40 replies to this topic

[Vulpesveritas]

If possible, I would like to get together an idea of how everyone's systems are doing in MWO, and to improve future aiding of new members to the website.
As such, if you would like to join in, please post and record the following:

All benchmarks should be done on the in-game preset settings and the following resolutions for each. Minimum recording time is 60 seconds, and it must be in-game on a map.

Low Setting: Benchmark at 1366x768 (or 1280x800) Please state which.

Medium Setting: Benchmark at 1600x900

High Setting: Benchmark at 1920x1080

Pilot name:
Mech Piloted during recording:
Map Played:
Settings level: (Ex. "LOW", "MEDIUM", "HIGH")
CPU: (If overclocked, state overclock frequency and base frequency.)
GPU: (If overclocked, state overclock frequency and base frequency.)
RAM Type/brand/amount: (state speed and timings if overclocked)
Recording Length:
MSI AB/ FRAPS: (Inclusion with video optional but helpful)
Minimum FPS:
Average FPS:
Maximum FPS:

Edited by Vulpesveritas, 04 October 2012 - 12:21 PM.

[Oderint dum Metuant]

Im going to include the basic test you know i have already done, but couldn't release due to NDA.

The test was to see how a FX4170 would handle the game in its current state this is by no means be all end all as specific machines and chips will vary and we had to use an FX 8120 with cores disabled and overclocked to simulate it.
It may be the 8120 is not an accurate test bed.

For noting during gameplay my GPU usage did not exceed 40% which shows that the game currently is heavily CPU bound.

Running the 8120 2 cores 2 modules @ 4.2ghz
MSI GTX 580
8GB Gkill Sniper 1600mhz
Sabertooth 990FX Motherboard.

With everything on high, the game produced at best 37FPS at worst 17FPS it did average around the 23-27 mark.

Low settings.

Produced FPS at best 42 at worst 23FPS but averaged a more stable 34-37FPS
GPU usage did not exceed 34%

Medium Settings

Produced more stable results, 25-27FPS no lower than that, highest was 37.
GPU usage 34%

Currently shows the game is horribly CPU bound as its barely even using the GPU which is surprising, and there is nearly not much difference in performance from a stock 8120 and the 4170 its about 10FPS.

The last comment is because my 8120 overclocked was stuck producing FPS in the 40's. I have since switched to an Asus Maximus V Gene and 2500k which produces much better FPS at high 50-60FPS currently the 2500k is running at stock speeds i will do proper tests with this over the weekend

[Catamount]

There needs to be a consistent resolution between submissions. The resolution one plays with should usually correlate with the settings one is able to play on. Laptops and budget machines with 720P screens won't have stellar performance most of the time, while anyone with a system capable of playing on medium/high should at least have a 900P or 1080P screen, so this should be a good lowest common denominator, allowing anyone who can play at one of these settings to be able to meet the screen resolution requirement. This helps eliminate one of the big extraneous variables in benchmarking, allowing the results to be better compared (two results are two different resolutions are all but useless for comparison).


this is what I suggest asking people for:

Low Setting: Benchmark at 1366x768 (or 1280x800)

Medium Setting: Benchmark at 1600x900

High Setting: Benchmark at 1920x1080

For medium/high, if the person benchmarking as a 16x10 monitor, running at 1680x1050 or 1920x1200, they can still drop down to the slightly lower resolution, whereas 16x9 monitors would not be able to switch to the higher one. For low, the difference in resolution between 1366x768 and 1280x800 is negligible (2.45%), so those can be used interchangeably.

Edited by Catamount, 04 October 2012 - 12:11 PM.

[Vulpesveritas]

Catamount, on 04 October 2012 - 12:07 PM, said:

There needs to be a consistent resolution between submissions. Since the resolution one plays with should usually correlate with the settings one is able to play on, this is what I suggest asking people for:

Low Setting: Benchmark at 1366x768 (or 1280x800)

Medium Setting: Benchmark at 1600x900

High Setting: Benchmark at 1920x1080

For medium/high, if the person benchmarking as a 16x10 monitor, running at 1680x1050 or 1920x1200, they can still drop down to the slightly lower resolution, whereas 16x9 monitors would not be able to switch to the higher one. For low, the difference in resolution between 1366x768 and 1280x800 is negligible (2.45%), so those can be used interchangeably.

Edited to express this, thank-you Cat.

[Aznpersuasion89]

Pilot name: Aznpersuasion89
Settings level: Max everything
Res: 1440x900 (also done @ 1360x768 with same results)
CPU: AMD Phenom II x4 975 Black 3.6 GHZ
GPU: Sapphire OC Edition 7850 256 Bit (GPU 920/ MEM 1250)
RAM Type/brand/amount: AMD Performance 8GB 8-8-8
Recording Length: 4:30
MSI AB/ FRAPS: MSI AB
Minimum FPS: Min has been as low as 20 FPS
Average FPS: anywhere in between
Maximum FPS: as high as 75 FPS V-Sync

GPU load ~50-60%

Edited by Aznpersuasion89, 04 October 2012 - 05:42 PM.

[Bishop Steiner]

OK, will fill in what I can.

I don't have any video capture programs, so if you can recommend a good one, I can amend video to this later.
Leaving this as a place holder, will edit it as I get info.

CPU: AMD Phenom Black Edition 6 Core, 3.8 ghz (turbo 4.1), stock.
GPU: EVGA Invidia GeForce 560GTX, stock.
RAM Type/brand/amount: 8 gb Patriot Viper Extreme 240 pin DDR3 1866

Mech Played: Trial Catapult K2
Map Played: Caustic
Recording Length: 8 minutes
Resolution: 1280x720 full screen
Settings level: low
MSI AB/ FRAPS: Dunno what this is, tell me, and will find out.
Minimum FPS: 31
Average FPS: 52
Maximum FPS: 64

Mech Played: Trial Catapult K2
Map Played: Forest
Recording Length: 5 minutes 12 seconds
Resolution: 1280x720 full screen
Settings level: medium
MSI AB/ FRAPS:
Minimum FPS: 21
Average FPS: 46, but once ion heavy combat, 29
Maximum FPS: 62 fps

Mech Played: Trial Catapult K2
Map Played:
Recording Length:
Resolution: 1280x720 full screen
Settings level: High
MSI AB/ FRAPS:
Minimum FPS:17
Average FPS:32
Maximum FPS: 49

[Stickjock]

the MSI AB/FRAPS are the video capture programs you can use Bishop...

Fraps is pretty much a dedicated video/screen shot cap program while MSI Afterburner is a GPU tweaking program that allows recording as well...

[mekabuser]

approve of this idea but its VERY safe to say that all dual cores will be taxed heavily at any clock speeds. My cpu E5200@3.5ghz a 1 ghz OC is pegged at 95% plus the whole time I play mwo.
This is unfortunate because iirc it was CW that dual cores were just fine for budget gaming rigs up till at least 3 years ago.Maybe even two So consequently alot of people have dual cores that run most everything just fine, except this.

Best news I can say is i5 2500k is 159 usd at microcenter and you can combo up with a decent MB for like 200 plus. They got deals.
Its important for pgi to optimize for dual core. TEST ON ONE GUYS, until people can eventually upgrade.

[Vulpesveritas]

Bishop Steiner, on 05 October 2012 - 08:50 AM, said:

OK, will fill in what I can.

I don't have any video capture programs, so if you can recommend a good one, I can amend video to this later.
Leaving this as a place holder, will edit it as I get info.

CPU: AMD Phenom Black Edition 6 Core, 3.8 ghz (turbo 4.1), stock.
GPU: EVGA Invidia GeForce 560GTX, stock.
RAM Type/brand/amount: 8 gb Patriot Viper Extreme 240 pin DDR3 1866

Mech Played: Trial Catapult K2
Map Played: Caustic
Recording Length: 8 minutes
Resolution: 1280x720 full screen
Settings level: low
MSI AB/ FRAPS: Dunno what this is, tell me, and will find out.
Minimum FPS: 31
Average FPS: 52
Maximum FPS: 64

Mech Played: Trial Catapult K2
Map Played: Forest
Recording Length: 5 minutes 12 seconds
Resolution: 1280x720 full screen
Settings level: medium
MSI AB/ FRAPS:
Minimum FPS: 21
Average FPS: 46, but once ion heavy combat, 29
Maximum FPS: 62 fps

Mech Played: Trial Catapult K2
Map Played:
Recording Length:
Resolution: 1280x720 full screen
Settings level: High
MSI AB/ FRAPS:
Minimum FPS:17
Average FPS:32
Maximum FPS: 49

As stated, MSI Afterburner and FRAPS are the video recording / framerate reading programs I am wanting to run this benchmarking on.
Also, Bishop, your CPU is an FX-6200, not a Phenom II X6 BE.

[CCC Dober]

Here's one for you on the lower end of the spectrum.

CPU: Intel Q6600 @ 3.0Ghz
GPU: KFA2 Geforce 660 Ti Ex OC
RAM: 6GB
RES: 1024x768

All settings cranked up to max but still smooth fps. The same system was choking real hard with a Geforce 9800 GTX+ even on lowest settings. Draw your own conclusions.

[Vulpesveritas]

CCC Dober, on 05 October 2012 - 09:38 AM, said:

Here's one for you on the lower end of the spectrum.

CPU: Intel Q6600 @ 3.0Ghz
GPU: KFA2 Geforce 660 Ti Ex OC
RAM: 6GB
RES: 1024x768

All settings cranked up to max but still smooth fps. The same system was choking real hard with a Geforce 9800 GTX+ even on lowest settings. Draw your own conclusions.

Would it be possible to have more accurate information?

[CCC Dober]

@Vulpe
I can do that for ya mate, no sweat =)

What do you want to know exactly?

[Vulpesveritas]

CCC Dober, on 05 October 2012 - 11:22 AM, said:

@Vulpe
I can do that for ya mate, no sweat =)

What do you want to know exactly?

Vulpesveritas, on 03 October 2012 - 03:58 PM, said:

If possible, I would like to get together an idea of how everyone's systems are doing in MWO, and to improve future aiding of new members to the website.
As such, if you would like to join in, please post and record the following:

All benchmarks should be done on the in-game preset settings and the following resolutions for each. Minimum recording time is 60 seconds, and it must be in-game on a map.

Low Setting: Benchmark at 1366x768 (or 1280x800) Please state which.

Medium Setting: Benchmark at 1600x900

High Setting: Benchmark at 1920x1080

Pilot name:
Mech Piloted during recording:
Map Played:
Settings level: (Ex. "LOW", "MEDIUM", "HIGH")
CPU: (If overclocked, state overclock frequency and base frequency.)
GPU: (If overclocked, state overclock frequency and base frequency.)
RAM Type/brand/amount: (state speed and timings if overclocked)
Recording Length:
MSI AB/ FRAPS: (Inclusion with video optional but helpful)
Minimum FPS:
Average FPS:
Maximum FPS:

What was stated in the first post preferably.

[CCC Dober]

Aww man, I'm stuck at 1024x768. Don't ask, it's rather embarassing LOL
I can put up some results for that resolution though ...

*edit*


CPU: Q6600 @ 3.0Ghz (originally 2.4 Ghz)
GPU: KFA2 Geforce GTX 660 Ti EX OC
RAM: 6GB DDR2 Kingston
Measuring tool: EVGA Precision X 3.0.3 OSD

Pilot name: CCC_Dober
Mech piloted during recording: Catapult-K2
Map played: Forrest colony
Settings: very high @ 1024x768 (vsync off, motion blur off, msaa off)
Minimum FPS: 19
Average FPS: 29
Maximum FPS: 40

Pilot name: CCC_Dober
Mech piloted during recording: Catapult-K2
Map played: Frozen City
Settings: low @ 1024x768 (vsync off, motion blur off, msaa off)

Minimum FPS: 22
Average FPS: 31
Maximum FPS: 43


Is it me or does it look like the CPU could use a bit more 'juice' ?

Edited by CCC Dober, 05 October 2012 - 12:44 PM.

[Vulpesveritas]

CCC Dober, on 05 October 2012 - 11:34 AM, said:

Aww man, I'm stuck at 1024x768. Don't ask, it's rather embarassing LOL
I can put up some results for that resolution though ...

run at low settings for a comparison to other low resolution tests.

[CCC Dober]

Vulpesveritas, on 05 October 2012 - 12:19 PM, said:

run at low settings for a comparison to other low resolution tests.

Like this?

CCC Dober, on 05 October 2012 - 11:34 AM, said:

...
CPU: Q6600 @ 3.0Ghz (originally 2.4 Ghz)
GPU: KFA2 Geforce GTX 660 Ti EX OC
RAM: 6GB DDR2 Kingston
Measuring tool: EVGA Precision X 3.0.3 OSD
...
Pilot name: CCC_Dober
Mech piloted during recording: Catapult-K2
Map played: Frozen City
Settings: low @ 1024x768 (vsync off, motion blur off, msaa off)

Minimum FPS: 22
Average FPS: 31
Maximum FPS: 43

=)

Edited by CCC Dober, 05 October 2012 - 01:14 PM.

[CCC Dober]

Lil update here:

I have to say that tweaking the user.cfg helped a lot. Kudos to the gents in this thread right here -> http://mwomercs.com/...rmance-visuals/

I'm using the Low/Medium NVIDIA SPEC by Demonr6 & IceNinja
Forrest colony went up by 5-10 fps on average and looks much clearer now (less fog). I'm not really a fan of 'impaired' vision so that's an added benefit.

[Barbaric Soul]

all settings set to maximum, motion blur off, 1920*1200 resolution, playing caustic valley in a K2 guassboat, my FPS were a pretty steady 60fps. Other maps I have seen upwards of 80 fps. I use MSI Afterburner of monitor my FPS and OC my GPU.

system specs-
CPU- 2600k OC'ed to 4ghz
RAM- 8gig DDR3 1600
GPU- HD7970 OC'ed to 1100MHz core 1400MHz memory
Monitor- 24" 1920*1200 resolution

[Spunkmeyer]

Pilot name: Spunkmeyer
Mech Piloted during recording: HBK, CPLT, AWS
Map Played: All
Settings level: Low
CPU: (If overclocked, state overclock frequency and base frequency.) not overclocked
GPU: (If overclocked, state overclock frequency and base frequency.) not overclocked
RAM Type/brand/amount: CRUCIAL, 8GB (2 x 4GB) Ballistix Tactical PC3-14900 DDR3 1866MHz CL9 (9-9-9-27) 1.5V SDRAM DIMM, Non-ECC
Processor: INTEL, i3-2130 Dual-Core 3.4GHz
Graphics Card: GIGABYTE, GV-R645OC-1GI, Radeon™ HD 6450 675MH
MSI AB/
FRAPS: Sorry, no fraps
Minimum FPS:7fps
Average FPS: 12fps
Maximum FPS: 13fps
Game Resolution: Windowed 1280x800 *fixed*

PGI really needs to add support for dual-core users, otherwise they are alienating a good portion of their playerbase. I've heard of support down the elusive pipeline, but no other ETA has been given. The graphics settings implemented a few patches ago generally do nothing to boost performance. Why is priority being given to things like the economy instead of performance?

Edited by Spunkmeyer, 06 October 2012 - 08:31 AM.

[Catamount]

Spunkmeyer, on 06 October 2012 - 06:02 AM, said:

Pilot name: Spunkmeyer
Mech Piloted during recording: HBK, CPLT, AWS
Map Played: All
Settings level: Low
CPU: (If overclocked, state overclock frequency and base frequency.) not overclocked
GPU: (If overclocked, state overclock frequency and base frequency.) not overclocked
RAM Type/brand/amount: CRUCIAL, 8GB (2 x 4GB) Ballistix Tactical PC3-14900 DDR3 1866MHz CL9 (9-9-9-27) 1.5V SDRAM DIMM, Non-ECC
Processor: INTEL, i3-2130 Dual-Core 3.4GHz
Graphics Card: GIGABYTE, GV-R645OC-1GI, Radeon™ HD 6450 675MH
MSI AB/
FRAPS: Sorry, no fraps
Minimum FPS:7fps
Average FPS: 11fps
Maximum FPS: 14fps
Game Resolution: Windowed 1280x720

PGI really needs to add support for dual-core users, otherwise they are alienating a good portion of their playerbase. I've heard of support down the elusive pipeline, but no other ETA has been given. The graphics settings implemented a few patches ago generally do nothing to boost performance. Why is priority being given to things like the economy instead of performance?

Please record at the requested resolutions; if you don't, we can't use the information. Different resolutions will make it impossible to compare the results (yours is close enough that we can probably extrapolate, but that's not ideal).