Hacker News new | past | comments | ask | show | jobs | submit login

Microkernels have always been slower than monolithic kernels, never faster.

There is zero chance that Zircon will be faster than Linux.

The classic example is Mach from CMU:

"When Mach was first being seriously used in the 2.x versions, performance was slower than traditional monolithic operating systems, perhaps as much as 25%... When Mach 3 attempted to move most of the operating system into user-space, the overhead became higher still: benchmarks between Mach and Ultrix on a MIPS R3000 showed a performance hit as great as 67% on some workloads."

https://en.m.wikipedia.org/wiki/Mach_(kernel)




Also an ancient example. Mach's latest version is ~30 years old (released 1994). See https://fediverse.blog/~/3542/the-three-generations-of-micro... for advancements since then such as an order of magnitude performance improve.


seL4 can barely do SMP. Modern hardware is all multicore. seL4 is not a performance winner outside of the embedded space.

https://docs.sel4.systems/projects/sel4/frequently-asked-que...


Stuck in the past?

That is why QNX and INTEGRITY have no sales.

Linux performance is so beyond what everyone else can achieve that is why nowadays we run it virtualized and pack applications into containers.


Apple just colocated the rest of the xnu kernel in the same address space as Mach and it seems to perform ok


Given apple is that true … not x vs y easily. But say like macOS and linux on arm and intel.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: