With respect to Lifespan -
For Hard drives the answer is indeterminate - ie there is no well established link between the number of writes and the the lifespan.
For SSD's, drives have a limited number of write cycles - the number depends on the make of the drive, technology used and how big the drive is (SSD's have hidden space for "over-provisioning" to increase the lifespan) The rewrite cycles typically varies between 3000 and 100000 - and unless you are stress testing the ssd to failure the drive will most likely become obsolete before it dies.
Speed and Efficiency
These are largely 2 sides of the same coin. When you are writing, things slow down - this is a lot more marked in hard drives then SSD's because SSD's are more responsive and have a much lower seek time - they don't need to wait for a spinning disk to get to the right position.
The impact of fragmentation on an SSD is very small (but, surprisingly, not non-existant), while it can be significant on a hard drive. Thus the way the files are deleted and rewritten are important to the efficiency and speed - particularly on a hard drive. Also, on a hard drive, for bulk writes, the outer tracks of the drive are typically around twice as fast as the inner tracks (because more data flies under the head), so as the disk gets fuller, your speeds decrease.
Preparing for failure
In both cases, its a good idea not to rely on the drives and to have backups. RAID will also help reduce the risks of drive failure. There are a couple of important considerations though -
SSD's are about 10 times as reliable as hard drives, but when they fail they normally fail catastrophically - ie suddenly and you can't get any data off them.
HDD's are a lot more likely to give warning that they are about to, or have started to fail, giving the opportunity to recover some/most data from them. (They can also fail catastrophically due, for example, to a failure on the control board - but they are a lot less likely to).
In both cases drives have health information which can be read (Google S.M.A.R.T). This normally gives an indication of how far through its useful life an SSD is, and may - or may not - provide meaningful information about an impending failure of a hard drive. (According to a very recent study by BackBlaze Seagates fail faster then other drives, but are generally better at predicting that - that said most peoples reading of the stats say that the particular model of a drive is more indicative of likely failure then the brand - although my takeaway is that Hitachi drives are a lot more reliable on the whole.)
Writing 50 GB to any of the described devices has not effect on speed, lifespan, or efficiency. – Ramhound – 2016-02-18T20:22:18.877
HDD's are pretty hardcore, not sure how it would affect the mechanics. SSD's on the other hand literally have a limited number of writes to each spot. – Cand3r – 2016-02-18T20:24:45.387
Why was this downvoted ? – davidgo – 2016-02-18T20:31:13.510
@Cand3r Any storage device can only sustain a limited number of accesses during a finite amount of time, simply because a finite time and a non-zero access time puts an upper bound on how many accesses can happen during that amount of time. That said, SSDs very rarely break due to flash failure and the severity of flash per-cell overwrite failures is often grossly overstated for modern SSDs. Compare Are SSD drives as reliable as mechanical drives (2013)? and How damaged is my SSD disc.
– a CVn – 2016-02-18T20:34:37.443Are you sure you mean 50 GB (50 - 100 times)? That's not a large volume of data. Perhaps you mean 50 TB? 50 PB? – ChrisInEdmonton – 2016-02-18T20:40:03.540
Well, large for my purposes at least. – Rob – 2016-02-18T20:41:07.537
@MichaelKjörling, I was not talking about time at all. The media of a SSD will not withstand the same number of writes as a HDD given a perfect environment, as in, controllers, write heads, cases and whatnot don't break in this 'simulation'. For the question tho, it suppose it doesn't matter, that large write will only write to 1 spot, not the same spot every time. Also, whats the word? Transistor on an SSD? Magnetic bit on HDD, can't remember for the life of me. – Cand3r – 2016-02-18T20:47:19.433
1@davidgo - Its a vague question about the lifespan of various media, with no specifics, there are several dozen questions that answer the questions proposed by this question (in my opinion). – Ramhound – 2016-02-18T20:52:55.340
@Cand3r SSDs don't map identical LBA addresses to identical flash cell locations every time. If they did, areas like that used for the file system would wear out in practically no time. Look up "wear leveling". – a CVn – 2016-02-18T21:07:18.693
@MichaelKjörling, you missed my point again but i worded it poorly, and I know what wear leveling is, he's going to be writing large volumes of data, so those aren't going to be wearing out 1 cell more than any other was what I was trying to say. – Cand3r – 2016-02-18T21:10:27.783