As others have pointed out, the point of zswap is to "swap" in and out of RAM.
I'd like to add that your premise is faulty anyway: since compressed objects are smaller doing I/O on them is faster. As long as cost of the I/O you saved is less than the cost of doing the compression, it's faster not slower. Transparently compressed I/O usually uses high-speed algorithms (LZ4, LZO, etc) to take advantage of this effect.
Also, when a machine is swapping you'll usually find that the CPU is near idle since useful work is blocked on the swap I/O. Spending a little CPU time to speed up that I/O is a win.
Re:Linux 3.11's feature set now confirmed
by TheCraiggers on Thursday July 18, 2013 @01:06PM
(#6064938)
An enlightening response. Thank you.
It's not the cost of the storage, it's how slow it to read and write to it. This is still going to be an order of magnitude faster than swapping to disk.
The point is to avoid writing to disk entirely by compressing RAM that may otherwise be swapped to disk when remaining RAM is low.
Even on tiny embedded devices, storage is cheaper than cycles IMO.