I'm curious about global vs per-force as well. I can imagine a PvP server where the scenario does all-forces sounds while each team does their own per-force notifications.

The video itself is, of course, mind blowing.
Yes, that's why I asked. I would imagine that people are able to come up with ...

I assume sound can be set to be force-based too and not global for everyone?


Also, now finally that Darude video made with colored lamps can also have background music playing, all in-game

I think they might be putting in too much in the same release. They should probably split it up a bit. Multithreading+bunch of minor optimization+nuclear+new science+combat balancing+hd sprites+fluid wagons and even more. Just one of those categories would be enough for a new release.
Most of ...

What the belt really gains is in the dependency analysis.
Isn't that analysis done on compiler side? If so, what stops from using it on regular x86?
There is absolutely nothing preventing this from occurring as this is what the hardware out-of-order unit does.
That's done in hardware in x86 ...

Which is exactly why you go out with changes to information in good time, like before that general discussion post, the info that the customers expected was about right, was around february 2017.
A looong time ago I think they said they were *hoping* to get something by then but they said they can ...

ratchetfreak wrote:What the belt really gains is in the dependency analysis.

Isn't that analysis done on compiler side? If so, what stops from using it on regular x86?

What are you talking about? Loads of information has been disclosed in FFF articles.

Devs have also stated they have no clue when will it release besides that it's almost certain it won't be before around two months.

You aren't running over the same data at the same time
I explicitly said same data at *different* times.
That can't be avoided.
You can if you don't want to force double-buffering of everything :)
Devs have already talked about their solution a little bit that avoids doing that by calculating a ...

Note: 8 cores also have 8 times as much L1 cache and 2-4 times as much L2 cache as a single core.
Running same data over several cores at different times in different threads also means more memory accesses to get the same data. Devs already stated how that's currently one of the biggest problems ...

They can compile and run real programs and get good results doing so
How good?
Intel just bumped their absolute maximum (best case) dispatch rate to 8 instructions per cycle (per core) as of Haswell. (It was 6 before that.) That's not exactly close to the Mill's 33.
There is little point in ...

You should put that stuff you wrote there in a separate article, possibly even non-FFF one. I'm sure a LOT of people would be interested in reading it but only a few ever see it due to it being hidden in a thread full of mental mastrubation :)
Yeah, we're thinking of that. For these questions were ...

TL;DR: Factorio has multithreaded update since around October 2016.

[a wall text made of pure awesomeness]

You should put that stuff you wrote there in a separate article, possibly even non-FFF one. I'm sure a LOT of people would be interested in reading it but only a few ever see it due to it ...

It depends on your memory access pattern. If you are streaming through memory serially, using every byte, then the number of CPU cycles available per byte of memory is actually very low. If you are hitting random memory addresses and only looking at one byte at each address, then yeah, then you're ...

Note that the fact that you can make something run in parallel does not mean you have created good (highly optimized) code. And lack of optimization (e.g. just copying all of memory, only handling the minimum "unit of work" size rather than special-casing center vs edge and tiling for best cache ...

Not exactly a fix for collecting the artifacts but how I've solved it is to build a belt outside my defensive walls that brings stuff to my artifact storage area. Sure, it'll still mean that you'll have huge fields of artifacts laying on ground but at least those fields won't expand infinitely ...

The core count keeps climbing. Some years back I had to buy an expensive dual socket board to get SMP. Nowadays my laptop has 8 cores.
I got my first "dualcore" in around 2005 in form of hyperthreaded Pentium4. Next was quadcore q6600 about 2 years later. The next CPU I got was less than a year ...

Try a different solution. Instead of each of the 108 inserters checking the warehouse if they can take some items go the other way and add more clock states.

Clock state 0: all inserters figure out how many items they can potentially grab (evaluate signals, set filters)
Clock state 1: the ...

My point is that the game is very well optimised but it seems to spend a lot of CPU cycles on entities that are inactive because they don't have power.
This is your gut feeling or did you test it somehow?

I don't really think that dealing with drain power usage is all that expensive in the big ...

Not exactly a fix for collecting the artifacts but how I've solved it is to build a belt outside my defensive walls that brings stuff to my artifact storage area. Sure, it'll still mean that you'll have huge fields of artifacts laying on ground but at least those fields won't expand infinitely since ...

Xeanoa wrote:Wait, there is a campaign beyond the tutorial?? o.o
Several hundred hours, and I didn't know...

Well, to be honest, I've not played them myself but there are 4 different "campaigns". I'm not really sure what their win condition is, though.