What is Branch Prediction? | Hacker News Reader

What is Branch Prediction?

stackoverflow.com

69 pointsralphchurch28 comments

The author of the answer has set multiple world records for calculating digits of pi. I highly recommend browsing through his answer history. You can learn more from his answers than from entire computer science classes.

http://stackoverflow.com/users/922184/mysticial

For those interested in a Stack-Ed perspective of Mysticial's l337ness (top 5 tags):

C++: http://www.stack-ed.com/#tag/c%2B%2B/user/922184/Mysticial

Performance: http://www.stack-ed.com/#tag/performance/user/922184/Mystici...

Language-Agnostic: http://www.stack-ed.com/#tag/language-agnostic/user/922184/M...

Optimization: http://www.stack-ed.com/#tag/optimization/user/922184/Mystic...

Branch-Prediction: http://www.stack-ed.com/#tag/branch-prediction/user/922184/M...

Why do people keep posting this every month???

People keep upvoting it every month.

Karma whoring.

Nailed it.

I notice ralphchurch's name is green. That means recently registered, right?

I believe so. And they have posted 5 stackoverflow questions in 11 days.

IIRC, this is called Power Leveling.

the funny thing is I tried this YESTERDAY even this same post, I wonder how he managed to post a duplicate...

and I did it to test if it would work, and it did not worked, and now it worked.. :/

now I guess that spamming SO does work after all.

Thanks for asking the question. I noticed the color was different, but wasn't too sure which color it was because of my colorblindness..

Does anybody actually care about karma on HN? If so, why?

Same reason as why we want money? We see it as a proxy for power.

Why do people want power? Evolutionary reasons.

Probably for men we see it as a way to reproduce a lot for women it's probably seen as a way to ensure their children are successful and reproduce.

It gets a bit depressing if you think about it to much. Just enjoy life and accept things.

But we don't just see money as a proxy for power. It is the closest proxy for power there is. HN karma, on the other hand, is barely a proxy for power at all.

Posting SO questions is a double whammy because you level up HN karma and you snag some SO badges as well.

For what it's worth, I've never seen that before, and it was really interesting. I'm glad someone posted it :)

Sounds like a job for Discourse: better dupe detection, usefully surfacing "evergreen" content to noobs, and anti-karma-whoring.

Well, it is a very good answer. Most likely people admire that.

Also, I've been here for three years and I don't recall seeing it before, so it can't be posted that often. You must just happen to catch it a lot. I wouldn't have known it was a dupe if you hadn't pointed it out.

Question: can a processor have two pipelines? When there's a branch it would be able to start loading both branches and just switch the pipeline when the final branch is determined.

It might still need to flush the pipelines when there are several branches in succession but you'd still be able to use branch prediction to predict the two most likely branches, which would help tremendously.

I think it's possible, but I think it's more efficient to use a completely separate...thread?...if you have a 2nd core. (Or if you have lots of empty space bubbles in your pipeline, you can just try forcing 2 threads in there, aka Hyperthreading)

That makes sense for most processor architectures. But there are exceptions. The Cell Processor (Sony PS3 Processor) has 8 RISC SPUs. These processors do not have branch prediction logic. This is because they were designed for computationally intensive graphics kernels which are useful for Sony PS3 games. There is a huge performance intensive to avoiding branches on these processors. So many times both routes are taken and then the appropriate value is chosen using a mask instruction. This does not stall the pipeline on the processor.

This is basically what computers do, it's called speculative execution. Modern computers actually have several pipelines, each for a different function. This is what makes a computer "superscalar". Basically, the CPU tries to keep the pipelines of all of the execution units filled at all times. If it's able to, it will schedule both sides of a branch to be executed, and flush out whichever half wasn't used.

speculative execution and superscalar are independent concepts. While superscalar does effectively mean multiple pipelines, you don't have to be superscalar for speculative execution. You can reduce pipeline stalls in a single pipeline by "speculating" right as well. The concept that GP is talking was a feature of of Itanium.

http://en.wikipedia.org/wiki/Explicitly_parallel_instruction...

" The EPIC architecture also includes a grab-bag of architectural concepts to increase ILP:

Predicated execution is used to decrease the occurrence of branches and to increase the speculative execution of instructions. In this feature, branch conditions are converted to predicate registers which are used to kill results of executed instructions from the side of the branch which is not taken. . . . Multi-way branch instructions improve branch prediction by combining many alternative branches into one bundle. "

Generally speculation only executes one side of the branch. Speculative threading has been proposed using SMT to execute both sides, but usually you're better off trusting the branch predictor and using threads as non-speculative state.

Speculation down both sides is independent of SMT and if I remember correct there were proposals to do this - especially where the branch predictor "knew" it wasn't doing a good job on a branch (no point going down both sides of a 99.9% predicted branch).

The problem is that you rapidly run out of execution resources (what if there are more branches to follow, for example) and you will burn lots of energy executing things that don't happen (not to mention speculatively loading things that aren't needed, etc). I remember seeing some papers on the limits of ILP and I think this was one of the 'extreme' ideas (speculation down multiple paths).

To execute both sides, the fetch unit needs to have multiple program counters which is essentially a form of SMT.

Sure, but it defeats the purpose of branch prediction. If you have a high confidence of which branch will be taken you don't want to waste execution units. The high level of data dependencies between branch points is another obstacle. It isn't as simple as hyperthreading.

How many time must this get re-posted before people get that it's a good Stack Overflow answer? Does HN really churn that many users each month?

A beautiful example that shows how useful genuine understanding of computers can be, even to high-level developers.

Reply on news.ycombinator.com