- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi, what kind of latency vzeroall has compared to vzeroupper?
From a code density's point of view a single vzeroall is clearly better than a sequence of xorps to zero a couple of xmm registers, but i can't find any infos on latency. Is it "for free" like vzeroupper (zero latency according to intel's optimization manual) of microcoded?.
From a code density's point of view a single vzeroall is clearly better than a sequence of xorps to zero a couple of xmm registers, but i can't find any infos on latency. Is it "for free" like vzeroupper (zero latency according to intel's optimization manual) of microcoded?.
Link Copied
0 Replies
Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page