Do we need branch prediction?

By: Geoff Langdale (, August 24, 2022 7:18 pm
Room: Moderated Discussions
Kara ( on August 24, 2022 2:07 pm wrote:
> Since like my second year programming I stopped writing branched code totally,
> few tens of thousands lines of code in, I haven't encounter anything that
> intrinsically requires speculative branch prediction, like logically.
> But I write only numes.
> Is there any program that can't be written without a branching that can't
> be rephrased (literally, all that is is rephrasing) into branchless?
> Mind risc-v don't offer BPU, there's extensions but not in the requirements.

if (--defcon == 1) {

Anything with a side effect in a branch is problematic (the printf example here is not great, as it just prints different things - much nastier would be read on one side and write on the other).

While launching missiles is an extreme side effect, my experience with trying to elide branches (which is extensive; I have written a bit about it on the semi-moribund, although mostly in SIMD) has convinced me that avoiding unpredictable conditional branches is only worth doing when the branch is (1) genuinely unpredictable, (2) actually frequently going in both directions and (3) when the amount of work being saved is significant.

Having a quantitative idea of the branch misses frequency, taken vs not-taken and the relative cycle costs of the work is critical - I've tripped over my fetish for writing branch-free code more than once with clever-clever solutions that are not better than just "eating the miss penalty 10% of the time".

If you're doing a fair bit of work on both sides of the branch, branch free alternatives can get expensive fast, especially as using a cmov isn't magically making your code "no longer bound by the condition" in terms of the data that results from the computation.
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
Do we need branch prediction?Kara2022/08/24 02:07 PM
  Do we need branch prediction?Rayla2022/08/24 03:04 PM
  Do we need branch prediction?Mark Roulo2022/08/24 04:25 PM
    Do we need branch prediction?Paul A. Clayton2022/08/24 06:28 PM
      Do we need branch prediction?Anon2022/08/24 06:52 PM
    Do we need branch prediction?rwessel2022/08/24 06:40 PM
    Do we need branch prediction?ananon2022/08/25 12:56 AM
      Do we need branch prediction?Carlie Coats2022/08/25 07:39 AM
        OT: You wanted rather than Mark Roulo2022/08/25 09:57 AM
          OT: You wanted "pre" rather than "code" rather than Mark Roulo2022/08/25 09:57 AM
            OT: You wanted "pre" rather than "code" rather than Carlie Coats2022/08/25 11:43 AM
    Do we need branch prediction?Kevin G2022/08/25 08:47 AM
      Do we need branch prediction?rwessel2022/08/25 11:10 AM
          Do we need branch prediction?rwessel2022/08/26 10:37 AM
  Do we need branch prediction?Geoff Langdale2022/08/24 07:18 PM
  Do we need branch prediction?Adrian2022/08/25 12:58 AM
  Do we need branch prediction?Brendan2022/08/25 07:54 AM
    Do we need branch prediction?Anon42022/08/25 11:45 AM
      Do we need branch prediction?anon22022/08/25 09:08 PM
  Predicable branching patterns (aykm; )Kara2022/08/27 01:00 PM
    Predicable branching patterns (aykm; )Jukka Larja2022/08/27 11:38 PM
    Predicable branching patterns (aykm; )Eric Fink2022/08/28 01:22 AM
      Predicable branching patterns (aykm; )Doug S2022/08/28 10:17 AM
    Predicable branching patterns (aykm; )---2022/08/28 02:16 PM
      Predicable branching patterns (aykm; )Anon2022/08/28 02:35 PM
  Q: Do we need branch prediction? A: We don't need it, but....2022/08/29 02:17 PM
    Quantum stuff.Kara2022/08/29 03:01 PM
      Q: Do we need branch prediction? A: We don't need it, but....2022/08/30 11:48 AM
    "Quantum Woo" (hint; that's NOT how it works) (NT)anonymous22022/08/29 04:19 PM
      Q: Do we need branch prediction? A: We don't need it, but....2022/08/30 12:14 PM
Reply to this Topic
Body: No Text
How do you spell tangerine? 🍊