Short answer: You're perfectly fine running WCG on a device with SSD. Modern SSDs last many terabytes written, which usually translates to 5-10 or more years of daily use until that's even an issue. I believe the most disk-intensive WCG project is Microbiome Immunity Project, which has maybe 50 MB downloads, but meh. I've standardized on Samsung SSDs (TLC NAND) which last forever. There are many other disk-intensive use cases to worry about (e.g. downloading massive games, video editing scratch disk, DVR/PVR security camera storage, Plex server, other professional work like CAD, torrenting) that cause more wear over time due to writes. WCG projects are practically nothing. Can't speak for other BOINC or distributed computing projects. The upcoming climate change projects on WCG may have larger work units, but I have no idea at this point.

I'd say the most important things with SSDs is to make sure you have the latest firmware and make sure TRIM is enabled, which helps with the drive's internal wear leveling process. TRIM is usually enabled by default.

I recommend a UPS if it's a desktop PC in case the power goes out, reducing the possibility of data loss or corruption.


I'm 100% flash storage other than my NAS.

1. Samsung 840 EVO 250 GB (Laptop that sometimes gets shut down)
2. Samsung 850 EVO 1000 GB (24/7 Desktop)
3. Samsung 860 EVO 512 GB (24/7 Desktop)
4. Lexar 16 GB USB flash drive (24/7 Desktop)

In Options --> Computing preferences --> Computing tab, I have "Request tasks to checkpoint at most every" set to 900 seconds, which is 15 minutes. I have a Linux crunchbox running off 16 GB USB flash drive (Lexar brand, Micron NAND flash) that's set to checkpoint at most every 1200 seconds (20 minutes). You can put whatever you want here. 1800 seconds is 30 minutes. 3600 seconds is 1 hour. Just know if you have to power off or reboot your machine, you lose that much more time per work unit. I'm happy with 15-20 minute checkpoint intervals for my SSDs.

On the "Disk and memory" tab, I have "Leave non-GPU tasks in memory while suspended" ticked so in the event that BOINC pauses or suspends, it picks up right where it left off from RAM instead of reverting back to the last checkpoint on disk. I also have "Page/swap file use at most" set to 1%, which is the lowest integer value it takes.

The Windows devices are set to never Hibernate (can disable Hibernate from an Administrator Command Prompt with the powercfg -h off command then reboot), and I've disabled Sleep mode in both BIOS (if possible) and in Windows Power Options set to "Never." In the Advanced Power Options, I double-checked that "Sleep after" is set to Never and "Turn off hard disk after" is set to 0 minutes or disabled or greyed out, since this option is meant for spinning disk HDD.

If using Windows 10 (maybe also in 8.1), in "Defragment and Optimize Drives," it's set to optimize the SSD once weekly, and Windows is smart enough to run the TRIM command instead of actually defragmenting, which should never be done on flash storage. If using an older version (such as Windows 7), verify that Defragment is disabled so Windows doesn't wear the heck out of the SSD.

Thanks for that awesome information.

Just to add to the excellent advice above.

I have 3 small capacity SSDs at 16-32GB and they have accumulated 214/179/82 years of run time with no issues at all. I usually bought brand names that I know like Kingston/Transcend and let them go for it.

So, as others have pointed out, you should have no issues at all with a modern SSD.