How is simspeed lag in 2018?

[SpoCk0nd0pe]

Huh, didn't know there could be autistic threads outside the balance forums.
Good job guys, really exceeding my expectations.

Yeah, like some RT troll^^

I hope one day we have an AI that sweeps through the internet and shuts those posts up with ‚Deleted: Idiot‘

Since the topic kind of goes that direction: Is there a good frame time analysis comparing the 8700k and the 2700x with tight b-die timings? Mindblanktech has a 7700k comparison showing the 2700x with better frame times across many modern titles. But I suspect the 8700k might do better with its larger L3.

Has anyone tested the validity of CPU scores compared to replay times? I would be very interested to see if that Intel dual core really does perform so well in SupCom. I cannot believe the sim isn‘t L3 latency bound, but would be amazed to be proven wrong

[E8400-CV]

250 is not a good cpu score.
170 is what a 6 year old intel (3770k) cpu will get with no overclock

Thank you!

170 on stock 3770K? Huh?

I was about to respond to all of that nonsense again, but quite honestly I think there is no hope. Pretty much only talking down on other people's rigs and distracting from the main points I made, that he has no real response to. Every time I make a point there is a dice rolled if there is an actual response to my argument lol.

Bad decision on my part to waste my time on a guy that does not even play the game.

The best part is that he himself stated that is CPU was the slowest in game. The irony is taking over.

Haha, you really have too much time. You get paid by the AMD propaganda department?

So much talk and this guy still didn't reach the conclusion of hosting his own lobbies.

No point if it just gets f*ckd by the -U bullshit.

[E8400-CV]

Huh, didn't know there could be autistic threads outside the balance forums.
Good job guys, really exceeding my expectations.

Yeah, like some RT troll^^

I hope one day we have an AI that sweeps through the internet and shuts those posts up with ‚Deleted: Idiot‘

Since the topic kind of goes that direction: Is there a good frame time analysis comparing the 8700k and the 2700x with tight b-die timings? Mindblanktech has a 7700k comparison showing the 2700x with better frame times across many modern titles. But I suspect the 8700k might do better with its larger L3.

Has anyone tested the validity of CPU scores compared to replay times? I would be very interested to see if that Intel dual core really does perform so well in SupCom. I cannot believe the sim isn‘t L3 latency bound, but would be amazed to be proven wrong

Years ago there was a topic but the test was with a fake Sandbox game.

i7-7700K vs 2700X OC'd on all sides with modern titles that are CPU intensive... yeah... that's an easy call. The i7-8700K will still beat that since it does away with the core shortage.

In FAF though, the extra cores do exactly nothing.

[IceDreamer]

The 8700K is currently the fastest gaming CPU. It is almost certainly the fastest CPU for FAF, along with the other Coffee Lake architecture CPUs.

That being said, the Zen+ architecture has really closed the gap a great deal. Yes, Coffee Lake is faster, but it's not faster enough to make a noticable difference in many titles, provided you're not aiming for 144Hz+ refresh. A number of deep dives have shown that with custom memory timings and overclocks on both platforms, R2 closes even further to get within 5-3% of the 8700K.

So the 8700K is going to be faster than a 2700X in FAF. However, unless you play games with other people who also have only 6th and 8th generation parts, it's not going to be slow enough for anyone to notice. The older FX and Sandy Bridge chips of other players are still going to be the bottlenect.

As a related note, I have a question to anyone reading this thread: How much of a simspeed improvement, in %, is enough to rationalise temporarily breaking most unit/game (non-UI) mods? 10%? 50%? What would you say?

[FtXCommando]

Better to ask the dudes that play phantom and other common mods. The people that browse the forum are gonna say 1% because most are not really affected by game mods being broken.

[E8400-CV]

[...]

As a related note, I have a question to anyone reading this thread: How much of a simspeed improvement, in %, is enough to rationalise temporarily breaking most unit/game (non-UI) mods? 10%? 50%? What would you say?

If it's just temporary, it suggest is doesn't have to break in the first place. Why would it break? There is a beta version, right? So anyone can test his / her mod and adapt it to the new release. Or am I missing something?

But yeah, any simspeed improvement is welcome

On the Phantom thing; that mod especially can benefit from a simspeed increase

[Wesmania]

From my experience managing client releases, I'm pretty sure 95% of modders won't start on that until everything is merged and released to stable breaking everything for everyone, just like client themes :/

[TheKoopa]

My i3-4160 hyperthreaded @3.6ghz gets a CPU score of 170 and I am literally never the slowest in a game, do what you want with that info

[ZLO_RD]

My i3-4160 hyperthreaded @3.6ghz gets a CPU score of 170 and I am literally never the slowest in a game, do what you want with that info

no one said that you need 4 cores for FA... you just need an normal intel CPU xDDD
your core i3 is same thing as i7 just with less cores afaik.

here is a Pentium G3420 3.2 ghz 196 score proof kinda
no so far away from core i3

[CookieNoob]

As a related note, I have a question to anyone reading this thread: How much of a simspeed improvement, in %, is enough to rationalise temporarily breaking most unit/game (non-UI) mods? 10%? 50%? What would you say?

Who is going to fix most (thats a lot) unit/game mods? You dont even have time to maintain the FA repo (which is what you are sitting in the council for). If there is no one to fix these mods afterwards, then temporarily means permanent in this case. This is not a game in its early active development stage with lots of active modders, but a 10 year old title with thousands of mods.

Since you probably didnt do only a single optimization, but a lot of different things, how about you merge those that dont break most mods and see how much performance gain can be reached with that. Then you can check how much additional gain you can get with the stuff that breaks everything and use that as a comparison. Optimization is never an "all or nothing" thing.