I recently got back into mining on the ethereum network and experimented with a couple of mining pools in the process.
One thing I immediately noticed when taking a look at the log from the mining software (Phoenixminer), is that the amount of "new jobs"/ "new work" received from the mining pool has drastically increased since I stopped mining back in 2017.
The new work package from the pool (requested by "eth_getWork" or pushed from the pool) contains the current block header pow-hash (AKA "work") that is used by the miner to find a valid solution.
I don't understand why this hash would change more frequent, than the average time it takes, that a new block is found on a network-wide scale... I remember receiving much fewer new jobs from the pool back in 2017. I have not done an in-depth analysis of the current status, but just looking at the terminal, it seems I'm getting a new job more than once per second. With an average block-time of ~13 seconds this seems way too fast.
Follow-up question: Would this high rate of "job-switching" not increase the amount of stale shares submitted on a higher latency connection (especially on rental platforms)?