Just shot thru 341 work units all errored out in less than 2 seconds.



Error was reported as shown here as read from null "0x0" trapped in milkyway.exe

After all 341 works units crapped out, a single seti task got to run and correctly reported the problem:



No other seti tasks ran as the error program was smart enough to stop. Be nice if the MW programmers here could implement better error checking. I uninstalled whatever device drive windows 10 had updated to and put the older AMD driver back in and MW is working fine (At least until the next windows update).