comment
Share on other sites

Link to post
Share on other sites

2 hours ago, JPStone said:

 

Okay that all makes sense but you would figure my CPU would be bottle-necked then if that was the case but looking in HWMonitor while running KSP the CPU usage is far from bottle-necked. I hit maybe 20%. I have not looked really at the per core usage so maybe as mentioned before the affinity is only 2 or 4 cores max so my 8 cores and 16 threads is not really helping me and over half of my potential compute power is just sitting idle. 

 

If that is the case you would figure that they would have found a way to fix that by now unless... Its all in their playbook for planned obsolescence??? LOL Thanks for the reply. 

That's the exact definition of a CPU bottleneck: two of your cores are pegged at 100% while the others sit there and do nothing, there's nothing you can do about it, except a CPU with better single core performance.

AMD Ryzen 5 5600 | MSI B450 Tomahawk | Corsair LPX 16GB 3000MHz CL16 | XFX RX 6700 XT QICK 319 | Corsair TX 550M 80+ Gold PSU

Link to comment
Share on other sites

Link to post
Share on other sites

18 hours ago, JPStone said:

 

Okay that all makes sense but you would figure my CPU would be bottle-necked then if that was the case but looking in HWMonitor while running KSP the CPU usage is far from bottle-necked. I hit maybe 20%. I have not looked really at the per core usage so maybe as mentioned before the affinity is only 2 or 4 cores max so my 8 cores and 16 threads is not really helping me and over half of my potential compute power is just sitting idle. 

 

If that is the case you would figure that they would have found a way to fix that by now unless... Its all in their playbook for planned obsolescence??? LOL Thanks for the reply. 

Most of the solutions to the n-body problem are not easy to do multithreaded and fast/on consumer hardware. I am aware of universe history simulations on sever farms that have 100s of CPUs or more... but KSP is not scaled for that, and uses 2 or 4 cores max... it's CPU speed limited. Fallout 4, is not... and can get by on very little, as it's all abstracted away...

Link to comment
Share on other sites

Link to post
Share on other sites

On 4/24/2018 at 2:34 AM, xFluing said:

That's the exact definition of a CPU bottleneck: two of your cores are pegged at 100% while the others sit there and do nothing, there's nothing you can do about it, except a CPU with better single core performance.

Oh, Okay. I was always under the impression a bottleneck was when you CPU was maxed out at 100% but still not able to push enough data to the GPU to keep it feed so then obviously your GPU is being starved by CPU. I guess its the same thing though seeing as you cant engage all cores for gaming so you can't truly hit 100% usage on the CPU even on super mega ultra setting. Hopefully soon they will find a workaround to limiting your CPU to 2 or 4 cores for gaming.

Link to comment
Share on other sites

Link to post
Share on other sites

3 hours ago, JPStone said:

Oh, Okay. I was always under the impression a bottleneck was when you CPU was maxed out at 100% but still not able to push enough data to the GPU to keep it feed so then obviously your GPU is being starved by CPU. I guess its the same thing though seeing as you cant engage all cores for gaming so you can't truly hit 100% usage on the CPU even on super mega ultra setting. Hopefully soon they will find a workaround to limiting your CPU to 2 or 4 cores for gaming.

They aren't limiting your CPU, any software has to be written to use cores. Can add months to development time, impacts games heavier than most. Not a problem for huge studios with large workforces, kinda an issue for smaller ones. Why for gaming single core is king, will be for years and years to come. 

If anyone asks you never saw me.

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×