Are multi player games slower than single player games?

I

ihira

Gawd
Joined
Sep 30, 2007
Messages
904
Hi folks, I just recently upgraded from a X1900XT to a G92 8800GTS.

The 8800GTS 512MB is awesome, I play at 1280x1024 4xAA my fps in Bioshock, Oblivion, whatever has increased dramatically anywhere from 1.8x~even 2x faster and it runs constant silk smooth 60+ fps.
However I didn't see that much of an increase in TF2. I still dip in 30's sometimes. Even CS:S can get into the 40's.
I was kind of hoping 60+ fps with those titles too seeing they're graphicaly lighter than Bioshock or Oblivion but is this normal? Are multi player games more CPU dependant than single players?
 
For instance, I have a GTX and a Pentium D, and TF2 still drops into "chunky" range.

This says nothing about the graphics power of the card, it only tells me that my CPU cannot keep up with feeding enough data into the graphics card to chew up. The efficiency of the GPU depends on the speed of the CPU (to a point). Hence the importance of building a well rounded system.
 
So are multi player games like TF2 CSS, more reliant to the CPU?
Because HL2ep2 on the same Source engine gets much higher fps than those two.
Never thought I'd say this but I guess a E6600 isn't good enough...
 
ihira said:
So are multi player games like TF2 CSS, more reliant to the CPU?
Because HL2ep2 on the same Source engine gets much higher fps than those two.
Never thought I'd say this but I guess a E6600 isn't good enough...
Click to expand...

Well I think your CPU should be good enough. I don't know why your performance would drop so low. You don't have VSYNC on or anything do you?
 
When I play TF2 I drop into the 30s and 40s too... I thought it would be smoother than that. Even when I lowered the settings, it was still running almost the same.
 
Koko56 said:
When I play TF2 I drop into the 30s and 40s too... I thought it would be smoother than that. Even when I lowered the settings, it was still running almost the same.
Click to expand...

I haven't tried lowering settings for my low framerate, but I'm reading this problem all over the place, and it appears to be a game performance issue that valve is aware of.

I run at max settings (I think with Vsync), so I'm sure that plays a part in my case ;)
 
ihira said:
So are multi player games like TF2 CSS, more reliant to the CPU?
Because HL2ep2 on the same Source engine gets much higher fps than those two.
Never thought I'd say this but I guess a E6600 isn't good enough...
Click to expand...

Overclock your CPU. an E6600 not being overclocked is like throwing an 8800Ultra in your grandmas machine. So much wasted power.
 
Sr7 said:
I haven't tried lowering settings for my low framerate, but I'm reading this problem all over the place, and it appears to be a game performance issue that valve is aware of.

I run at max settings (I think with Vsync), so I'm sure that plays a part in my case ;)
Click to expand...

Hopefully it'll get fixed. :D
 
peacetilence said:
The sever you play on can limit performance if it's being strained.
Click to expand...

Sorry, but at no point in the rendering process does your PC stop and go "Hey server, how many FPS should we give this guy?"

The server does not affect your FPS. The gameplay occurring on that server does.
 
ihira said:
Hi folks, I just recently upgraded from a X1900XT to a G92 8800GTS.

The 8800GTS 512MB is awesome, I play at 1280x1024 4xAA my fps in Bioshock, Oblivion, whatever has increased dramatically anywhere from 1.8x~even 2x faster and it runs constant silk smooth 60+ fps.
However I didn't see that much of an increase in TF2. I still dip in 30's sometimes. Even CS:S can get into the 40's.
I was kind of hoping 60+ fps with those titles too seeing they're graphicaly lighter than Bioshock or Oblivion but is this normal? Are multi player games more CPU dependant than single players?
Click to expand...

Whoa there! You're looking at a whole different set of dynamics with on-line play. The NIC and router come into play here, bigtime!. Got an el-cheapo 2nd tier NIC and router? Expect lower performance. Even some namebrand routers handle packets poorly in online gaming. Router options, configurations, encryption, they all play a role as well. Wired, NIC to Router, or wireless - another factor. Ram can also seriously effect performance - shouldn't even consider competitive online play with 1 Gig or lower. Could you give us a little more information about your NIC and router?

Scratch the comment about memory - you sittin' at 2 GB already.
 
I always thought TF2 lagged a bit because of the sheer amount of stuff that is happening constantly. Lots of players, lots of shooting, lots of physics...
I never really thought networking configs can impact FPSs on online games. That, I didnt know and am interested what factors come into affect.

FYI I'm on a Linksys WRTG54L wired. no NIC, its onboard gigabit (Do people buy NICs these days?).
 
All Core 2 Duos really light up once you get past 3.0GHz, try it out, especially for a CPU-dependent apps like source games. Also, as stated, server/networking/NIC also play a part in your FPS/lag/"chunky" playing. Really, the entire system gets stressed because so many different components are being pushed at the same time.
 
Wow, I thought I was the only one having the problem with TF2 and the fps drops, I run it at 1900x1200 with 4x AA and 8xAS.
 
Ejay said:
Whoa there! You're looking at a whole different set of dynamics with on-line play. The NIC and router come into play here, bigtime!. Got an el-cheapo 2nd tier NIC and router? Expect lower performance. Even some namebrand routers handle packets poorly in online gaming. Router options, configurations, encryption, they all play a role as well. Wired, NIC to Router, or wireless - another factor. Ram can also seriously effect performance - shouldn't even consider competitive online play with 1 Gig or lower. Could you give us a little more information about your NIC and router?

Scratch the comment about memory - you sittin' at 2 GB already.
Click to expand...

Packets != FPS. Don't confuse the issues.

Packets can result in lag, not low FPS

i.e. you can get 2 fps with a 2ms ping (you're talking to the server without delay/network lag), and you can get 200fps while lagging network-wise, they're two totally different things.
 
Valve is working on the TF2 performance issues:)

Update, so you know where we're at:

We found several bad perf cases in the tests we ran toward the beginning of the week; a couple of them are being worked on at the moment:

1) Bad framerates when shooting water or a wall with a shotgun or the minigun when you are standing right next to that water or wall. I'm working on that now; and am partially done.

2) Bad CPU performance in some unexpected places. This would tend to affect people with lower-end CPUs, even if they have high-end GPUs. We're still investigating this one. My suspicion currently is that this may be related to the drop reported on Nov 21.
Click to expand...
 
Yeah.. theres definitely something wrong with your computer. Theres no way an 8800GTS 512 should drop into the 40's on one of the most efficient engines ever created, Source. You should be getting over 200fps in the source benchmark with that setup. Even on a server with 40 people it should run smooth as glass. There has to be something wrong. My X1900XT never lagged in source with that resolution and max settings when I had it.
 
S3th13 said:
Yeah.. theres definitely something wrong with your computer. Theres no way an 8800GTS 512 should drop into the 40's on one of the most efficient engines ever created, Source. You should be getting over 200fps in the source benchmark with that setup. Even on a server with 40 people it should run smooth as glass. There has to be something wrong. My X1900XT never lagged in source with that resolution and max settings when I had it.
Click to expand...

It's not source as a whole though, it seems to be specific to TF2 as the poster above you has shown.
 
Sr7 said:
Sorry, but at no point in the rendering process does your PC stop and go "Hey server, how many FPS should we give this guy?"

The server does not affect your FPS. The gameplay occurring on that server does.
Click to expand...

there is a server side setting for FPS that can result in lower choke for people in cs:s, it may apply in TF2.



since getting my 8800Gt stock and a e4400 @ 3.1ghz, i run 1920 x 1200 with ingame all high, and i recall AA @ 8xca? and AF on 8x also and keep around 60fps on all maps i have played.
 
Sr7 said:
Packets != FPS. Don't confuse the issues.

Packets can result in lag, not low FPS
Click to expand...

This really isnt true. When you consider that your CPU has to chew up all the incoming and outgoing information between the server and your computer you can occasionally see a drop in performance in multiplayer games compared to single player.

Play some FPS games that have single player maps like the multiplayer maps like the BF series and youll see a difference in FPS.
 
S3th13 said:
Yeah.. theres definitely something wrong with your computer. Theres no way an 8800GTS 512 should drop into the 40's on one of the most efficient engines ever created, Source. You should be getting over 200fps in the source benchmark with that setup. Even on a server with 40 people it should run smooth as glass. There has to be something wrong. My X1900XT never lagged in source with that resolution and max settings when I had it.
Click to expand...

In the Source benchmark test in CS:S (this is what youre talking about right?) I got a average 235 fps.
TF2 is another story. On 32 man servers at Dustbowl opening or Gravelpit B where all chaos is happening, it definately sinks into 40s, sometimes even 30s for me according to FRAPS. And it seems like I'm not the only one with G92 8800GT(S) or GTXs reporting this issue...
 
pfunkman said:
This really isnt true. When you consider that your CPU has to chew up all the incoming and outgoing information between the server and your computer you can occasionally see a drop in performance in multiplayer games compared to single player.

Play some FPS games that have single player maps like the multiplayer maps like the BF series and youll see a difference in FPS.
Click to expand...

incoming and outgoing information is a ridiculously minute amount of CPU, do the research if you don't believe it. Run around in an offline map and record your FPS. Then go run around in an online server (so long as it's the same visuals without people in your view). You won't notice a difference.
 
Sr7 said:
incoming and outgoing information is a ridiculously minute amount of CPU, do the research if you don't believe it. Run around in an offline map and record your FPS. Then go run around in an online server (so long as it's the same visuals without people in your view). You won't notice a difference.
Click to expand...

I have already done these tests and can tell you without a doubt multiplayer does perform slower in most games than single player.

The reason i used the BF series as an example is because the single player maps are identical to the multiplayer maps and there is a noticeable difference between the two in performance. Even CS me and the wife play with bots for fun just her and i VS the bots and i can tell you there as well you will get better FPS than the online multiplayer.
 
pfunkman said:
I have already done these tests and can tell you without a doubt multiplayer does perform slower in most games than single player.

The reason i used the BF series as an example is because the single player maps are identical to the multiplayer maps and there is a noticeable difference between the two in performance. Even CS me and the wife play with bots for fun just her and i VS the bots and i can tell you there as well you will get better FPS than the online multiplayer.
Click to expand...

Well then you and I will have to disagree. I'm sorry, but network traffic simply cannot account for 2/3 to 1/2 the FPS.
 
Sr7 said:
Well then you and I will have to disagree. I'm sorry, but network traffic simply cannot account for 2/3 to 1/2 the FPS.
Click to expand...

Im not saying it can account for that much im simply saying it can impact game performance.
 
pfunkman said:
I have already done these tests and can tell you without a doubt multiplayer does perform slower in most games than single player.

The reason i used the BF series as an example is because the single player maps are identical to the multiplayer maps and there is a noticeable difference between the two in performance. Even CS me and the wife play with bots for fun just her and i VS the bots and i can tell you there as well you will get better FPS than the online multiplayer.
Click to expand...

I actually get lower FPS with bots in CS:S... I think this maybe because the CPU has to manage all the bots, whereas online people just run around and send co-ordinates.
 
ihira said:
...it definately sinks into 40s, sometimes even 30s for me...
Click to expand...
I'm running it at the same settings as you, and I don't think mine dips any lower than that on an A64 3000+ / x1950pro 256.

So you can all stop blaming the CPU...
 
LuminaryJanitor said:
I'm running it at the same settings as you, and I don't think mine dips any lower than that on an A64 3000+ / x1950pro 256.

So you can all stop blaming the CPU...
Click to expand...
FPS is going to be server AND situation dependent, you can't overall compare two CPUs.
 
Mr. K6 said:
FPS is going to be server AND situation dependent, you can't overall compare two CPUs.
Click to expand...
Why not?

The OP sees this problem regularly. So it's fairly safe to assume that in the 15 hours I spent playing TF2, I encountered something at least approaching those conditions. And trust me, anything which can choke an E6600 is probably going to leave you with single-digit framerates on a 3000+.

And yet I'm getting the same performance as the OP, at the same settings. The problem lies elsewhere.
 
You missed the point of my post. Unless you were playing on the exact same server at the exact same time, the comparison isn't valid. Multiplayer games stress so many different components (as I said before) that it's very difficult to control the many variables to a point to make a decent comparison. Source is known to be very CPU dependent. You can test for yourself: run at stock speed and run overclocked and you'll see a significant difference, especially on large servers (try 64 player CS:S, for example).
 
MrGuvernment said:
there is a server side setting for FPS that can result in lower choke for people in cs:s, it may apply in TF2.



since getting my 8800Gt stock and a e4400 @ 3.1ghz, i run 1920 x 1200 with ingame all high, and i recall AA @ 8xca? and AF on 8x also and keep around 60fps on all maps i have played.
Click to expand...

Server FPS does not affect player FPS. Look at BF2. The server has a stat that says it runs at "30fps". Meanwhile, you can easily get 200fps if you so choose. They mean two completely different things.

Tickrate is not what you're talking about either, that's to do with the number of updates/amount of bandwidth going between the client and the server.
 
LuminaryJanitor said:
I'm running it at the same settings as you, and I don't think mine dips any lower than that on an A64 3000+ / x1950pro 256.

So you can all stop blaming the CPU...
Click to expand...

Sorry but I think you havn't been in stressful maps/situation in TF2 yet.
Try 32player dustbowl opening with 4xAA and all highest settings.
I have a friend that has a E6600 + X1950PRO 256 and he sinks into the 30s~even 20s depending on how much action is going on according to FRAPS. Thats pretty similar perfoemance when I had my old X1900XT.
 
Mr. K6 said:
You missed the point of my post. Unless you were playing on the exact same server at the exact same time, the comparison isn't valid.
Click to expand...
I didn't miss your point. I know it's not an entirely level playing field. But it's still the same game, and even if I haven't run into the most stressful situation it can produce, the longer you spend playing the game, the closer you're likely to get.

Probability suggests that the worst the OP's seen shouldn't be that much worse than the worst I've seen. The fact that the OP has seen it happen regularly makes it even more likely that I've seen it happen once, or at least something very close.

ihira said:
Try 32player dustbowl opening with 4xAA and all highest settings.
Click to expand...
I have, many times, running 4xAA, 16xAF, at maximum settings, at the same res as you. If your system is dropping into the 30s there, then mine should have been in the teens, or worse. Like many of the above posters have said, something appears to be wrong.
 
You must log in or register to reply here.
Back
Top