Comment #6 on issue 4310 by [email protected]: 64 bit performace is
slower than 32 bit
https://code.google.com/p/v8/issues/detail?id=4310
For example because all pointers are twice the size, 64-bit platforms
consume more memory, which puts more load on caches and memory bandwidth.
On the other hand, in some cases additional instructions available in the
64-bit instruction set provide tangible speedups.
V8 also behaves a little different, e.g. "Smis" are represented differently
on 64-bit platforms (32-bit payload, 32-bit shift; whereas 32-bit platforms
use 31-bit payload, 1-bit shift). The 64-bit way of storing Smis is faster
in some cases, but requires an additional instruction in other cases.
Specifically on arm64, the stack pointer has funny alignment requirements,
which means V8 must do more work in some situations.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings
--
--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
---
You received this message because you are subscribed to the Google Groups "v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
