What is the difference between -cpu host and -cpu max? It seems to me that they are the same

Or another way asked.... what is the difference between all supported host features vs all features supported by the accelerator?

I am assuming that the accelerator would also be the host features since you can only use one option here.

From man page:

x86 host KVM processor with all supported host features
x86 max Enables all features supported by the accelerator in the current host


KVM is a kernel module, not a hypervisor that you can use.I guess you are referring to the -cpu option in QEMU?

Here accelerator refers to KVM. You can use other accelerators by using the -accel option. The -cpu help flag gives you a pretty good explanation.

-cpu max means emulate a cpu that has all the features supported by KVM (limited by the set of features supported by your physical machines, of course)

-cpu host means emulate a cpu that is the same as your host cpu (limited by the set of features supported by the accelerator)

There isn't a whole lot of difference between the two, unless the accelerator you are using has very limited supported feature set.

It is recommended to just stick with -cpu host, which is also the default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.