FM17 Performance Benchmarking Thread

[bennytee]

Just been reinstalling windows on a friends laptop and thought i'd run the benchmark on it, spec is as follows

CPU: Intel Celeron 900

CPU Frequency: 2200 MHz

RAM: 2GB ddr2

OS: Win XP 32-bit

Storage: HDD

Test A

Time: 37 min 51 sec

Test B

Time: If I get a spare hour or 3 i'll let you know

 

Luckily for him he doesn't play FM and only uses his laptop for word processing and browsing

[m@rk]

that is some patience!

[Powermonger]

7 hours ago, bennytee said:

Just been reinstalling windows on a friends laptop and thought i'd run the benchmark on it, spec is as follows

We can't move on until you complete Test B 

Do it for the team!

[bennytee]

2 hours ago, Powermonger said:

We can't move on until you complete Test B 

Do it for the team!

 

I actually did but then i had to do a few things and my mind wandered, when I got back it was done so its anywhere between 45 mins and 2 hours!  

It's almost in real time, you could argue it's more realistic  

I loaded up a game with just the Welsh Premier loaded and the speed was fine.  Just goes to show that FM will pretty much work on any system if you limit the leagues and player count.

[giblets_of_fire]

On 27/03/2017 at 22:47, giblets_of_fire said:

Well I decided that my laptop could use a fresh start & did a fresh install of Windows (10). Having installed literally just the NVidia drivers and steam my time has gone from 5:56 to 4:54 on test A so something was definitely slowing it down.

Now I just need to determine if it was some software I had installed or one of the myriad of graphics, mods etc that I have on the game (used a completely vanilla version for 4:54)

At least I know the hardware is kosher........

So im aware that im in danger of hijacking this thread but i think this is interesting!

Once i'd got my new windows install fully up and running my 6700HQ with 16gb DDR4 was taking 5:10 to 5:25 to do Test A. However there was weird quirk where if i ran it immediately when turning it on after work (i.e switched off all day) it would only take 4:35ish 

This mirrored a trend i'd had last year with Dirt Rally where the frame rate would dive by 20% after a few minutes play for no apparent reason. So on Friday night (im that cool!) i realised that i hadn't updated the BIOS since i got the laptop so as a speculative punt i updated to a new version from the MSI website.

Well now that discrepancy is still there but the initial run takes 4:10 and the ones after between 4:20 & 4:25, so found another whole minute!

By reinstalling Windows and updating the BIOS i've gone from 6:10 to 4:20 for Test A. Quite the improvement!!

Edit: Done 3 tests on Benchmark B and they come back between 7:55 & 8:10, very quick!

[lahara]

The poor gaming reviews of ryzen forced me back in the arms of intel, so here are my benchmarks for that system, tried to run it at different OC's

CPU: I5-7600k

CPU Frequency: 4000 MHz

RAM: 16GB DDR4 2133 MHz

OS: Win 10 64-bit

Storage: SSD

Test A

Time: 3 min 15 sec

Test B

Time: 6 min 38 sec

 

CPU: I5-7600k

CPU Frequency: 4000 MHz

RAM: 16GB DDR4 3000 MHz

OS: Win 10 64-bit

Storage: SSD

Test A

Time: 3 min 04 sec

Test B

Time: 6 min 04 sec

 

CPU: I5-7600k

CPU Frequency: 4600 MHz

RAM: 16GB DDR4 3000 MHz

OS: Win 10 64-bit

Storage: SSD

Test A

Time: 2 min 49 sec

Test B

Time: 5 min 35 sec

[myHorses2You]

My results:

CPU: AMD FX-8350 

CPU Frequency: 4.0GHz

RAM: 16 GB Dual-Channel DDR4 @ 3200MHz

OS: Win 10 Home 64-bit

Storage: SSD Samsung EVO 850 - 250GB

FM17 Benchmark A

Time: 5 min 34 sec

FM17 Benchmark B

Time: 9 min 25 sec

[Powermonger]

Thanks @lahara and @myHorses2You, your tests have been added now to the table.

I added all your permeations Lahara as others may find them useful. Your last 7600K test performed the same as @eple 6700K build I noticed.

@bennytee I added your Celeron test too, just in case there are some sadists older laptop users who wish to also play FM.

We have quite a good cross-section of differing PC builds now.

3 hours ago, giblets_of_fire said:

So im aware that im in danger of hijacking this thread but i think this is interesting!

Good to add that information. Someone later reading this thread wondering why their computer is running slower than someone else with a similar setup, they might find that information useful and solve their problem.

[bennytee]

On 01/04/2017 at 14:09, Powermonger said:

 

@bennytee I added your Celeron test too, just in case there are some sadists older laptop users who wish to also play FM.

We have quite a good cross-section of differing PC builds now.

 

  I'm looking at an old Athlon 64 desktop in the corner of my room at the moment.  Dare I put myself through the torment??!!!!

I do wonder whether for next years Benchmark, assuming there is one, might it not be a good idea to list the model of laptop and maybe have these results seperate?

I say this because the amount of overheating, throttling etc seems to vary wildly across brands.  I like to give as much advice as I can to people over on the Laptop buying advice thread and it would be great to say "we've benchmarked this with the actual game on this specific laptop" rather than having to give educated guesses based on hardware specs and supposed reputation. 

Just a thought.

[Powermonger]

2 hours ago, bennytee said:

  I'm looking at an old Athlon 64 desktop in the corner of my room at the moment.  Dare I put myself through the torment??!!!!

I do wonder whether for next years Benchmark, assuming there is one, might it not be a good idea to list the model of laptop and maybe have these results seperate?

I say this because the amount of overheating, throttling etc seems to vary wildly across brands.  I like to give as much advice as I can to people over on the Laptop buying advice thread and it would be great to say "we've benchmarked this with the actual game on this specific laptop" rather than having to give educated guesses based on hardware specs and supposed reputation. 

Just a thought.

I have my old PC and wife's laptop I could probably benchmark, just got to get around to do it

Good idea about adding a new column to record laptop model, especially when laptop efficiency can be so varied. I may even update this years table with that information if I can get it from the testers.

[m@rk]

10 hours ago, Powermonger said:

I have my old PC and wife's laptop I could probably benchmark, just got to get around to do it

Good idea about adding a new column to record laptop model, especially when laptop efficiency can be so varied. I may even update this years table with that information if I can get it from the testers.

my laptop - one of the slowest on the list is a Samsung RV520 (i3-2330M one) - I doubt you can still buy them, or why you'd even bother for FM!

[Harper]

Make: MSI - GL62M 7RE-623

CPU: Intel Core i7-7700HQ

CPU Frequency: 2800/3800 MHz 6MB Intel® Smart Cache (Task Manager was wavering between 3.3-3.56 GHz during testing)

RAM:  8GB DDR4 2400MHz (Task Manager never exceeded 4.4GB during testing)

OS: Win 10 Home 64-bit

Install Storage*: SSD M.2 

Save Game Storage: HDD 7200RPM

Test A

Time: 4 min 34 sec

Test B

Time: 9 min 19 sec

*Base model comes with HDD as the only storage

[Powermonger]

Thanks @m@rk and @Harper, I'll add this shortly to the table with some other info.

As per @bennytee suggestion, thought we might as well update this years results with PC/laptop details.

Just a shout out to the following people who did testing before:

@fabioke, @roykela, @eple, @Baldyloxxx, @RubenL, @Nacaw, @bananahk, @forzamr_b, @Dic, @lahara, @Urg, @jwchriste, @will221996, @Akim, @scott-h22, @Millsy2, @ukmets, @jckc221013jamie, @Oussebon, @myHorses2You
 

If you could please let me know if the computer you tested on was a custom built desktop PC, store bought PC (include make/model) or laptop (make/model), I'll add these details to the existing chart.

I have a fair idea who have desktop PCs and those with laptop builds but to maintain consistency I'd prefer not to make assumptions

Cheers guys.

[roykela]

35 minutes ago, Powermonger said:

Thanks @m@rk and @Harper, I be shortly adding this sort to the table with some other info.

As per @bennytee suggestion, thought we might as well update this years results with PC/laptop details.

Just a shout out to the following people who did testing before:

@fabioke, @roykela, @eple, @Baldyloxxx, @RubenL, @Nacaw, @bananahk, @forzamr_b, @Dic, @lahara, @Urg, @jwchriste, @will221996, @Akim, @scott-h22, @Millsy2, @ukmets, @jckc221013jamie, @Oussebon, @myHorses2You
 

If you could please let me know if the computer you tested on was a custom built desktop PC, store bought PC (include make/model) or laptop (make/model), I'll add these details to the existing chart.

I have a fair idea who have desktop PCs and those with laptop builds but to maintain consistency I'd prefer not to make assumptions

Cheers guys.

Custom built desktop PC here.

[Urg]

1 hour ago, Powermonger said:

Thanks @m@rk and @Harper, I'll add this shortly to the table with some other info.

As per @bennytee suggestion, thought we might as well update this years results with PC/laptop details.

Just a shout out to the following people who did testing before:

@fabioke, @roykela, @eple, @Baldyloxxx, @RubenL, @Nacaw, @bananahk, @forzamr_b, @Dic, @lahara, @Urg, @jwchriste, @will221996, @Akim, @scott-h22, @Millsy2, @ukmets, @jckc221013jamie, @Oussebon, @myHorses2You
 

If you could please let me know if the computer you tested on was a custom built desktop PC, store bought PC (include make/model) or laptop (make/model), I'll add these details to the existing chart.

I have a fair idea who have desktop PCs and those with laptop builds but to maintain consistency I'd prefer not to make assumptions

Cheers guys.

Custom built desktop PC.

[bennytee]

custom - 4670k

HP Envy 17 - 4710mq

My Celeron beast was an EMachines E525

[bananahk]

custom pc

[Nacaw]

i5 6600k - custom built desktop 

[Millsy2]

custom built 

[JamieTC13]

i5 7400 Custom pc 

 

[scott-h22]

On 4/6/2017 at 10:23, Powermonger said:

Thanks @m@rk and @Harper, I'll add this shortly to the table with some other info.

As per @bennytee suggestion, thought we might as well update this years results with PC/laptop details.

Just a shout out to the following people who did testing before:

@fabioke, @roykela, @eple, @Baldyloxxx, @RubenL, @Nacaw, @bananahk, @forzamr_b, @Dic, @lahara, @Urg, @jwchriste, @will221996, @Akim, @scott-h22, @Millsy2, @ukmets, @jckc221013jamie, @Oussebon, @myHorses2You
 

If you could please let me know if the computer you tested on was a custom built desktop PC, store bought PC (include make/model) or laptop (make/model), I'll add these details to the existing chart.

I have a fair idea who have desktop PCs and those with laptop builds but to maintain consistency I'd prefer not to make assumptions

Cheers guys.

Store brought laptop. 

http://www.laptopsdirect.ie/stormforce-wildfire-core-i7-6700hq-8gb-1tb-128gb-ssd-geforce-gt940-dvd-rw-7270-9069/version.asp

[Dic]

Phenom 955 and i3 4160 - custom built desktop

4700hq - laptop, Asus G56JR

[bennytee]

On 4/5/2017 at 10:45, m@rk said:

my laptop - one of the slowest on the list is a Samsung RV520 (i3-2330M one) - I doubt you can still buy them, or why you'd even bother for FM!

i'd get one, if it was cheap enough and portable for travelling, maybe just running a nation or 2.  Thats the beauty of FM, it will run on almost anything if you configure it right.

[Akim]

also custom build

[will221996]

On ‎06‎/‎04‎/‎2017 at 17:23, Powermonger said:

Thanks @m@rk and @Harper, I'll add this shortly to the table with some other info.

As per @bennytee suggestion, thought we might as well update this years results with PC/laptop details.

Just a shout out to the following people who did testing before:

@fabioke, @roykela, @eple, @Baldyloxxx, @RubenL, @Nacaw, @bananahk, @forzamr_b, @Dic, @lahara, @Urg, @jwchriste, @will221996, @Akim, @scott-h22, @Millsy2, @ukmets, @jckc221013jamie, @Oussebon, @myHorses2You
 

If you could please let me know if the computer you tested on was a custom built desktop PC, store bought PC (include make/model) or laptop (make/model), I'll add these details to the existing chart.

I have a fair idea who have desktop PCs and those with laptop builds but to maintain consistency I'd prefer not to make assumptions

Cheers guys.

custom built

[Powermonger]

Hey guys, I've revised and reformatted the results table, adding type of computer people are using and also tester name. I have also combined the two tables so Test A and Test B results are side by side for easier review.

@Harper your test results are added to the table now too.

Let me know if you see any mistakes everyone.

[bennytee]

On 4/10/2017 at 10:41, Powermonger said:

Hey guys, I've revised and reformatted the results table, adding type of computer people are using and also tester name. I have also combined the two tables so Test A and Test B results are side by side for easier review.

@Harper your test results are added to the table now too.

Let me know if you see any mistakes everyone.

Cracking work mate 

[myHorses2You]

On 06.04.2017 at 12:23, Powermonger said:

Thanks @m@rk and @Harper, I'll add this shortly to the table with some other info.

As per @bennytee suggestion, thought we might as well update this years results with PC/laptop details.

Just a shout out to the following people who did testing before:

@fabioke, @roykela, @eple, @Baldyloxxx, @RubenL, @Nacaw, @bananahk, @forzamr_b, @Dic, @lahara, @Urg, @jwchriste, @will221996, @Akim, @scott-h22, @Millsy2, @ukmets, @jckc221013jamie, @Oussebon, @myHorses2You
 

If you could please let me know if the computer you tested on was a custom built desktop PC, store bought PC (include make/model) or laptop (make/model), I'll add these details to the existing chart.

I have a fair idea who have desktop PCs and those with laptop builds but to maintain consistency I'd prefer not to make assumptions

Cheers guys.

Custom build PC..

[Powermonger]

15 hours ago, bennytee said:

Cracking work mate 

Thanks

I've gone and updated the table again with more information for each persons CPU makeup of cores, threads, cache and year released. Felt it would useful to know that detail when comparing results.

@myHorses2You I updated your build details in the table too. Cheers for that.

[jwchriste]

On 4/6/2017 at 04:23, Powermonger said:

Thanks @m@rk and @Harper, I'll add this shortly to the table with some other info.

As per @bennytee suggestion, thought we might as well update this years results with PC/laptop details.

Just a shout out to the following people who did testing before:

@fabioke, @roykela, @eple, @Baldyloxxx, @RubenL, @Nacaw, @bananahk, @forzamr_b, @Dic, @lahara, @Urg, @jwchriste, @will221996, @Akim, @scott-h22, @Millsy2, @ukmets, @jckc221013jamie, @Oussebon, @myHorses2You
 

If you could please let me know if the computer you tested on was a custom built desktop PC, store bought PC (include make/model) or laptop (make/model), I'll add these details to the existing chart.

I have a fair idea who have desktop PCs and those with laptop builds but to maintain consistency I'd prefer not to make assumptions

Cheers guys.

Both my 3930k and 7700k systems were custom built.

[will221996]

just tested with new ram. same setup as before. The ram is corsair vengeance low profile. 32 gb in 4 dimms, but in dual channel(ryzen only has dual channel) it is 3200 Mhz.

Make: Custom

CPU: AMD Ryzen r7 1700x

RAM:  32GB DDR4 3200Mhz

OS: Win 10 Pro 64-bit

Install Storage: SSD SATA

Save Game Storage: HDD 7200RPM

Test A

Time: 4 min 20 sec (10 second faster... probably margin of error)

Test B

Time: 8 min 31 sec( 1 minute faster.. very noticeable.) any devs among us to explain why faster ram(im assuming fm cant use more than 12 gb of ram(the amount previously tested)) makes such a big difference?

[bennytee]

During an Easter visit to my folks house I thought i'd give the benchmark a run on my old mans desktop that I built for him for a fraction over £230 quid. (no OS and pre Brexit referendum)

CPU: G3258 Intel Pentium Anniversary edition

CPU Frequency: 4.3Ghz OC 

RAM: 8 GB DDR3 @ 1600MHz

OS: Win 10 Home 64-bit

Storage: SSD 

3mb cache
2 cores, 2 threads
Q2'14

FM17 Benchmark A

Time: 5 min 19 sec

FM17 Benchmark B

Time: 10 min 40 sec

 

In terms of price vs performance I think its a beast really, I had it overclocked to about 4.7Ghz too at one point but toned it way down for my old man before giving it to him.

[Powermonger]

Thanks @will221996 & @bennytee, table has been updated with your test results.

[bennytee]

@Powermonger sorry to be pedantic but the Celeron 900 has 1 core 1 thread and 1mb cache. 

Released Q1'09

[Powermonger]

1 hour ago, bennytee said:

@Powermonger sorry to be pedantic but the Celeron 900 has 1 core 1 thread and 1mb cache. 

Released Q1'09

No worries, fixed now. I think I copied the info from the config above it but forgot to update the CPU details. 

[fabioke]

@Powermonger HP NOTEBOOK PAVILION "gaming" 15-AK020NB

[SebastianRO]

I just bought a new laptop and I am completely over the moon

It's a Lenovo Legion Y520: Intel Core i5 - 7300 HQ + nVIDIA GTX 1050 Ti + 8GB DDR4 @2400 MHz.

Before I bought it, I asked some people if I should get this over the i7 6700 HQ (ASUS RoG) and all of them told me that the i7 would be 20% faster. However, stubborn as I am, I went for the Lenovo Legion, which offered me a much better GPU (GTX 1050 Ti is about 75% better than the 960M) for the same price. So, I thaught "Let's try this benchmark and see how the new Kaby Lake i5 fairs over the last generation i7 Skylake" and these are the results:

I dont have an SSD in my laptop yet, keep this in mind.

Benchmark A:  04:49 sec

Benchmark B:  09:42 sec

 

As you can see in comparison, the Kabylake i5 actually edges the Skylake i7 in the first test by 2 seconds and loses in the second test by 5 seconds. Now I'm certain that if I had an SSD (like that 6700 had), the B benchmark would have been won by the i5 as well. To conclude this comment I will say this: Lenovo nailed it with the Y520 series. It's very affordable (I paid about 820 euros) and it's running, looking and feels great. Considering the power it has under it's hood, I think this one will be relevant even after 2-3 years in AAA titles. As for FM ... well ... more :).

[Powermonger]

Thanks @SebastianRO, I have updated the table now with your testing. Very little between your i5-7300HQ and the i7-6700HQ, maybe the extra 4 threads just edged ahead of your 7300HQ but there is so little between them it looks like you picked well.

[Powermonger]

Just been benchmarking my old custom PC seeing as there is no i5-4570 tested yet.

Intel i5-4570 @ 3.2 GHz 4/4 6MB

DDR3 16GB 1333 MHz

SSD

Windows 10 64-bit

Benchmark A: 4min 19 secs

Benchmark B: 8min 24 secs

Will update the table later tonight.

[Guest]

On 2017-4-24 at 12:43, Powermonger said:

Thanks @SebastianRO, I have updated the table now with your testing. Very little between your i5-7300HQ and the i7-6700HQ, maybe the extra 4 threads just edged ahead of your 7300HQ but there is so little between them it looks like you picked well.

If you're not planning on running leagues in full detail choosing an i5 over the equivalent i7 is always the way to go.

[Powermonger]

Table updated now and I also refined the CPU frequency column, adding a CPU: BASE and CPU: TURBO as what was there before was a bit inconsistent as it was a mixture of base frequencies and overclocked turbo frequencies.

[CFuller]

Make: Lenovo H30-05 (store-bought)

CPU: Intel Core i7-4790

CPU Frequency: 3.6GHz

RAM: 8GB DDR3 1600MHz

OS: Win 10 64-bit

Storage: HDD

Test A

Time: 4 min 26 sec

Test B

Time: 8 min 39 sec

[Powermonger]

@CFuller the table has been updated with your test now, thanks!

[spa]

Custom built

CPU: Intel Pentium G4400
CPU Frequency: 3300 MHz

RAM: 8GB DDR4 2133 Dual channel Kingston

OS: Win 10 64-bit

Storage: HDD

Test A
Time: 5 min 09 sec

Test B
Time: 10 min 34sec

[Powermonger]

Thanks @spa, I've updated the table with your test results.

[bennytee]

@jwchriste and/or @Powermonger 

Strange request but do you recon you guys could run test A but with the detail level set to maximum?

Somebody is asking about it on the new laptop/PC thread and i'd be curious to know how long that would take on your super charged rigs.

[Powermonger]

2 minutes ago, bennytee said:

@jwchriste and/or @Powermonger 

Strange request but do you recon you guys could run test A but with the detail level set to maximum?

Somebody is asking about it on the new laptop/PC thread and i'd be curious to know how long that would take on your super charged rigs.

Sure, I can give it a try and I'll post up the results. I'll just run it at base clock as a baseline.

[bennytee]

cheers, just ran it myself on my i7 4210mq and it took 55min 52sec

[Emperor J]

Custom built (mainly for streaming/unexpected financial windfall)

CPU: AMD Ryzen 7 1800x

CPU Frequency: 3.6GHz OC'd to 4.1GHz

RAM: 16GB DDR4 3000MHz

OS: Win 10 64-bit

Storage: HDD

Test A

Time: 5 min 12 sec

Test B

Time: 9 min 13 sec

[Powermonger]

On 5/15/2017 at 08:07, bennytee said:

cheers, just ran it myself on my i7 4210mq and it took 55min 52sec

Finally got a chance to do myself.

Benchmark A (Max/Max): 27min 14 secs

Not sure that is a setting I would want to play at with all nations and leagues selected.