Keybase and GitHub user computer_whisperer has been hard at work detailing out the mathematical reality of whether there is current risk to k32 posts with the rapid increase in plotting speed brought on by the “Mad Max plotter“. The Chia Network team will be addressing this tomorrow, but go ahead and read computer_whisperer’s theory here.
His claim is that there is simply not enough VRAM to hold the necessary data to form a plot on the fly, so a GPU will need to spill out to main memory and storage in order to actually perform the calculations and store the tables. I do not have the data at hand to check his work, but it does look correct to me, and even if it is off the bottleneck will always be the PCIE bus from the GPU to system memory until we start getting GPUs with enough memory on board to hold all that data.
As you can see, if the plotting device (gpu?) doesn’t have enough on-board memory to completely hold the data line moving along the bottom, then you have to put that data somewhere else.computer_whisper