How is simspeed lag in 2018?

[uzurpator]

theoretically. now does "core jumping" deliver sufficient cooldown to translate into performance?... that is the question.

Short answer - No.

Long answer - core jumping lowers performance, because each time a process is expulsed from execution and entire cache needs to be emptied as well. Performance hit might be slight, might be substantial, depending on the task at hand.

That being said - desktops are usually not affected by thermals for it to matter. Laptops, to the contarary. The cooling tower on my CPU ( i7 5820 ) weighs as much as an entire laptop. Portables are designed to be used in bursts, because there is only so much cooling to be gotten from such a small format. That is because performance per/watt/clock is essentially the same for desktop and laptop variants for a given family of cpus.

[SpoCk0nd0pe]

This game uses just two threads. One for simulation, one for graphics. Having "all cores at 30%" is just the OS jumping threads to different cores. From that you can establish that limiting and FA to two cores and OCing the beejeezus out of them is the best way to get maximum simulation speed.

theoretically. now does "core jumping" deliver sufficient cooldown to translate into performance?... that is the question.

As far as I know the goal is to reduce power use (leakage) and wear. There shouldn‘t be much of a performance hit. I think I remember benches saying it‘s below 1% but I‘m too lazy to search atm.

[Franck83]

140 CPU rating on a Desktop i5 4670k. No overclocking. DDR3-2133 Dual Chanel. CPU set on High Priority.

[tatsu]

that's the theory but what I've found is that compute micro-pauses are not controlled by OS or motherboard but by the CPU itself.

it halts calculations for a millisecond then resumes and does this at a frequency that varies according to temperature. this is actually incredibly efficient at shearing off a couple degrees.

And apparently the CPU never really stops doing this. whether it's at 80°C or 30°C although as I said it does do it a whole lot less at temperatures like 30°C.

I have to wonder if temperature is able to comfortably radiate away into the inactive dies then out into the radiator if the CPU would judge the inner environment less hostile?

[E8400-CV]

So it's 2019 now.... and still the game is ruined by lagger.

Just had a game that ran mostly at -3.... due to one fucker. When will there be a kick option for such assholes?

And even if death... they don't leave.

[Gruffman]

So it's 2019 now.... and still the game is ruined by lagger.

Just had a game that ran mostly at -3.... due to one fucker. When will there be a kick option for such assholes?

And even if death... they don't leave.

I'd settle for a working FOE system so you can easily spot scumbags who ruin games that you never want to play with again

[E8400-CV]

Sounds nice, but that only keeps out repeat offenders. The influx from new laggers still gives 25% of games running too slow.

[relent0r]

Yea I hate to admit it but I brought a new mb/cpu just for this game. Had an amd 8120 for years and it was fine with pretty much all games except faf. Got an intel 9600K as it had really good single thread performance as far as bang for buck goes.

[E8400-CV]

Yea I hate to admit it but I brought a new mb/cpu just for this game. Had an amd 8120 for years and it was fine with pretty much all games except faf. Got an intel 9600K as it had really good single thread performance as far as bang for buck goes.

Yes, but as long as there is one guy with crap PC in your game.... you get nothing for doing the right thing. And FAF just allows them.

I just was in a game... -3 at 20 minutes. "Professor_van_M".... and *ding* I remembered his nickname. He is doing that same stunt since forever. We need a way to get rid of these players w/o "report and we will take a look".

[relent0r]

Yea I hate to admit it but I brought a new mb/cpu just for this game. Had an amd 8120 for years and it was fine with pretty much all games except faf. Got an intel 9600K as it had really good single thread performance as far as bang for buck goes.

Yes, but as long as there is one guy with crap PC in your game.... you get nothing for doing the right thing. And FAF just allows them.

I just was in a game... -3 at 20 minutes. "Professor_van_M".... and *ding* I remembered his nickname. He is doing that same stunt since forever. We need a way to get rid of these players w/o "report and we will take a look".

So your thinking they should allow us to restrict benchmark results when creating custom games? similar to player ratings. Or maybe a warning when players have higher than xxx when joining a map that is 20km or over.