Halo Infinite Patch Coldly Kicks GPUs With Less Than 4GB Of VRAM To The Curb
Thankfully, the issue is not a significant problem. All modern desktop and mobile gaming GPUs over the past 5 years have been operating with 4GB of memory at a minimum. Even when 3 GB GPUs were current, it wasn’t a popular memory capacity. The only time 3GB was standard was during the early 2010s when the graphics industry was moving beyond 2GB frame buffers to 4GB and 6GB as the choice for flagship hardware. This was particularly true of AMD, which favored 3GB heavily during the 7000 series, R9 200, and R9 300 generations.
NVIDIA did have a few 3GB GPUs during this time, including the 3GB GTX 580 and 3GB GTX 780. But overall, the company preferred to stick with even numbering, using 2GB and 4GB for most of its GPUs. The last time NVIDIA used 3GB of VRAM was in the GTX 1060 3GB, and it was due to unusual circumstances. The GPU’s awkward 192-bit memory configuration means that six memory ICs had to be used, instead of four or eight. This limited NVIDIA’s memory choice to just six 512MB ICs in its cheaper sibling of the GTX 1060 6GB. Capacities between 512MB and 1GB - like 768MB, did not exist at the time.
Even when the GPU was new, the 3GB capacity was already problematic with GPU reviewers. The low capacity prevented the GPU from utilizing high-resolution textures in some games and caused micro stuttering in many AAA games as resources were swapped between VRAM and system RAM to counter the low capacity. Now, in a world where 12GB, 16GB, and 24GB video memory are commonplace, the loss of interest in 3 GB support is understandable. If anyone’s still running one of these GPUs, now is a great time to upgrade your graphics card. Though, we hope 343 Industries will give us an official explanation for the GPU requirement changes, just to make sure this 3GB limitation is not a glitch.
