Jump to content
Sports Interactive Community

Football Manager 16 Benchmark Thread

Guest

By Guest,
in Football Manager General Discussion

Recommended Posts

PineappleBlender

PineappleBlender

Posted
Posted
By then if you would want a truly stress test for these lines, the difference in cores/threads would show more if the number of competitions simed in full details would increase. How much that is practical (I personally don#t sim that much on full details) is another thing, but it's obvious the number isn't that highly. The OC'ed Pentium G (2 Cores / 2 Threads) is outperforming some Core i5s (4/4). That's Intel's desktop budget line, the level below the Core i3s (2 cores/4 threads). This wouldn't happen even in a typical AAA real-time graphics 3d video game, and those don't scale as much as FM's match sim does, not even at their most optimized, presumably because depending on the number of cores/threads available, the matches are simply calculated in parallel. The more threads available, the more matches can be calculated in parallel. The load is more like application software rather than your average 3d video game, but then as it's a measure of how long it takes to calculate something rather than about consistency of frame rates of images (all stuttering or fluid picture), that's rather logical. :-)

Does FM16 actually do stuff in parallel though? For FM14 at least, I demonstrated that the quick match engine only uses a single core; I had 7 cores sitting idle doing nothing when it should have destroyed the benchmark.

Until that problem is fixed, I'm never buying another FM game. I'm not going to support software written like it's still the mid 90s, especially when it's an embarrassingly parallel problem.

Link to post
Share on other sites
Guest

Guest

Posted
Posted

Considering that FM16 is showing a 30% reduction in processing time for Test A & a 60% reduction for Test B (both primarily use the quick ME) there must have been change in how processing is handle.

Obviously only SI can confirm how this has been achieved.

Link to post
Share on other sites
thomit

thomit

Posted
Posted

CPU: Intel Core i7 6700HQ

CPU frequency: 2.6 - 3.5 MHz

RAM: 8 GB DDR4 2133 MHz CL13

Windows 10 64 bit

Storage: SSD

Benchmark A: 3.23

Benchmark B: 5.23

Edit: With my old i7 920 (desktop, no oc) I used to run FM 15 with all leagues in play, but only the top 5 european leagues as "playable", the rest was "view only". With my new i7 6700HQ (laptop) I run FM 16 with all leagues, all playable, as well as some extra additions of players via the advanced database setup ... I'm happy with the processing speed, both on the old and the new.

Link to post
Share on other sites
Svenc

Svenc

Posted
Posted
Does FM16 actually do stuff in parallel though? For FM14 at least, I demonstrated that the quick match engine only uses a single core; I had 7 cores sitting idle doing nothing when it should have destroyed the benchmark.

Yeah, that's the thing I was getting at. The quick sim, that's quick sim. It doesn't create 90 minute "real-time" matches at all. Every thing else seems single-thread too mostly, how much of a benefit there would be if it wasn't that's one for SI's programming staff. But on full match details, that is the actually fut fat ME, the game grabs every thread it can [presumably so that multiple matches can be simulated in parallel]. It did with much older version of FM than FM 2014 already. When I test stuff and holiday, FM is pretty much the only program that on occasion makes the fans on my comp spring to a higher speed, notably. :D Don't know whether there's a limit, but the full match details during match days, they grab any thread they can and pump them to the max. You would see this in the task manager every time there is a match day commencing. There's hugely spikes in the CPU load then and all cores are utilized.

That's the reason why in the above ranking a 2 thread Pentium G budget chip outperforms all Core i5s, and the Core i3 is level with them. otherwise they couldn't possibly. We used to test this on previous versions as well, Core i3s of that generation (slower than this one) were struggling to compete with aging first gen Core 2 Quads. But we had put the details to "maximum" for everything. Barside makes a really good point though that for your average player (including me), during normal saves how many comps do you sim in full match details (as to be configured in the details menu during a save, not at the start of the game when you pick leagues)? At minimum, all the competitions you compete in by default are simulated in full match details, naturally, so that all your rival's matches are even and comparably and watchable and goals of the month are watchably etc. But other than that the sky (or the comps loaded) are the limit. And they slow things down significantly, much more so on chips who can't handle multiple threads.

Link to post
Share on other sites
Guest

Guest

Posted
Posted

You also need to factor in how well the user takes care of their system, unlike structured benchmarking the systems tested here could include various drains in performance.

CPU: Intel Core i7 6700HQ

CPU frequency: 2.6 - 3.5 MHz

RAM: 8 GB DDR4 2133 MHz CL13

Windows 10 64 bit

Storage: SSD

Benchmark A: 3.23

Benchmark B: 5.23

:mad: Damn it

Link to post
Share on other sites
Guest

Guest

Posted
Posted

Decided to run a full detail comparative on my machines, I used the Test B file to include transfer deadline day & then took the FM15 Benchmark file which I set to have the week holiday period to also end when the transfer window closed.

Both saves were set to Maximum full detail for international & club competitions, this means that every match that takes place is processed using the full match engine.

FM15

i7-920

Time: 39m 9s

Transfer completed: 826

i7-6820HK

Time: 36m 5s

Transfers completed: 852

FM16

i7-920

Time: 31m 42s

Transfers completed: 2145

i7-6820HK

Time: 29m 0s

Transfers completed: 2152

Initially I was disappointed with the time differences but then I looked at the transfer activity & FM16 processed 2.5 times as many completed transfers than 15 which does mess with the results & will probably call for a bit of fine-tuning in any future comparative tests, at least it does show a marked improvement in processing times & it would be interesting what times other people get, especially those with non-HT quad core or dual core cpu's.

Link to post
Share on other sites
thomit

thomit

Posted
Posted
You also need to factor in how well the user takes care of their system, unlike structured benchmarking the systems tested here could include various drains in performance.

:mad: Damn it

A reason for this could be due to the fact that these processors' turbos works differently, even in the same laptop chassis - which I think is the case here, is it not? I believe we both have the Clevo P650/651 chassis. There's a very fresh review of the Clevo P651RG over at Notebookcheck.net, with a 6820HK in it. Their tests show that the 6820 can't utilize it's turbo as efficiently as the 6700HQ can. They're saying - for example - that this machine stops doing any turboing when the load is sufficiently demanding - like running a stress-testing software. It does not start to throttle down though, which is actually a very good result for any laptop. But with the 6700HQ in the same chassis, this does not happen. My own tests has shown me that even when I run FurMark and Prime95 simultaneously, my turbo still does it's turboing. Only to between 2.7 - 2.8 MHz (all cores), but still. This is unheard of in the world of laptops. I am not aware of any other processor/chassis laptop combo that can do that. 99% of all laptops starts to seriously throttle down from their base clocks in such circumstances; a very few may be able to run with it's base clocks. Running FurMark alone, I still get turbo to between 2.9 and 3.1 MHz - the max for turbo on all cores. The 6700HQ works in perfect harmony with the chassis (cooling system) of this particular laptop; the 6820HK may be one tiny step too far to be able to acheive that. But of course, the 6820HK is overclockable, and can be forced to provide better results. Can't do that with the 6700HQ.

I do run a fairly streamlined setup, cutting out uneccessary processes/services and programs, but not going to any extremes in that area at all.

Link to post
Share on other sites
Guest

Guest

Posted
Posted

A little more testing unfortunately confirms that for the most part FM is still only utilising 1 core when processing (used Intel XTU monitor), cores 2-4 of my Skylake i7 appear to only be activated when FM was processing full detail matches so there is still plenty of room for improvement to make use of multiple cores & leave HT to kick in for match processing.

The real world performance improvements this year have been impressive & do give me confidence that the team at SI can work on the issue of underutilised multiple core cpu's, of course this is based on the assumption that spreading the load across all available cores will provide additional real world performance benefits at no cost to game stability.

Link to post
Share on other sites
  • 4 weeks later...
AhMostyn

AhMostyn

Posted
Posted

Upgraded to SSD - 20 secs faster for A, negligible for B

CPU: Intel Core i5 6500

CPU Frequency: 3200MHz

RAM: 16GB DDR4 2133MhZ

OS: Win 10 64-bit

Storage: SSD

Test A

Time: 3m 24s

Test B

Time: 5m 38s

Link to post
Share on other sites
Ferni

Ferni

Posted
Posted
[TABLE=class: grid, width: 950]

<tbody>[TR]

[TD]CPU[/TD]

[TD]Frequency[/TD]

[TD]RAM[/TD]

[TD]Frequency[/TD]

[TD]Storage[/TD]

[TD]OS[/TD]

[TD]Test A[/TD]

[TD]Test B[/TD]

[/TR]

[TR]

[TD]Intel i7-6700K[/TD]

[TD]4700 Mhz OC[/TD]

[TD]16GB DDR4[/TD]

[TD]3000 Mhz[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]2m 25s[/TD]

[TD]3m 55s[/TD]

[/TR]

[TR]

[TD]Intel i5-6600K[/TD]

[TD]4400 Mhz OC[/TD]

[TD]16GB DDR4[/TD]

[TD]3000 Mhz[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]2m 32s[/TD]

[TD]4m 14s[/TD]

[/TR]

[TR]

[TD]Intel i7-5820K[/TD]

[TD]4500 Mhz OC[/TD]

[TD]16Gb DDR4[/TD]

[TD]2133 Mhz CL13[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]2m 42s[/TD]

[TD]4m 2s[/TD]

[/TR]

[TR]

[TD]Intel i7-6820HK[/TD]

[TD]4000 Mhz OC[/TD]

[TD]16GB DDR4[/TD]

[TD]2133 Mhz CL13[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]2m 54s[/TD]

[TD]5m 2s[/TD]

[/TR]

[TR]

[TD]Intel i5-4670K[/TD]

[TD]4100 Mhz OC[/TD]

[TD]16GB DDR3[/TD]

[TD]1600 Mhz CL9[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]3m 1s[/TD]

[TD]5m 4s[/TD]

[/TR]

[TR]

[TD]Intel i7-6700HQ[/TD]

[TD]2600 Mhz (Turbo 3500Mhz)[/TD]

[TD]8GB DDR[/TD]

[TD]2133 Mhz CL13[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]3m 23s[/TD]

[TD]5m 23s[/TD]

[/TR]

[TR]

[TD]Intel i7-4790[/TD]

[TD]3666 Mhz (Turbo 4000Mhz)[/TD]

[TD]8GB DDR3[/TD]

[TD][/TD]

[TD]HDD[/TD]

[TD]Win 10 64bit[/TD]

[TD]3m 31s[/TD]

[TD]5m 7s[/TD]

[/TR]

[TR]

[TD]Intel i7-2600K[/TD]

[TD]4200 Mhz OC[/TD]

[TD]8GB DDR3[/TD]

[TD][/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]3m 37s[/TD]

[TD]5m 52s[/TD]

[/TR]

[TR]

[TD]Intel i7-6820HK[/TD]

[TD]2700 Mhz (Turbo 3600Mhz)[/TD]

[TD]16GB DDR4[/TD]

[TD]2133 Mhz CL13[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]3m 37s[/TD]

[TD]5m 59s[/TD]

[/TR]

[TR]

[TD]Intel i5-6500[/TD]

[TD]3200 Mhz[/TD]

[TD]16GB DDR4[/TD]

[TD]2113 Mhz[/TD]

[TD]HDD[/TD]

[TD]Win 10 64bit[/TD]

[TD]3m 44s[/TD]

[TD]5m 51s[/TD]

[/TR]

[TR]

[TD]Intel i7-4710MQ[/TD]

[TD]2500 Mhz (Turbo 3500Mhz)[/TD]

[TD]16GB DDR3[/TD]

[TD]1600 Mhx CL9[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]3m 47s[/TD]

[TD]6m 5s[/TD]

[/TR]

[TR]

[TD]Intel i7-920 (D0)[/TD]

[TD]3995 Mhz OC[/TD]

[TD]6GB DDR3[/TD]

[TD]1537 Mhz OC CL9[/TD]

[TD]HDD[/TD]

[TD]Win 10 64bit[/TD]

[TD]3m 49s[/TD]

[TD]6m 14s[/TD]

[/TR]

[TR]

[TD]Intel Pentium G3258[/TD]

[TD]4200 Mhz OC[/TD]

[TD]8GB DDR3[/TD]

[TD]1600 Mhz CL9[/TD]

[TD]SSD[/TD]

[TD]Win 7 64bit[/TD]

[TD]3m 53s[/TD]

[TD]7m 24s[/TD]

[/TR]

[TR]

[TD]Intel i7-3820[/TD]

[TD]3600 Mhz[/TD]

[TD]16GB DDR3[/TD]

[TD]1600 Mhz[/TD]

[TD]HDD[/TD]

[TD]Win 10 64bit[/TD]

[TD]4m 4s[/TD]

[TD]6m 57s[/TD]

[/TR]

[TR]

[TD]Intel i5-3570K[/TD]

[TD]3400 Mhz[/TD]

[TD]8GB DDR3[/TD]

[TD]1600 Mhz CL9[/TD]

[TD]SSD[/TD]

[TD]Win 7 64bit[/TD]

[TD]4m 7s[/TD]

[TD]6m 48s[/TD]

[/TR]

[TR]

[TD]Intel i5-750[/TD]

[TD]3887 Mhz[/TD]

[TD]8GB DDR3[/TD]

[TD]1333 Mhz[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]4m 7s[/TD]

[TD]6m 52s[/TD]

[/TR]

[TR]

[TD]Intel i5-2500K[/TD]

[TD]3300 Mhz (Turbo 3600Mhz)[/TD]

[TD]8GB DDR3[/TD]

[TD]1600 Mhz[/TD]

[TD]SSD[/TD]

[TD]Win 7 64bit[/TD]

[TD]4m 16s[/TD]

[TD]7m 19s[/TD]

[/TR]

[TR]

[TD]Intel i3-4150[/TD]

[TD]3500 Mhz[/TD]

[TD]8GB DDR3[/TD]

[TD]800 Mhz[/TD]

[TD]HDD[/TD]

[TD]Win 10 64bit[/TD]

[TD]4m 16s[/TD]

[TD]7m 35s[/TD]

[/TR]

[TR]

[TD]Intel i7-860[/TD]

[TD]3010 Mhz OC[/TD]

[TD]8GB DDR3[/TD]

[TD]800 Mhz[/TD]

[TD]SSD[/TD]

[TD]Win 10 64bit[/TD]

[TD]4m 31s[/TD]

[TD]7m 29[/TD]

[/TR]

[TR]

[TD]Intel i7-2670QM[/TD]

[TD]2200 Mhz (Turbo 3100Mhz)[/TD]

[TD]8BG DDR3[/TD]

[TD]1600 Mhz[/TD]

[TD]HDD[/TD]

[TD]Win 7 64bit[/TD]

[TD]5m 01s[/TD]

[TD]8m 11s[/TD]

[/TR]

[TR]

[TD]AMD Phenom II X4[/TD]

[TD]3600 Mhz[/TD]

[TD]8GB DDR3[/TD]

[TD]1333 Mhz[/TD]

[TD]SSD[/TD]

[TD]Win 7 64bit[/TD]

[TD]5m 56s[/TD]

[TD]9m 29s[/TD]

[/TR]

[TR]

[TD]Intel i5-2410M[/TD]

[TD]2301 Mhz[/TD]

[TD]8GB DDR3[/TD]

[TD]1333 Mhz[/TD]

[TD]SSD[/TD]

[TD]Win 7 64bit[/TD]

[TD]6m 26s[/TD]

[TD]13m 28s[/TD]

[/TR]

[TR]

[TD]Intel Core2Duo[/TD]

[TD]2400 Mhz[/TD]

[TD]4GB DDR3[/TD]

[TD]1067 Mhz[/TD]

[TD]HDD[/TD]

[TD]OSX Yosemite[/TD]

[TD]12m 19s[/TD]

[TD]23m 48s[/TD]

[/TR]

</tbody>[/TABLE]

Any preferences on the table format? I've gone with ordering by Test A results but could just as easily order by Test B as that is the more strenuous test or even split into two tables for each test.

underestimating AMD ? :herman:

Link to post
Share on other sites
  • 2 weeks later...
CFuller

CFuller

Posted
Posted

I've just run both benchmark tests on my coming-up-to-6-year-old laptop. I'll hopefully be installing a new SSD at some point this coming week, and when that's done, I'll run a couple more tests to see if there's any noticeable difference.

CPU: Intel Core i3 330M

CPU Frequency: 2133 MHz

RAM: 8GB DDR3

OS: Win 10 64-bit

Storage: HDD

Benchmark A - 9 min 25 sec

Benchmark B - 16 min 37 sec

Link to post
Share on other sites
  • 2 weeks later...
CFuller

CFuller

Posted
Posted

I've now installed a SSD in my laptop, and here are my latest results with it:

CPU: Intel Core i3 330M

CPU Frequency: 2133 MHz

RAM: 8GB DDR3

OS: Win 10 64-bit

Storage: SSD

Benchmark A - 9 min 9 sec

Benchmark B - 16 min 2 sec

Link to post
Share on other sites
  • 2 weeks later...
Earnie is God!

Earnie is God!

Posted
Posted

This is my crappy laptop which I bought refurbished last summer.

CPU: Intel Core2 Duo P8400

CPU Frequency: 2260 MHz

RAM: 4.00GB Dual-Channel DDR2 @ 398MHz (6-6-6-18)

OS: Win 10 Home 64-bit

Storage: HDD

Test A

Time: 9 min 56 sec

Test B

Time: 16 min 37 sec

Glad to see it hasn't come bottom of the table though!

I'm having a new laptop delivered next week and will do the test with that one when it arrives; should half the time or better.

Link to post
Share on other sites
  • 2 weeks later...
Earnie is God!

Earnie is God!

Posted
Posted

CPU: Intel i5-6200u dual core

CPU Frequency: 2.3GHz (turbo 2.8GHz)

RAM: 8.00GB Single-Channel DDR3 @ 798MHz (11-11-11-28)

OS: Win 10 Home 64-bit

Storage: HDD

Test A

Time: 7 min 1 sec

Test B

Time: 11 min 46 sec

Not as quick as I had hoped but still a considerable improvement on my old laptop

Link to post
Share on other sites
  • 4 weeks later...
Powermonger

Powermonger

Posted
Posted

Just returning to FM after about a 5 year absence playing, this thread has been useful in deciding whether I needed to upgrade my PC or not.

My results:

CPU: Intel i5-4570 quad core

CPU Frequency: 3.2GHz

RAM: 16.00GB Dual-Channel DDR3 @ 1600MHz

OS: Win 10 Pro 64-bit

Storage: SSD

Test A

Time: 3 min 40 sec

Test B

Time: 5 min 53 sec

It looks like I am OK for now, but a long way from being close to the quickest.

 

Link to post
Share on other sites
  • 1 month later...
  • 3 months later...
  • 4 weeks later...
Powermonger

Powermonger

Posted
Posted
On 30/07/2016 at 12:14, Powermonger said:

Just returning to FM after about a 5 year absence playing, this thread has been useful in deciding whether I needed to upgrade my PC or not.

My results:

CPU: Intel i5-4570 quad core

CPU Frequency: 3.2GHz

RAM: 16.00GB Dual-Channel DDR3 @ 1600MHz

OS: Win 10 Pro 64-bit

Storage: SSD

Test A

Time: 3 min 40 sec

Test B

Time: 5 min 53 sec

It looks like I am OK for now, but a long way from being close to the quickest.

 

Recently I upgraded my PC to the following specs and re-ran the FM16 benchmark files:

CPU: Intel i7-6700K

CPU Frequency: 4.0GHz

RAM: 16.00GB Dual-Channel DDR4 @ 3200MHz

OS: Win 10 Pro 64-bit

Storage: M.2 SSD NVMe Samsung EVO 960 500GB

Test A

Time: 2 min 24 sec

Test B

Time: 3 min 46 sec

Quiet an improvement over my old PC build. 

On 05/12/2016 at 15:15, eple said:

Anyone want to set up this for FM17?

Eager to see how FM17 performed on my new PC, I recreated Barside's FM16 benchmark savegames in FM17 but got worse results. 

Benchmark A: 2 minutes 56 seconds

Benchmark B: Not tested  

Not sure why it is slower, new save game has been setup with 116 leagues from 51 nations on a large database, resulting in an estimated 167,000 players. I made the same detail settings to the EPL as per the past benchmark setups. So it should be like for like and run on a non-modified FM17 installation  

When I get time I'll complete the Benchmark B file setup and upload the files for others to try later today. 

Link to post
Share on other sites
eple

eple

Posted
Posted
7 hours ago, Powermonger said:

Eager to see how FM17 performed on my new PC, I recreated Barside's FM16 benchmark savegames in FM17 but got worse results. 

Benchmark A: 2 minutes 56 seconds

Benchmark B: Not tested  

Not sure why it is slower, new save game has been setup with 116 leagues from 51 nations on a large database, resulting in an estimated 167,000 players. I made the same detail settings to the EPL as per the past benchmark setups. So it should be like for like and run on a non-modified FM17 installation  

When I get time I'll complete the Benchmark B file setup and upload the files for others to try later today. 

I recently upgraded some parts and have a similar set up

Went from i5-2500k, 8 Gb/1600mhz ram to i7-6700k and 16gb/3000mhz ram

Did a test on FM before and after the upgrade where I simulated three weeks.

I ran it multiple times and got similar but not identical results.

Quickest run through before was 6 min 20 sec

With new parts it was 3 min 55 sec

then I overclocked the CPU to 4.7ghz which brought it down to 3 min 35 sec.

I would share the save, but I'm on the private beta build, so I assume any save I make won't be compatible (and a breach of the NDA if shared).

Link to post
Share on other sites
Powermonger

Powermonger

Posted
Posted
1 hour ago, fabioke said:

2017 Thread!

Already up, you can find it here:

5 hours ago, eple said:

I recently upgraded some parts and have a similar set up

Went from i5-2500k, 8 Gb/1600mhz ram to i7-6700k and 16gb/3000mhz ram

Did a test on FM before and after the upgrade where I simulated three weeks.

I ran it multiple times and got similar but not identical results.

Quickest run through before was 6 min 20 sec

With new parts it was 3 min 55 sec

then I overclocked the CPU to 4.7ghz which brought it down to 3 min 35 sec.

I would share the save, but I'm on the private beta build, so I assume any save I make won't be compatible (and a breach of the NDA if shared).

Interesting results. How many players are in the file you tested? I haven't tried overclocking yet, been concentrating on getting a baseline for comparison. 

Link to post
Share on other sites
eple

eple

Posted
Posted
1 minute ago, Powermonger said:

Interesting results. How many players are in the file you tested? I haven't tried overclocking yet, been concentrating on getting a baseline for comparison. 

It was a save that was 7 seasons in so I'm unsure about player count (and if it matters that the save is many years in). Starting DB was around 70-80k players and I ran it for 3 weeks starting at the 30th of June.

Overclocking the 6700k was very easy by the way.. and I haven't really pushed it. CPU usually hovers 35-45c and 50% fan speed. During heavy load it goes up to around 55-60c and 70% fan speed keeps it in check. I have a Noctua cooler.

Link to post
Share on other sites
Powermonger

Powermonger

Posted
Posted
40 minutes ago, eple said:

It was a save that was 7 seasons in so I'm unsure about player count (and if it matters that the save is many years in). Starting DB was around 70-80k players and I ran it for 3 weeks starting at the 30th of June.

Overclocking the 6700k was very easy by the way.. and I haven't really pushed it. CPU usually hovers 35-45c and 50% fan speed. During heavy load it goes up to around 55-60c and 70% fan speed keeps it in check. I have a Noctua cooler.

Ok, just curious. The FM16 benchmark file in this thread has 188,000 players and the FM17 one I made and tested has 167,000 players but processed slower. It also opened much slower in FM Genie Scout when I was looking at player and staff counts in comparison to the FM16 file.

if I get time tonight I'll do some overclocking. I have a Corsair H100i V2 liquid cooler and CPU temp is usually between 26 and 30 deg but jumps to about 36 when processing FM. 

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Go to topic listing

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...