i made this post, about the resources necessary to run Bitcoin Cash with the actual 32 MB blocks: https://www.reddit.com/r/btc/comments/mw753g/updated_post_of_why_bigger_blocks_will_not_make/?utm_source=share&utm_medium=web2x&context=3

So lets see how much resources Bitcoin Cash would need to use, having 1 GB blocks, not taking in account the hardware evolution, or software updates that we will have in the future, just using what we have today. ( to reach 1 GB, we would need to increase the blocks in 32 times, 32 * 32 = 1024, so i will take the numbers of the earlier post, and multiply them by 32.)

Bandwidth costs

Bitcoin Cash at the maximum blocksize today needs 853 Kb/s, 853 * 32 = 27,296, so with 1 GB blocks we would need about 27.3 Mb/s of bandwidth, 100 Mb/s internet is available to most people on first world contries today, so not even close to a problem.

Hardware costs

In the earlier post, i showed that a Raspberry Pi 4 can handle 256 MB blocks with 70% CPU load and 2 GB RAM usage, and it have a 1.5 GHZ quad core CPU.

so a modern PC with a Ryzen 3 3300 X, wich have 4 cores and 8 threads, and a 3.8 GHZ clock, should be 3x more capable than the Rasperry Pi 4, and handle the 1 GB blocks fine with 8 GB of RAM.

Storage costs

At this level, a HDD would not have enough IOPS to handle the amount of data real time, but a sata SSD could handle this easily, lets see how much the blockchain would grow:

1 GB * 6 * 24 *365 = 52.56 TB per year

Using the 2 TB Samsung 870 for $210 as example again, we would need of 27 of those SSDs every year, so that would be $5670 per year of storage.

But remember that not everyone needs run a full node, small stores and mercahnts could run a pruned node to verify payments, only big companies that uses Bitcoin Cash needs to run a full node, like exchanges, car manufactures and big stores, those guys can afford to run a full node. And even at $5670, running a full node would be cheaper in the long run, that to pay credit cards fees on the long run, also if we were filling 1 GB blocks, that would be about 3,000 TPS, so sites like read.cash would have a lot of revenue by showing ads in a banner, and the profits would be much larger than the costs to run a full node, everyone would benefit of this level of adoption, and the number of full nodes would be the same of today or increase.