How To Boost Older GPU Performance, Since You Can’t Buy a New One

How To Boost Older GPU Performance, Since You Can’t Buy a New One

Right now, GPU prices are fluctuating between “insane” and “really insane.” The very cheapest cards may still be reasonably affordable, but everything above that has been shoved completely beyond reach for consumers. With prices like these, it’s impossible to recommend anyone buy a card for gaming right now — so what can you do to improve the performance of what you already have?

Before we kick off, a few caveats are in order. First, the tweaks we’ll explore, even taken in aggregate, will certainly help but are unlikely to work miracles. Performance tuning can improve frame rates and turn a game that’s chugging in places into one that runs noticeably and meaningfully better, but there’s no way to turn an HD 7970 or GeForce GTX 680 into a Vega 64 or GTX 1080. Second, because every game is different, some of these tips may be more or less effective depending on the title.

Two miscellaneous tips before we start: First, always make certain you are playing games in fullscreen mode, not windowed, and not borderless (windowed mode, without a visible border). Windowed mode is slower than fullscreen, in virtually every title, because your GPU is drawing both your desktop and the game itself at the same time.

Second, make sure you’ve dusted both your GPU and CPU. While unlikely to be a major slowdown cause, it’s far from impossible for a GPU’s performance to drop because the card is throttling under load. Dust is an excellent insulator and if you’ve got an older card, you’ve got no performance to spare in the first place.

Game-Specific Optimization

The first thing we recommend doing to squeeze a little performance out of an aging card is to experiment with game settings themselves. While most reviewers and gamers test titles according to presets (Low, Medium, High, etc), this is a practical time-saving necessity for the former and a matter of convenience for the latter.

Gamers generally know that certain features explicitly tied to AMD or NV GPUs (think GameWorks) can incur heavy performance penalties on other architectures, but the same can be true for other features as well. It’s not unusual for a game’s implementation of ambient occlusion, tessellation, or antialiasing to hit one company’s GPU harder than another, and this can even vary depending on GPU family. Yes, simply lowering game settings or resolution can improve frame rate, but toggling specific features can get you nearly the same results for a smaller reduction in performance.

The slideshow below shows the impact of different detail levels in the 2014 Shadow of Mordor, with the optional HD texture pack installed. You can see differences here, but they’re subtle. Medium detail definitely doesn’t pack as much eye candy as Ultra, but it still presents an attractive-looking and engaging game. Even turning off features like ambient occlusion or disabling tessellation doesn’t hurt the visual appeal much. Different games have different “break” points, some look dramatically different at Medium compared to High, while others “break” at Medium versus Low. Oftentimes the benefits of “Very High” or “Ultra” are quite small compared with their performance hit.

Don’t be afraid to climb into the Advanced Options and start changing various sliders. Because different GPUs take different performance hits from different options, you may need to do a little detective work, but it’s not unusual to boost frame rates by 5-10 percent just from small tweaks.

If you need a tool for checking performance in DX11 and earlier titles, FRAPS is still a good choice. (DX12 performance monitoring is more complicated and a bit beyond the scope of this article). It’s also a good idea to use presets as rough targets. If you know a game runs acceptably fast on “Low” and too slowly on “Medium,” don’t be afraid to use “Low” as a starting point for tweaking features upwards. You may find there’s more eye candy available to you than you initially thought. Treat the situation like a buffet dinner — if you don’t see much difference between “Medium” and “High” textures in a title, check to see if dropping to Medium gives you enough headroom to turn up a setting you do care about. You may find that expensive features add relatively little to the game, giving you space to enable other visual improvements you want.

Running at a lower resolution is also a useful way to spot-check performance improvements, though this is title-specific. First person shooters tend to scale better with resolution changes than a game like Civilization VI.

Driver Tweaks

Once upon a time (the late 1990s), it wasn’t uncommon for AMD and Nvidia to drop drivers that would improve game performance by 10-20 percent, even in older titles. Those days are mostly gone, provided you aren’t installing a launch-day update for a new game. But there are still a few ways you can tweak driver panel settings to squeeze out a little more performance.

How To Boost Older GPU Performance, Since You Can’t Buy a New One

On Nvidia cards (AMD has an analogous option in its own driver settings), you can set texture filter quality to “High Performance” as opposed to the default “Quality” option, force Ambient Occlusion on or off (if applicable), and force anisotropic texture filtering off.

Tweaking image quality in-driver won’t gain you a ton of performance, but it should improve things by a few percent. It’s also useful to check, to make sure you haven’t used global settings for one title and then forgotten to change them for another.

Check for Online Tweaks and Unofficial Patches

This one, again, is highly game-dependent. Some games don’t support modding or have small user communities. But in some cases, end users take it on themselves to create patches that fix various aspects of a title, including issues that impact performance. Games like Skyrim and Fallout 4 have often received unofficial patches that can boost performance or optimize game textures.

Overclock Your GPU (Desktop Only)

We’re only touching on the topic here, not diving into it, but overclocking your GPU is typically good for a few percent of performance at the very least. Again, your results are going to vary based on which card you have and how aggressively the manufacturer tuned it beforehand. Tools like MSI Afterburner can be used to overclock most GPUs.

Note: Everything written here applies to desktop GPUs. wfoojjaec does not recommend attempting to overclock a laptop GPU.

How To Boost Older GPU Performance, Since You Can’t Buy a New One

If you are planning to test a GPU overclock, we recommend small tweaks (no more than a 5 percent increase to memory or GPU clock at any one time) and slow going. Stay away from adjusting your GPU’s voltage until you’ve researched the topic and the acceptable range of your card, and don’t tell Afterburner to automatically apply settings at boot until you’ve confirmed the settings you actually want.

Expect this process to lead to a lot of reboots and/or lockups if your goal is to squeeze every last drop of performance out of your card — you’ll have to do some testing to find out where the right break points actually are, and it’s entirely possible to have ten games that run rock-solid on a GPU at one clock, but an 11th title that won’t run at anything but stock speeds. Such is the nature of this metaphorical beast.

If you aren’t overclocking by pushing voltages higher, the chances of damaging the card are fairly small, though we recommend you make sure the fans are dusted before you start pushing things. Most GPUs can typically a handle a 5-10 percent overclock without complaint.

Adding it All Up

None of these solutions are going to work if you’re trying to coax more life out of a 10-year-old card, but they can measurably improve your overall performance. If you can tweak game settings for a 7 percent boost, pick up a 7 percent overclock, and tweak driver settings for a further 3 percent, you’ve got a 17 percent overall performance improvement. In a game struggling to hit 30fps (let’s call it 28fps for fun), a 1.17x improvement gets you up to 33fps. That’s not a ton, no — but the corresponding impact of each additional FPS is larger the lower your frame rate is. The perceived difference between 28 and 33fps is much larger than the gap between 60 and 65fps, even though both are 5fps faster than the other.

And of course, these are conservative estimates. In certain titles and particularly if unofficial performance-enhancing patches are available, you might see significantly larger gains, especially if you’re already bumping up against the VRAM limit of your current GPU. Resolution cuts can help a lot in this case — don’t be afraid to drop to lower resolutions if you’re seeing heavy stuttering, especially if you have an older or lower-end card with a limited amount of memory.

No, tuning your existing card is not the same as having a reasonably priced GPU market for upgrades, and reasonable GPU prices are what we’d prefer. But if you’re stuck for the moment, hopefully this’ll help you improve the overall situation.

Continue reading

MSI’s Nvidia RTX 3070 Gaming X Trio Review: 2080 Ti Performance, Pascal Pricing
MSI’s Nvidia RTX 3070 Gaming X Trio Review: 2080 Ti Performance, Pascal Pricing

Nvidia's new RTX 3070 is a fabulous GPU at a good price, and the MSI RTX 3070 Gaming X Trio shows it off well.

Ryzen 9 5950X and 5900X Review: AMD Unleashes Zen 3 Against Intel’s Last Performance Bastions
Ryzen 9 5950X and 5900X Review: AMD Unleashes Zen 3 Against Intel’s Last Performance Bastions

AMD continues its onslaught on what was once Intel's undisputed turf.

Intel Is Spreading FUD About Supposedly Huge Ryzen 4000 Performance Drops on Battery
Intel Is Spreading FUD About Supposedly Huge Ryzen 4000 Performance Drops on Battery

Intel believes it has presented evidence that negates the value of AMD's Ryzen 4000 product stack. Intel is mistaken.

Current x86 vs. Apple M1 Performance Measurements Are Flawed
Current x86 vs. Apple M1 Performance Measurements Are Flawed

There's an intrinsic difference between x86 and ARM CPU designs that makes comparing performance difficult — and it didn't get noticed in the initial wave of coverage.