Forged Alliance Forever

Re: CPU & Latency

2012-08-20T08:06:22+02:00

Mr-Smith wrote:
na i know im the slowest most time and gpu should not be a problem (ati mobility radone hd5850) well i think its the i5.

idd it's very unlikely the GPU.

I'm a bit surprised though that you lag "so hard" , it's a castrated i5 but still, I was under the impression it would perform 4v4 and such quite ok. Still have a check in the console, ren_networkstats and check in the columns your sim speed, compared to the other players in the game.

Statistics: Posted by -_V_- — 20 Aug 2012, 08:06

Re: CPU & Latency

2012-08-19T18:23:34+02:00

-_V_- wrote:
It takes way too long to get a game and it's not getting better each day, sadly.

indeed

Statistics: Posted by Raging_Squirrel — 19 Aug 2012, 18:23

Re: CPU & Latency

2012-08-19T17:09:55+02:00

na i know im the slowest most time and gpu should not be a problem (ati mobility radone hd5850) well i think its the i5.

Statistics: Posted by Mr-Smith — 19 Aug 2012, 17:09

Re: CPU & Latency

2012-08-19T13:46:08+02:00

when the game runs really slowly:

just search on the forum, how to show the simspeed (speed of simulation - only needs cpu power) of all the people in your game.
short version: open console, enter: "ren_shownetworkstats" and push enter. in the table you see a column with "maxsd" (or similar). this is the fastest simspeed, which can be handled by that guys cpu. it ranges from -10...+10. if you are always by far the slowest (and have values below -3), you should only play games with a lower unit limit and less AIs.

---------------------------------------------------------------------------------------------------------------------------------

when the game lags (e.g. 5-10 fps):

lower the graphics options and resolution of your game. if it does not work try this:

try to use "core maximizer". download it from gpg forum, put it in your c:/programdata/faforever/bin folder, run the file (cmax.exe) and launch your replay. the fps will be better, but the simspeed a little worse. if you look for a convenient way to start core maximizer with FAF, search on the forum. i have already described a solution.

Statistics: Posted by eXcalibur — 19 Aug 2012, 13:46

Re: CPU & Latency

2012-08-19T07:09:46+02:00

5v5 , so i guess a lot of units and quite a nice size map, right ?

Intel® Core™ i5-460M Processor (3M Cache, 2.53 GHz) => not enough to handle that much, correctly , beyond a certain point.
Roughly only the 2 first cores of the CPU is used for FA. My i7 2600k lags too in 4v4, and it already reached -10 , so don't be surprised if you can't run positive sim speed all the time with your i5 . Oh and when it happens , if I don't have core maximizer on, the first core is running full, if i have core max on, the load is spread on the other cores (the first one is not 100% loaded anymore) but it WON'T run faster. Pretty useless unless.

But before you go too harsh on your comp, make sure you're the faulty one when there's super lag. There may be people with slower comps than yours, therefore they'll be the bottleneck.

Other stuff to check, that is a mobile proc if I'm not mistaken, so it may also be your GPU that is the bottleneck (if it's not a dedicate GPU).

Statistics: Posted by -_V_- — 19 Aug 2012, 07:09

Re: CPU & Latency

2012-08-18T21:43:21+02:00

anyone know what causes lag? im running i5 460m watching 5v5 replay and lagg as hell... not sure why just have 50% cpu load...

Statistics: Posted by Mr-Smith — 18 Aug 2012, 21:43

Re: CPU & Latency

2012-08-18T19:43:03+02:00

FYI, I got bashed a lot throughout my FA years, I played countless times versus lappen's crew for instance or other preformed super teams, and that with just med people on my team. Did we lose yeah ? But then we got better.

I really started to try to get better at "air" because of one jerk that I "hated" quite a lot as a player and who bashed me over and over. Gave me motivation, then I got better and beat him. Ethom if you're there

Ask Mize how many times he got completely bashed with his old crew. He got better, and not by playing with players with lower experience.

Well, anyway, point is, I can be wrong but I feel most players in the community don't try very hard to get better now. I really want new (interesting) oil to shed haha.

Setting numbers as in the rating when you host (and kick people) is not really good. Go with noobs, med and good , well fair enough. Epic raping is not useful either. But numbers, that's overdoing it by a large margin.

I wish new or med people would spend more time in improving by watching more replays and testing the shit out of this game in sandbox, instead of crying "balance" whenever wherever. I got so tired of this shit that I stopped playing or close, and I would now very likely play a stupidly stacked game against me, as long as it's not ranked. I'm not willing to sacrifice my "soon to come back" t4 avatar yet , but play for play why not. It takes way too long to get a game and it's not getting better each day, sadly.

Oh and for the funny part, YOU going to a med game would ruin it as well *IF* I follow your logic. You don't belong to the games with the very best ones , as you implied, fair enough, but you CLEARLY don't belong either to the med ones, really not.

Shit I lost track again, sorry. Lag is freaking annoying, although a bit is part of the game and HAS TO BE tolerated. But then I'm afraid that if straight numbers are put on each one's head, the divisions we already see because of the rating, will grow even bigger and that would be a shame. Usually the lagging player will accept the request of the others gamers to just leave.

Ultimately if he doesn't , TML, and friendly fire are still options

Statistics: Posted by -_V_- — 18 Aug 2012, 19:43

Re: CPU & Latency

2012-08-17T16:56:10+02:00

Im sorry V but if you go in a mid game you will ruin it, the choice to group within skill level means that mid players like myself can play games without constantly being bashed by the "elite" on FAF.

As for lag, well if im not happy with the players in the lobby I leave, im not wasting 2-3 hours on a Setons.

It is harder to get games I agree but I would rather wait and play a decent lag free game than a lagfest or a bash, which I only seem to get one or another lately.

Statistics: Posted by Casimus — 17 Aug 2012, 16:56

Re: CPU & Latency

2012-08-13T13:10:30+02:00

rabidradish wrote:
the elitist jerks.

On that note, I would say this part of the community is really not the worst anymore. I am, well was part of this elitist jerk group, but it was merely a division between experienced and not experienced players. Now you have all sort of retarded separations : between 800-1700 , 900-1300, below 1500 , bla bla. And don't you dare be a bit out of those bounds, most of the time you'd be kicked at sight.

Completely sad. Considering that the peak period of games is getting reduced, it's even worse.

Statistics: Posted by -_V_- — 13 Aug 2012, 13:10

Re: CPU & Latency

2012-08-12T14:22:02+02:00

Anaryl wrote:
10 is max on that benchmark? That would be a beast of a machine. I've got an i5 2500K running at 3.3Ghz (stock speed), and the benchmark gave me 4.4.

That's strange because I got 4.4 and I'm running an i7 3930 with a 7970 with 32gbs of ram.

Pretty useless benchmark if it cant tell between an i5 and an i7.

Ultimately the idea of discriminating between more players in the community is kinda retarded. FAF is not in a position to deny players, and it's only getting worse. The peak time is getting shorter and the off peak is getting longer. If you really want to put the writing on the wall, this is a great way to go about it.

Like I said, more than 2 cores is a waste for this game. 2500K does 3.7GHz single core, 3930K does 3.8. So that's 100MHz extra but it's likely that HT has slight performance hit, so yes scores should be the same.

And it's not about discriminating against players. I like to see as many players as possible, as long as they stick to playing on maps their system is suited for. If anything, it's likely that the mega laggers have scared more people away than the elitist jerks.

Statistics: Posted by rabidradish — 12 Aug 2012, 14:22

Re: CPU & Latency

2012-08-12T11:01:01+02:00

its imposible to get score 10 in this. Benchmark tests how many time took to your CPU to calculate many ingame calculations, like physic, move, quaternion rotation, project ray, detect colision etc. and some simple math. Slower cpu will need more time to complite benchmark.

So this benchmark test only how fast your cpu is, not how many memory etc. you have.

Statistics: Posted by ozonex — 12 Aug 2012, 11:01

Re: CPU & Latency

2012-08-12T10:52:18+02:00

Anaryl wrote:
it is likely the case that the benchmark does not care for bigger L3 cache, a bigger IGP and hyperthreading.
the last 2 are also of no value to FA, so the difference between i5 and i7 should not be significant.

So, then how would you get a 10 on this benchmark. There isn't exactly many higher spec'd PCs. At least from a benchmarking point of view, it should be able to tell the diff between a mid range gaming rig and a high end workstation. That's kinda the entire point.

the tool measures single thread performance. it does not test the GPU. it does not use any significant amount of ram.
so its suited for measuring FA performance (assuming it does some typical calculations that are used in RTS).
preliminary evidence suggests that a highend PC would not run FA much better than a mainstream PC.
This is for example true for the size of RAM since FA is a 32 bit application and for the GPU, since anything above N frames per second is indistinguishable to humans.
The only difference for simspeed is probably due to higher clockspeed and a bigger turbo boost.
So by not measuring the actual speed of processors the tool gives a more accurate result.

Statistics: Posted by rootbeer23 — 12 Aug 2012, 10:52

Re: CPU & Latency

2012-08-12T08:29:44+02:00

Anaryl wrote:
10 is max on that benchmark? That would be a beast of a machine. I've got an i5 2500K running at 3.3Ghz (stock speed), and the benchmark gave me 4.4.

That's strange because I got 4.4 and I'm running an i7 3930 with a 7970 with 32gbs of ram.

Pretty useless benchmark if it cant tell between an i5 and an i7.

it is likely the case that the benchmark does not care for bigger L3 cache, a bigger IGP and hyperthreading.
the last 2 are also of no value to FA, so the difference between i5 and i7 should not be significant.

Anaryl wrote:
Ultimately the idea of discriminating between more players in the community is kinda retarded. FAF is not in a position to deny players, and it's only getting worse. The peak time is getting shorter and the off peak is getting longer. If you really want to put the writing on the wall, this is a great way to go about it.

[/quote]

IMO less than 10% of players are a concern on maps larger than 10x10/8 players.
i fail to see how it is r... to care about game quality. Whether you discriminate in the lobby or make people leave in a laggy game is really not much different in the end, only with less drama.
It also saves us the embarrasment of publicly believing FA should be able to run on .
So should FAF not provide a means to discriminate and deny certain people from ruining games? What have you gained if you can play more games before the sun sets, but a few of those will be a waste of time?
Then there is still the other 50% of games which are on 10x10 or smaller and open to everyone.
It is a useful feature that only affects a minority of players on certain map sizes and then only with a bunch of game hosts that care about it. No biggie. Just use brain for maximum effect.

Statistics: Posted by rootbeer23 — 12 Aug 2012, 08:29

Re: CPU & Latency

2012-08-12T07:47:14+02:00

Holy shit.

In total agreement with Anaryl.

Damn I thought it was super easy to make a decent benchmark tool. Guess I was mistaken

Statistics: Posted by -_V_- — 12 Aug 2012, 07:47

Re: CPU & Latency

2012-08-11T14:24:54+02:00

ozonex wrote:
ok, for now:

best score i seen is: 6.0
for my c2d e6600 3.4ghz i have: 0.0
worst score i seen is: -4.5

so, for now i can tel if someone have more than 2.0 can play setons easly.
2.0-0.0 can make problems in games 6v6 and bigger.
Score less than -1.0 is just bad

gives the result with precision of 0.5, but this is not important, becouse you see that one cpu is or not good for big games. I use for benchmark gameengine, to simulate real calculations.

BENCHMARK

I get an error message trying to run the benchmark (Application validation did not succeed. Unable to continue)

Casimus wrote:
My CPU is a AMD Phenom II x6 2.8ghz.

-_V_- wrote:
[whispers]say to Cas he was running -3 5 mins ago[/whispers],

Oh, the irony. No offence to any fanboy, but AMD is really bad at this game.

VIP wrote:
got 5.5
tbh its useless to make benchmark because of 1 map, every c2d 2.4 can handle every most playable map (5x5 10x10)

There are more laggy maps, Shards comes to mind.

Want to play this game without lag on 20x20/40x40 maps? Get i5 2500K or i5 3570K and overclock to 4GHz+. And it still might dip -1.

Apart from that, this game is completely single thread limited, more than 2 cores is a waste.

Statistics: Posted by rabidradish — 11 Aug 2012, 14:24