As a deck owner, it’s not that powerful, you’re never going to drive it at full settings on most modern games, so the size of the vram does not matter all that much.
Also, the GPU already does the cache work itself if the vram is full anyway, and GDDR is much faster than regular DDR, which is why you see stuttering on 8gb GPUs when texture resolution is pushing the limits.
With the spot price if GDDR6 modules it’s frankly disappointing to only see 8gb.
Plus, add in the fact that FSR/DLSS models take up valuable vram size to work for framegen and stuff, it reduces even more the availability for actual data.
Would it be safe to assume their processor/gpu magic that brought us the deck has advanced enough since then to compensate?
I’m no hardware guru, but wouldn’t it be possible to use a swap file or other methods to simulate extra ram if optimized and efficient enough?
As a deck owner, it’s not that powerful, you’re never going to drive it at full settings on most modern games, so the size of the vram does not matter all that much.
Also, the GPU already does the cache work itself if the vram is full anyway, and GDDR is much faster than regular DDR, which is why you see stuttering on 8gb GPUs when texture resolution is pushing the limits.
With the spot price if GDDR6 modules it’s frankly disappointing to only see 8gb.
Plus, add in the fact that FSR/DLSS models take up valuable vram size to work for framegen and stuff, it reduces even more the availability for actual data.