How big is the FPS drop supposed to be in 1.31?

[soRaz]

I just did the whole test-procedure, and reported my findings in the survey.

I then redid the "no-clutter, no-shadow" test in 1.30, and got 69 fps, and in 1.31 I got just 25 fps.

I checked that both installations used the same settings (1900x1200, SSE2 etc etc). I have a Nvidia FX3700M, 1GB graphics card. (Quadro series i think).

I realise this is probably still just a debug build, but it would be nice if someone could say "you are supposed to loose ~ 10 fps on average".

Edit: d'oh, just realised the laptop was running on battery during the tests. After plugging in power I did the no-clutter-no-shadows test again.

I got 109fps in 1.30 rocksolid, and

I got 50fps average in 1.31, with fluctuations from 25 to 76 fps, all while standing perfectly still looking west down the tank track.

Edited December 28, 2009 by soRaz

[raftrey]

I too notice a dramatic decrease in FPS while testing at about the same rate in lost.

About half the FPS

I currently run in game between 50 to 100 FPS (give or take) in game currently, also depending on where I am at.

While testing with the given procedure I could not get over 50 FPS.

My test went from a low 20FPS on test 2 to 50 on Test 7.

This is half to FPS of what I am currently getting.

Will this be the case or will CRS correct whatever is affecting the FPS.

Keep in mind, my current system, with a town full of friendlies as well as enemies I get 75 to 90 FPS.

Doing a test Alone with no one on and not connected to a server I get less then 50 FPS

Does not give me confidence on the full release as of yet

[hick]

I just did the whole test-procedure, and reported my findings in the survey.

I then redid the "no-clutter, no-shadow" test in 1.30, and got 69 fps, and in 1.31 I got just 25 fps.

I checked that both installations used the same settings (1900x1200, SSE2 etc etc). I have a Nvidia FX3700M, 1GB graphics card. (Quadro series i think).

I realise this is probably still just a debug build, but it would be nice if someone could say "you are supposed to loose ~ 10 fps on average".

Edit: d'oh, just realised the laptop was running on battery during the tests. After plugging in power I did the no-clutter-no-shadows test again.

I got 109fps in 1.30 rocksolid, and

I got 50fps average in 1.31, with fluctuations from 25 to 76 fps, all while standing perfectly still looking west down the tank track.

I was getting ready to say... I'm seeing 50 fps solid in that test with 1.31 and I have the FX3600M, half the RAM and 12 less pipes.

What laptop you have btw and why are you using workstation GPU's?

I have a Precision M6300 and run SW2009 with it.

[soRaz]

I was getting ready to say... I'm seeing 50 fps solid in that test with 1.31 and I have the FX3600M, half the RAM and 12 less pipes.

What laptop you have btw and why are you using workstation GPU's?

I have a Precision M6300 and run SW2009 with it.

I have a Precision M6400; I think the FX3700M is the mobile version of the Quadro GPU.

It may well be that I have too much crap running, while performing the test, and that may negatively affect the FPS seen, but the second test (*with* power cord attached), was run within 2 minutes of eachother, so anything affecting test 1 (109-109-109-109-109-109-109 fps) should affect test 2 (26-50-76-50-76-26-31-50).

I use Vista 64bit, SP2. Not sure if that means anything wrt to this.

[xanthus]

Nvidia FX3700M, 1GB graphics card. (Quadro series i think)

There's your problem. Pretty sure that card is not officially supported by the game.

Can't really complain about gaming on a video card not made for gaming which uses drivers not meant gaming.

[hick]

There's your problem. Pretty sure that card is not officially supported by the game.

Can't really complain about gaming on a video card not made for gaming which uses drivers not meant gaming.

the FX3700M is a G94 core and performs as well as the GeForce equivalent.

Just like the FX3600M is the same hardware as the GeForce 8800GT and actually slightly outperforms it in OpenGL games. (Drivers optimized for OpenGL API vs Direct3D API).

[hick]

I have a Precision M6400; I think the FX3700M is the mobile version of the Quadro GPU.

It may well be that I have too much crap running, while performing the test, and that may negatively affect the FPS seen, but the second test (*with* power cord attached), was run within 2 minutes of eachother, so anything affecting test 1 (109-109-109-109-109-109-109 fps) should affect test 2 (26-50-76-50-76-26-31-50).

I use Vista 64bit, SP2. Not sure if that means anything wrt to this.

*drool M6400, what I wouldn't give to have quad-core support and 16gb RAM on mine

Yea, the FX3700M is one of the mobile Quadro cards, used to be the top of the line, but Nvidia just came out with the FX3800M so... it's now in second (although I can't really see that much of a difference spec wise, except that the 3700 was configured with between anywhere from 512mg to 1gb RAM and the 3800 only has 1gb RAM).

Hearing a lot of good things about Windows 7 as well, although if this is a production laptop you may want to hold off just from a stabilty standpoint.

What are the rest of the specs on yours, RAM, CPU, HDD?

[badger77]

There's your problem. Pretty sure that card is not officially supported by the game.

Can't really complain about gaming on a video card not made for gaming which uses drivers not meant gaming.

Well he's just saying there is a big hit from 1.30 to 1.31 and it's not just him and the "non gaming " idea you have.

I have a GTX260 216 XFX gaming card and my fps drop from 190 to 80 using the same spot and same settings

I am sure there is a fair bit of improvement to come and some optimisation. This is a beta not even ready for online beta yet so don't read too much into it.

I do think we will see a decent old hit though hence the need for better machines and improved min spec.

[soRaz]

What are the rest of the specs on yours, RAM, CPU, HDD?

I run with 4GB ram, a 60GB SSD for Windows and the most used programs, and a 200GB standard HD for everything else. CPU wise, I have 2.8ghz dual core.

So almost any program starts within 1-2 seconds. Except for games.

I expect to upgrade to 12 gb ram, and all ssd's within a few months, when I change to Win7.

[hick]

I run with 4GB ram, a 60GB SSD for Windows and the most used programs, and a 200GB standard HD for everything else. CPU wise, I have 2.8ghz dual core.

So almost any program starts within 1-2 seconds. Except for games.

I expect to upgrade to 12 gb ram, and all ssd's within a few months, when I change to Win7.

Why not go to the full 16? The price difference isn't that big.

I'm surious what you are running with that beast?

I'm eventually going to put a 64gb SSD in for Win7 and try to find a 500gb SATA drive for my files. I'd really like a 128GB SSD so I can load my programs on it as well but I can't seem to find one at a reasonable cost.

How many RAM slot does that hting have 2 or 4? I'm finiding that even the older-gen DDR2 for the the M6300 going above 4GB gets stupidly expensive (4gb chips are $180 each).

Anyhoo, seems like aside from the GPU, you have very similar specs to me, The only time i don't get a rock steady FPS is when something starts rning in the background (updaters, AV scan, etc). otherwise, it can hold this game pretty stable at 60FPS no matter what.

[soRaz]

Why not go to the full 16? The price difference isn't that big.

I'm surious what you are running with that beast?

I'm eventually going to put a 64gb SSD in for Win7 and try to find a 500gb SATA drive for my files. I'd really like a 128GB SSD so I can load my programs on it as well but I can't seem to find one at a reasonable cost.

How many RAM slot does that hting have 2 or 4? I'm finiding that even the older-gen DDR2 for the the M6300 going above 4GB gets stupidly expensive (4gb chips are $180 each).

Anyhoo, seems like aside from the GPU, you have very similar specs to me, The only time i don't get a rock steady FPS is when something starts rning in the background (updaters, AV scan, etc). otherwise, it can hold this game pretty stable at 60FPS no matter what.

The 12 was to reuse the 4 i have already, but you may have a point. AFAIK it has 4 slots, but the 4Gb blocks are, as you say, stupidly expensive.

At any time at work, I run 2-3 VS2008's, SQL Server (both Express, and Enterprise), DB2, Websphere MQ, Eclipse, 5-6 Excels, 3-6 Word instances, and Lotus Notes, plus a Virtual PC for checking stuff in IE6.

[badger77]

Well to be fair I don't think you'll use more than 12.

Even with all that stuff running but check it out.

[hick]

The 12 was to reuse the 4 i have already, but you may have a point. AFAIK it has 4 slots, but the 4Gb blocks are, as you say, stupidly expensive.

At any time at work, I run 2-3 VS2008's, SQL Server (both Express, and Enterprise), DB2, Websphere MQ, Eclipse, 5-6 Excels, 3-6 Word instances, and Lotus Notes, plus a Virtual PC for checking stuff in IE6.

ah, fun fun.

[peegee77]

In reply to the OP..... in another thread I posted this

Really would appreciate a RAT answer on this. Is there a direct comparison between 1.30 offline and 1.31 offline (in which case the performance drop is very worrying)? Or is the 1.31 beta client crunching a load of non-game stuff in the background (debugs, dev tools etc) which is giving the CPU a thrashing in beta but will be gone in the release?.

and got this reply from JAEGER

We are well aware about the high CPU load of the game.

Performance will be improved significantly for the release client with the changes in the works for data and code.

Given all the variables, I would be very surprised if any of the RATS would put a figure on "improved significantly". I may be wrong.