2015 Performance Test for FMH2015 New Match Engine (with result)

[rseven]

Hi, mates,

Many user complain the slow speed of new match engine, but how slow it is?

I already request a performance test in Chinese FMH gamer social network, and some mates there have posted their test result.

The results are list below for reference.

The time here is only playback time, the beginning show and tactics adjustment is not counted. 

[tr][td]Device[/td][td]CPU[/td][td]RAM[/td][td]Time[/td][/tr]

Device                    CPU                     RAM             Time

IPad Air 2                A8                        1G                1m15s

IPhone 6                 A8                        1G                1m29s

IPad mini 2              A7                        1G                1m39s

IPhone 5s                A7                        1G                1m56s

 

LG G2                     Snapdragon 800   2G                4m21s (By Ty Wilson)

HTC One M8          Snapdragon 801   2G                4m27s (By Fractured Life)

VIVO Xplay 3S        Snapdragon 801   3G                4m30s

HTC One M8          Snapdragon 801   2G                4m31s (By maceyp)

SAMSUNG N9005  Snapdragon 800   3G                4m31s

SAMSUNG S5        Snapdragon 801   2G                4m36s

Sony Xperia Z2       Snapdragon 801   3G                4m39s

SAMSUNG N9006   Snapdragon 800   3G                4m40s

LG G3                     Snapdragon 801   2G                4m41s (By Fractured Life)

100+V6                    MT6592T              2G                4m48s

Huawei Honor 6       Kirin 920               3G                4m56s

Xiaomi RedMi S1     MSM8628             1G                6m00s

Xiaomi 2S                APQ8064 Pro        2G                6m21s

SAMSUNG S4         Exynos 5410          2G               7m00s

SAMSUNG S4         Exynos 5410          2G               7m22s (By JustM1kePlays)

Huawei Mate 1        K3V2                      2G               8m20s

Lenovo A780e         MSM8610              1G             10m58s

SAMSUNG S3         Exynos 4412          1G             11m00s

Asus Memo HD7     MT8125                 1G             12m35s (By catalin070697)

 

 

 

You are welcome to join if you are interested. I will summary the test result here and report to SI. Hope SI can use these feedback to improve their work.

 

Institution to do the test.

1. Download pkm file here. To keep on same base, let's test on same pkm file. It is only 24k, a small file.

2. Put the pkm file into your fmh2015 user folder.

    For Android, put in fmh2015_data\user_data

    For iOS, put in fmh2015 documents

3. Start game and click right bottom "view match"

4. Load "Man Utd v Chelsea"

5. Set speed to "Very Fast" and start play back. and start to your timer.

6. Stop your timer when paly stop.

7. post your result in this post with same format and information.

Thanks.

 

Edit: 

Two result  added from Chinese network. 

Add Vibe mates result in Blue on Nov.28

Edited December 24, 2014 by rseven

[Dec]

Very interesting. Nice work. The difference in time is ridiculous!

[rseven]

Very interesting. Nice work. The difference in time is ridiculous!

Yes.

The difference between iOS device and Android device is so big even these android device almost top levels. 

I hope SI to optimize the ME for android device.

[vtos]

 

I hope SI to optimize the ME for android device.

 

Probably it's not about ME optimization for Android, it's believed that Android is slower by itself than iOS due to its inner depths and structure features so we can hardly hope the ME on Android will be as fast as on iOS.

IMO, of course.

[catalin070697]

Asus Memo Pad HD7 | Cortex A7 | 1G | 12m35s

Big difference between this:

and this:

[rseven]

Asus Memo Pad HD7 | Cortex A7 | 1G | 12m35s

Big difference between this:

and this:

 

 

Do you use new match engine in your save?

I guess you use original engine the one used in fmh2014.

And when you play back, it will force you to use new ME.

This is what I think about why you have big difference.

[Lee-Roy]

The editor looks like it's been used to change kit colours, am I correct?

You have to ask yourself if such a thing would spoil the test (and before you say I don't think so, that's not a definitive answer) even in a minor way. Sure the colours shouldn't affect the match engine, but then again it's got to retain the information to change the colour of the dots that are moving around. Even in a minor way it could be adding on just that little bit extra processing time. Or it may not be but its worth considering.

[jaymarvels]

I don't think editing colours would effect the ME as its just a bunch of numbers stored in memory, its like replacing 123 with 456

[catalin070697]

Do you use new match engine in your save?

I guess you use original engine the one used in fmh2014.

And when you play back, it will force you to use new ME.

This is what I think about why you have big difference.

Yes I used original ME to show what big difference is (could play 10 games on Original instead of 1 on Enhanced)

The editor looks like it's been used to change kit colours, am I correct?

First picture is rseven's file so I don't know if GE was used but I've also wondered when i saw the kit colours (thinking that Chelsea is in blue)

[Lee-Roy]

Yeah I realised you was using R7's file, that's why I asked as it sorts affects the whole test

I don't think editing colours would effect the ME as its just a bunch of numbers stored in memory, its like replacing 123 with 456

Yeah but what occurred to me is, would the ME be affected by having to remember to keep 456 instead of 123. Obviously I'm not computer savvy on your guys level so I'm basing my thoughts on simple logic that makes sense to me lol

[rseven]

I don't think editing colours would effect the ME as its just a bunch of numbers stored in memory, its like replacing 123 with 456

The color changed I think is due to I creat this pkm from my Ipad.

[jaymarvels]

Yeah I realised you was using R7's file, that's why I asked as it sorts affects the whole test

I don't think editing colours would effect the ME as its just a bunch of numbers stored in memory, its like replacing 123 with 456

Yeah but what occurred to me is, would the ME be affected by having to remember to keep 456 instead of 123. Obviously I'm not computer savvy on your guys level so I'm basing my thoughts on simple logic that makes sense to me lol

Doubtful as it'll be the same number of bytes in memory

[rseven]

pkm file is very small, only 24k. I believe it should be a "script file".

All resource is load from local device. 

[JustM1kePlays]

Samsung S4

[anfieldsquirrel]

I'll have a go of this, how do I get a timer up on the game screen on a Samsung btw :lol:

#androidnewbie

[Fractured Life]

I'll post results for a HTC One M8 and LG G3 shortly...

 

 

EDIT: LG G3 (Snapdragon 801, 2GB Ram) 4mins 41s

EDIT: HTC ONE M8 5:36. However, completely forgot I have the faces pack downloading in the background. Will rerun when I'm not having such a brainfart...

 

Re-ran the HTC One M8 (Snapdragon 801, 2GB Ram) after the facepack was finished downloading, 4mins 27s. Bragging rights at work for me

[maceyp]

Im new to Football Manager on Handheld but been a PC player since CM3 days. I wasn't sure which engine to select so I have decided to do this test on my HTC One M8.

 

Half-Time: 02:34

Full-Time (Total): 04:31

 

I guess its slow compared to the top ones but i think its pretty quick for mobile

[rseven]

Im new to Football Manager on Handheld but been a PC player since CM3 days. I wasn't sure which engine to select so I have decided to do this test on my HTC One M8.

Half-Time: 02:34

Full-Time (Total): 04:31

I guess its slow compared to the top ones but i think its pretty quick for mobile

Your htc one m8 is in top of android device.

Don't worry.

[JustM1kePlays]

I'll have a go of this, how do I get a timer up on the game screen on a Samsung btw :lol:

#androidnewbie

download a floating timer or stopwatch from Google play store mate.

[Callum98]

Great test, shows a great deal of difference between all these devices.

[Ty Wilson]

4:21 on LG G2, making me seriously wonder if i should upgrade to g3. I have seen other benchmarks between both phones and they always seem grim, I want something twice as powerful if I end spending hundreds.

[Fractured Life]

Populating all those extra pixels comes at a price. At the end of the day if you only care for how fast it will play a match just root the g2 and pare everything back which is unneeded, optimize it and you can probably go quicker again. 

 

If I get a chance I might try it on my old rooted one x. Was pretty fast once I put it on cyanogenmod.

[jaymarvels]

But at 4:47 for a G3...top end phone....

[Fractured Life]

Dude it was 4:41, huge difference! Lol

 

They're seeming pretty equal really on this generation. One m8 is probably a little bit faster as sense U.I. is pretty good compared to the LG and Samsung.

 

 

What's the difference with the iPad/phone in terms of the game? Is there more to the android version? Mate had 14 on iPhone 5s and once I showed him the android version he went out and got a G3 and paid again for fmh14 as he thought it was much better. 

[jaymarvels]

Dude it was 4:41, huge difference! Lol

 

They're seeming pretty equal really on this generation. One m8 is probably a little bit faster as sense U.I. is pretty good compared to the LG and Samsung.

 

 

What's the difference with the iPad/phone in terms of the game? Is there more to the android version? Mate had 14 on iPhone 5s and once I showed him the android version he went out and got a G3 and paid again for fmh14 as he thought it was much better. 

 

IPad Air 2                A8                        1G                1m15s

IPhone 6                 A8                        1G                1m29s

IPad mini 2              A7                        1G                1m39s

IPhone 5s                A7                        1G                1m56s

 

I'm pro android, but kinda proves how tightly knit iOS is tied to the processor

[Fractured Life]

I saw the times aye, just mean what is different in the game? So basically we are saying that FMH15 is the exact same across both platforms, but is 3 times faster on a iPhone than a top spec android?

 

I don't recall many tests where my m8 was 3 times slower than an iPhone at anything. Seems strange to me.

[GillSCFC1999]

I think it may be to do with the fact that less is possible on iPhone anyway. And with skins, logos, edited databases, etc... us Android users have to deal with this as a compromise. OR... maybe more runs in the background on Android phones, even if it's memory capabilities are higher.

[jaymarvels]

Both will share the same code base...so by my reckoning its down to either the way FMH utilises the CPU/GPU/Memory or the way Android/iOS handles it

[Fractured Life]

@Jay, ha I just realised it was you created the scout tool for android? Loved it last year, bought the data editor on 15 just to see the

PA's of youngsters I'm scouting as I knew it would be a while before I could buy yours on play store again. Thanks for your work on that last year,  using it when starting in the lower leagues was a God send!

[jaymarvels]

I think it may be to do with the fact that less is possible on iPhone anyway. And with skins, logos, edited databases, etc... us Android users have to deal with this as a compromise. OR... maybe more runs in the background on Android phones, even if it's memory capabilities are higher.

 

Its only less as Apple lock down file system access, JB users end up seeing the same file structure as Android users.