Tom’s Hardware tested this software version of BitLocker last year and found it could slow drives by up to 45 percent.
WTF‽ In Linux full disk encryption overhead is minimal:
While in pure I/O benchmarks like FIO there is an obvious impact to full disk encryption and other synthetic workloads, across the real-world benchmarks the performance impact of running under full disk encryption tended to be minimal
There’s like five million ways you can use disk encryption on Linux though and not all of them are very performant. So keep that in mind if you see other benchmarks showing awful performance (use the settings Phoronox used).
I suspect Microsoft made some poor decisions in regards to disk encryption (probably because of bullshit/insecure-by-design FIPS compliance) and now they’re stuck with them.
WTF‽ In Linux full disk encryption overhead is minimal:
https://www.phoronix.com/review/hp-devone-encrypt/5
There’s like five million ways you can use disk encryption on Linux though and not all of them are very performant. So keep that in mind if you see other benchmarks showing awful performance (use the settings Phoronox used).
I suspect Microsoft made some poor decisions in regards to disk encryption (probably because of bullshit/insecure-by-design FIPS compliance) and now they’re stuck with them.