One of the most interesting things that your PC does is take half of the available RAM and turn it into shared GPU memory. But does this improve your system performance, or does it negatively affect performance?
In this article, we’ll find out what is shared GPU memory and what you can do to boost your system’s performance.
What Is Shared GPU Memory?
Shared GPU memory is a form of virtual memory that your GPU or graphics card uses when it runs out of dedicated VRAM or video memory. How much your shared memory is allocated is dependent on half of the amount of RAM you have. If your PC is running on 32GB RAM, then your shared memory is going to be 16GB, at least.
Operating systems will use RAM as a source of shared graphics memory because it’s faster than SSD or HDD. But RAM is still slower than VRAM, which modern graphics cards have. After all, a discrete graphics card could be running a GDDR6 VRAM and while RAM could still be behind with DDR4 or DDR5.
But your operating system or OS won’t use shared memory until your graphics card is out of VRAM, and even if this does happen, there’s no degradation in performance even if your video game or app is tapping into this shared memory on a filled GPU memory.
In fact, using shared GPU memory will allow a game or program to run because excess data that couldn’t be stored in VRAM will be stored on RAM. If your PC didn’t have a shared memory, your game or app would crash as soon as the graphics memory is full.
When RAM is used for shared GPU memory, it doesn’t mean that the shared memory is only for graphics card’s use. The OS will still use the shared memory to run other apps when it needs to, so you don’t have to worry about half of your sytem RAM being utilized as a back up for your dedicated graphics card.
But this is not the case for integrated GPU. Shared GPU memory refers to the maximum amount of memory that your GPU can use. Integrated GPUs don’t have dedicated memory that they can use to run apps or games. To make this possible, integrated graphics cards will use system RAM or system memory. For integrated GPUs from Intel and AMD, their shared GPU memory is half of the RAM size.
What’s The Main Difference Between Shared GPU Memory and Dedicated Memory?
Dedicated GPU memory refers to physical memory installed on a discrete graphics card. These are the high-speed memory modules, such as GDDR or HBM, and are placed near the GPU’s chip. Dedicated GPU memory is used for games, rendering, and other graphics-related tasks.
On the other hand, shared GPU memory is sourced or taken from your RAM. Unlike dedicated GPU memory, shared GPU memory is virtual and not physical. This is only an allocation of memory on your PC’s RAM.
If your system has an integrated GPU, meaning it has no discrete or third-party graphics card, then it won’t have its own dedicated memory. Integrated GPUs will take instead from system RAM.
Should Shared GPU Memory Be Increased or Decreased?
If you have a dedicated graphics card or a discrete graphics card, you don’t have to worry about how much shared memory you have. Shared GPU memory will only take up half of your system’s total RAM. If you have 8GB of RAM, 4GB will be allocated for shared GPU. Increasing or decreasing your shared GPU memory won’t really have an effect on performance but it doesn’t mean you should do either too.
For integrated GPUs, people were probably told to increase the amount of “reserved” or “dedicated” iGPU memory through BIOS and tweaks in registry settings, but this does nothing. In fact, performance may degrade at this point. This is because Intel chipsets and AMD chipsets have BIOS settings where system RAM is allocated exclusively for GPU usage. If you increase your graphics memory, you’re actually decreasing the amount of system RAM on your PC.
Games and apps might warn you about low memory size when you’re using iGPU but you can just ignore those messages. Your OS will still use RAM for video memory if you do run out of VRAM. But note that at this point, you’ll have weird visuals on games such as missing textures or slow-loading graphics.
Some games might not even run if you have an iGPU memory that’s already pre-allocated, so this might be the only time when increasing it via BIOS settings is ideal. Just remember to set them back to default once you’re done.
Wrapping It Up
Shared GPU memory is sort of a backup VRAM for when you’re about to get tapped out on your discrete GPU’s VRAM buffer. But with the modern graphics cards that we’re seeing right now, playing games, doing graphic design, or even video editing doesn’t even take up a lot of memory.
However, laptops might have a different issue. Unless it’s a gaming laptop, the GPU runs on half of the system’s total memory. Most laptops run on integrated GPUs, so the issue is the system’s RAM will be half occupied already. The big difference between an iGPU and a discrete GPU is that the latter will always have more VRAM and can handle more demanding tasks.