FM21 Performance Benchmarking Thread

[Phil930]

Those Ryzens crush the tests - really impressive.  Now I want to build an AMD system, but as noted above, i'll hold for DDR5 also.

[ZOXEXIVO]

1 hour ago, Daveincid said:

Heyy would it be possible for you to run benchmark D? It would be great to see the difference to the 5900x:)

Ok, I will try this tomorrow

[ZOXEXIVO]

1 hour ago, matt_forest said:

What do you mean "It's a shame to see results like this" ? Your own results or other peoples benchmark times?

It is my results.

5 years ago, I already raised the topic that a game that should have initial support for multithreading is not using it well.

Having made a benchmark in 2020, I see that the vacation process takes 2-3 cores of my 16 core CPU.

I have my pet project to develop football manager simulation logic - https://github.com/ZOXEXIVO/football-simulator (without manual control, and other things). I want create it with fully support of multithreading.

 

Edited November 27, 2020 by ZOXEXIVO

[Brother Ben]

1 minute ago, ZOXEXIVO said:

It is my results.

5 years ago, I already raised the topic that a game that should have initial support for multithreading is not using it well.

Having made a benchmark in 2020, I see that the vacation process takes 2-3 cores of my 16 core CPU

Benchmark D will see to that....

[Brother Ben]

12 minutes ago, ZOXEXIVO said:

I have my pet project to develop football manager simulation logic - https://github.com/ZOXEXIVO/football-simulator (without manual control, and other things). I want create it with fully support of multithreading.

 

Okay wow that sounds interesting, I can't pretend to understand how any of it works but I'd love to hear more

[ZOXEXIVO]

13 minutes ago, Brother Ben said:

Okay wow that sounds interesting, I can't pretend to understand how any of it works but I'd love to hear more

The project is in its early stage) I create FM 2017 - style UI in html

Edited November 28, 2020 by ZOXEXIVO

[Chewy82*]

Type: Laptop

Model: ASUS VIVOBOOK 14 M413IA-EB573T

CPU Model: AMD Ryzen 5 Mobile 4500U

CPU Base Frequency: 2.3GHz

CPU Turbo Frequency: 4 GHz

RAM: 8GB

RAM Clockspeed: 3192mhz

GPU: Integrated AMD Radeon Vega Graphics

Graphics Level in 3D: Medium

Storage Type: SSD

Benchmark A: 2 Minutes 50 Seconds
Benchmark B: 12 Minutes 24 Seconds
Benchmark C: 17 Minutes 43 Seconds

[Junkhead]

On the fence re an upgrade at the moment, this thread will swing it for me one way or the other this year. Will be popping back in the new year to see if I'm about to become a fair few hundred £'s lighter!

Thanks all. Maybe. 😳😂

[Smurf]

14 hours ago, ZOXEXIVO said:

 

It's a shame to see results like this
In 2021, a game that can potentially be parallelized to all cores almost does not use them.
Apparently, everything is bad in the code and it won't be possible to fix it for a long time.

These attempts to measure the speed look pathetic, because this is the task of developers, not users.

Well I ran the last Laptop buyers guide thread and @kevhamsterkindly agreed to start it over. The first laptop thread I ran went to well over 200 pages. The one I was involved in before that was over 200 pages. And the last one went to 100 pages. And Kev is well on his way with over 20 pages already in a few weeks.

Point is there are thousands of people looking for new laptops/computers and the biggest question asked is "how many leagues can I load?"

Fact is, we dont know. It's in the opening thread. We cannot test every system.

But the benchmark thread is amazing as the results show us what is working and what is not. 

We dont work for SI or FM we are mere mortals trying to help fellow gamers. 

It really is no coincidence either that https://www.notebookcheck.net/Mobile-Processors-Benchmark-List.2436.0.html ranks the processors and it is uncanny to how similar the list is for the Benchmark thread.

With this data we can offer the thousands of people asking year in year out what is best value for their money.

 

[ZOXEXIVO]

I update my result and add D Benchmark time: 15 minutes 38 seconds (28.08 - 04.09)

I am impressed, how FM used my CPU in benchmark D: sometimes it takes 100%

[Daveincid]

vor 1 Minute schrieb ZOXEXIVO:

I update my result and add D Benchmark time: 15 minutes 38 seconds (28.08 - 04.09)

I am impressed, how FM used my CPU in benchmark D: sometimes it takes 100%

Nice! So the 5900x is only a bit faster with 12cores than the 3950x with 16cores! AMD really does a great job at the moment!

[ZOXEXIVO]

1 hour ago, Smurf said:

Fact is, we dont know. It's in the opening thread. We cannot test every system.

But the benchmark thread is amazing as the results show us what is working and what is not. 

We dont work for SI or FM we are mere mortals trying to help fellow gamers. 

I understand your good intentions, but in fact it is useless work, which will be useless every year.
All such measures should come from the developers of special units who are engaged in measuring bottlenecks in the game and optimizing them.

They have access to the code and can see slow and weak points in the game on the graphs. The game should fully utilize all your harware.
Apparently, the current FM release cycle does not allow time for optimization and the developers are only busy creating and editing new features.
In fact, managers and developers have created a free workforce for themselves that does not need to be paid.

Edited November 28, 2020 by ZOXEXIVO

[EdL]

1 minute ago, ZOXEXIVO said:

I understand your good intentions, but in fact it is useless work, which will be useless every year.
All such measures should come from the developers of special units who are engaged in measuring bottlenecks in the game and optimizing them.

They have access to the code and can see slow and weak points in the game on the graphs
Apparently, the current FM release cycle does not allow time for optimization and the developers are only busy creating and editing new features.
In fact, managers and developers have created a free workforce for themselves that does not need to be paid.

This is just totally off the mark, these guys aren't doing work for us, we have internal performance testing ofc, but this is primarily for the users to compare performance, and as aide its interesting for me as there will always be more variety than we can test with.

As for optimisation of course we are doing it!

[EdL]

Type: Desktop

CPU Model:Ryzen 9 3950X

CPU Base Frequency: 3.5GHz

CPU Turbo Frequency: 4.7 GHz

RAM: 32GB

RAM Clockspeed: 2933Mhz

GPU: Radeon 5600XT

Storage Type: SSD - Samsung 970 EVO Plus

OS: Windows 10 1909 X64

 

Benchmark A: 1 min 49 seconds

[ZOXEXIVO]

6 minutes ago, EdL said:

This is just totally off the mark, these guys aren't doing work for us, we have internal performance testing ofc, but this is primarily for the users to compare performance, and as aide its interesting for me as there will always be more variety than we can test with.

As for optimisation of course we are doing it!

Thanks for answer

[Phil930]

1 hour ago, ZOXEXIVO said:

I update my result and add D Benchmark time: 15 minutes 38 seconds (28.08 - 04.09)

I am impressed, how FM used my CPU in benchmark D: sometimes it takes 100%

Now you're just showing off with all those cores/logical processes!  Love it.

[tonidopa]

Type: Desktop
CPU Model: i5-7600
CPU Base Freq: 3.50GHz
CPU Turbo Freq: 4.1GHz
RAM: 16GB DDR4
RAM Clockspeed: 2400Mhz
GPU: Asus Dual 1060 GTX OC 6GB DDR5
Storage Type: SSD
OS: Win 10 Pro x64
3D GRAPHICS: High

• Benchmark A: 2:49.04
• Benchmark B: 11:27.68
• Benchmark C: 19:58.40
• Benchmark D: 59:21.41

Edited November 28, 2020 by tonidopa
Add benchmark times

[Daveincid]

@EdL @ZOXEXIVO

It would be interessting to see if XMP-Profile activated makes a big difference.  I noticed quite a big improvement with 3600Mhz CL18 vs default in FM20 on my 3950x.

[Smurf]

1 hour ago, ZOXEXIVO said:

I understand your good intentions, but in fact it is useless work, which will be useless every year.
All such measures should come from the developers of special units who are engaged in measuring bottlenecks in the game and optimizing them.

They have access to the code and can see slow and weak points in the game on the graphs. The game should fully utilize all your harware.
Apparently, the current FM release cycle does not allow time for optimization and the developers are only busy creating and editing new features.
In fact, managers and developers have created a free workforce for themselves that does not need to be paid.

At the end of the day I am not doing it for SI. Nine years ago I bought a computer for FM and it didnt work. I spent close to a grand on it. Wasted money. As fM is the only game I play, I put stock into understanding the game and required specs for a good solution.

You can buy two computers for the same price. One could be severely underpowered and not suitable for FM.

I really dont want others to fall into the same trap. Once you buy the computer you really arent in a position to buy again. Its a lot of money.

All i am doing is advising what people spend their 500, 600, 700,800 etc quid on.

And the benchmark thread helps me and others understand what we are spending money on, more importantly, other peoples money.

I am sure @EdLand others appreciate it is literally for our own use and if it helps the SI team somehow then thats even better.

So i dont know why you are grouchy about this.

The threads help everyone. And its my time to waste.

[EdL]

My personal machine

Type: Desktop

CPU Model:Ryzen 7 2700X

CPU Base Frequency: 3.7GHz

CPU Turbo Frequency: 4 GHz

RAM: 16GB

RAM Clockspeed: 2933Mhz

GPU: NVIDIA GeForce GTX 1060 6GB

Storage Type: SSD - Samsung 970 EVO

OS: Windows 10 1909 X64 (additional info but could be good to know as it can have an affect)

Benchmark A: 2 min 22 Sec

Benchmark B: 10 min 36 Sec

 

[tonidopa]

Wonder if this can be measured in Stadia as well?

[xatharas]

Stupid question, but I'm new to FM and benchmarking, how do I start the benchmarks? I load the game and then what do I do? I got all 4 saves and can start them.

 

Thx in advance

 

Edited November 28, 2020 by xatharas

[Brother Ben]

22 minutes ago, xatharas said:

Stupid question, but I'm new to FM and benchmarking, how do I start the benchmarks? I load the game and then what do I do? I got all 4 saves and can start them.

 

Thx in advance

 

go to go on holiday

 

then just click go on holiday (it should be a week default)

 

Use a stopwatch to record the times

 

[matt_forest]

5 hours ago, ZOXEXIVO said:

I understand your good intentions, but in fact it is useless work, which will be useless every year.
All such measures should come from the developers of special units who are engaged in measuring bottlenecks in the game and optimizing them.

They have access to the code and can see slow and weak points in the game on the graphs. The game should fully utilize all your harware.
Apparently, the current FM release cycle does not allow time for optimization and the developers are only busy creating and editing new features.
In fact, managers and developers have created a free workforce for themselves that does not need to be paid.

(A) If one person avoids buying a crappy computer after looking at some of the benchmarks on here then to me that proves it's not a useless excercise.

(B) I get your point about the release cycle, that's the problem with a sports game as they're always going to be tied to the start of a new football season starting. I think FM and FIFA could benefit in the long run if every now and then they came out with an update rather than a full blown game.

[xatharas]

Type: Desktop

CPU Model:Ryzen 7 3700X

RAM: 32GB

RAM Clockspeed: 3600Mhz

GPU: NVIDIA GeForce RTX 2070 Super 8GB

Storage Type: SSD - Samsung 970 EVO

 

 

A: 1m55

B: 8m55

C: 10m56

D: 21m38

[Gee_Simpson]

Looks like AMD are the undisputed kings and value for money now!

I'll post my results up in a bit.

[VP.]

Type: Desktop

CPU Model: i7 - 8700

CPU Base Frequency: 3.20GHz

CPU Turbo Frequency: 4.60GHz

RAM: 16GB

RAM Clockspeed: 2400Mhz

GPU: NVIDIA GeForce 1050TI- 4GB

Graphics Level in 3D: High

Storage Type: SSD

 

Benchmark A: 2 min 06 Sec

Benchmark B: 9 min 45 Sec

Benchmark C: 13 min 08 Sec

[ScrambledEgg]

Haven't seen many older desktop system results posted so far so thought I'd chime in:

Type: Desktop

Model: Custom

CPU Model: i7 - 4770K

CPU Base Frequency: 3.5GHz

CPU Turbo Frequency: 4.40 GHz (OC)

RAM: 16GB

RAM Clockspeed: 1600Mhz (DDR3)

GPU: GTX 1650 Super 4GB

Graphics Level in 3D: Very High

Storage Type: NVMe SSD (via PCI Express Card)

Benchmark A: 02 min 33 Sec

Benchmark B: 10 min 48 Sec

Benchmark C: 16 min 15 Sec

Benchmark D: 43 min 02 Sec

 

CPU has all core overclock @ 4.4Ghz on an AIO liquid cooler - I ran Benchmark A at stock for comparison and the result was 2 min 54 sec, I will run the others at stock in the next few days as I'm interested to see how much slower bench D will be.

I had some CPU monitoring software running and logging in the background and there was only really bench D that had any sustained periods at 100% usage, the others have spikes but never for more than a few seconds (maybe 20 seconds at times on bench C).

Purchased FM every year since the mid 90's (Championship Manager days) and played on this system since 2013 - if any optimisation has been made to the processing it has been very subtle, after a few weeks of playing 2021 (about 25 hours), it feels about the same as it always has on this system although I accept that the level of data and analysis has increased so perhaps SI should be applauded for that.

With an average setup of leagues and medium database I've never felt the experience is bad; I would like to be able to load a larger database however I find that in doing this the performance is fine for the first five seasons or so but then just gets progressively worse every season to the point it is very sluggish, say, 20 seasons in - obviously this is referring to previous editions, I don't have that kind of time to play these days!

 

[Brother Ben]

3 hours ago, ScrambledEgg said:

Haven't seen many older desktop system results posted so far so thought I'd chime in:

Type: Desktop

Model: Custom

CPU Model: i7 - 4770K

CPU Base Frequency: 3.5GHz

CPU Turbo Frequency: 4.40 GHz (OC)

RAM: 16GB

RAM Clockspeed: 1600Mhz (DDR3)

GPU: GTX 1650 Super 4GB

Graphics Level in 3D: Very High

Storage Type: NVMe SSD (via PCI Express Card)

Benchmark A: 02 min 33 Sec

Benchmark B: 10 min 48 Sec

Benchmark C: 16 min 15 Sec

Benchmark D: 43 min 02 Sec

 

CPU has all core overclock @ 4.4Ghz on an AIO liquid cooler - I ran Benchmark A at stock for comparison and the result was 2 min 54 sec, I will run the others at stock in the next few days as I'm interested to see how much slower bench D will be.

I had some CPU monitoring software running and logging in the background and there was only really bench D that had any sustained periods at 100% usage, the others have spikes but never for more than a few seconds (maybe 20 seconds at times on bench C).

Purchased FM every year since the mid 90's (Championship Manager days) and played on this system since 2013 - if any optimisation has been made to the processing it has been very subtle, after a few weeks of playing 2021 (about 25 hours), it feels about the same as it always has on this system although I accept that the level of data and analysis has increased so perhaps SI should be applauded for that.

With an average setup of leagues and medium database I've never felt the experience is bad; I would like to be able to load a larger database however I find that in doing this the performance is fine for the first five seasons or so but then just gets progressively worse every season to the point it is very sluggish, say, 20 seasons in - obviously this is referring to previous editions, I don't have that kind of time to play these days!

 

You make a good point about it being slower down the line, what i've had to do is drop leagues the further in I get.  

I've got an i5 4670k from the same generation and I still find it fine but the results on here make me want to upgrade I have to say

[crghurst]

Type: Laptop

Model: PC Specialist Build

CPU Model: AMD Ryzen 7 4800H

CPU Base Frequency: 2.9 GHz

CPU Turbo Frequency: 4 .2 GHz

RAM: 16 GB

RAM Clockspeed: 2660 Mhz

GPU: NVIDIA GeForce RTX 2060

Graphics Level in 3D: Very High

Storage Type: SSD M.2

 

Benchmark A: 2 min 21 Sec

Benchmark B:  12 min 16 Sec

Benchmark C: 14 min  18 Sec

Benchmark D: 27 min  58 Sec

[ScrambledEgg]

2 hours ago, Brother Ben said:

You make a good point about it being slower down the line, what i've had to do is drop leagues the further in I get.  

I've got an i5 4670k from the same generation and I still find it fine but the results on here make me want to upgrade I have to say

Yeah if the 5950X becomes available soon I think I may have to get myself a new system for Christmas, I could squeeze a few more years out of my current set up but the results of others on this thread kind of make me realise I will actually notice a difference.

[Gee_Simpson]

Well after seeing the i7 9700k results I'm absolutely delighted with how my i9 9900k (which isn't that much better) performed. I didn't bother with the overclock to 5ghz on all cores this time like I did for the FM20 test as I don't use my PC overclocked so didn't see the point, although it would obviously have given me an even better result!

Type: Desktop

Model: Custom Build

CPU Model: i9 9900k

CPU Base Frequency: 3.6 GHz

CPU Turbo Frequency: 5 GHz

RAM: 16GB

RAM Clockspeed: 3200Mhz

GPU: NVIDIA GeForce GTX 1660 - 6GB

Graphics Level in 3D: Very High

Storage Type: SSD

 

Benchmark A: 1 min 38 sec

Benchmark B: 7 min 7 sec

Benchmark C: 8 min 54 sec

Benchmark D: 19 min 53 sec

 

So yeah, just that little advantage the i9 9900k has over the i7 9700k shows in Bench D especially! Obviously Bench D still lags behind the Ryzens but I'm really happy with the results! Happy to get under 20 mins in Bench D. This will certainly do me for a good few years I think, which is exactly what I was hoping for when I purchased this processor around a year ago.

After comparing the other 3 Benchmarks to the Ryzens, my i9 9900k mostly keeps up or sometimes betters them up until the last Benchmark. I don't ever run full detail anyway, so it doesn't bother me much really. Only in Bench D the Ryzens have a considerable advantage.

Edited November 30, 2020 by Gee_Simpson

[ScrambledEgg]

Further to my post yesterday, and in what feels like an act of self-harm, I have run all the benchmarks at stock to see what sort of performance difference there is on a 4770K - 15%(ish) is quite a jump so would definitely recommend anyone with older hardware to sling an overclock on to eek a bit more performance out if your CPU allows it - provided you're not running a stock cooler.

I don't do anything special, I just use Intel XTU, add 0.1 to the voltage and slide all cores up to 4.4. I have had it stable under a stability test up to 4.6Ghz but I was getting the odd crash in games so pulled it back to 4.4 given it doesn't really make much difference in GPU heavy titles anyway.

[Brother Ben]

13 minutes ago, ScrambledEgg said:

Further to my post yesterday, and in what feels like an act of self-harm, I have run all the benchmarks at stock to see what sort of performance difference there is on a 4770K - 15%(ish) is quite a jump so would definitely recommend anyone with older hardware to sling an overclock on to eek a bit more performance out if your CPU allows it - provided you're not running a stock cooler.

I don't do anything special, I just use Intel XTU, add 0.1 to the voltage and slide all cores up to 4.4. I have had it stable under a stability test up to 4.6Ghz but I was getting the odd crash in games so pulled it back to 4.4 given it doesn't really make much difference in GPU heavy titles anyway.

Quite a difference there to be fair. 

I however lost out badly in the silicon lottery on my 4670k and can only get it stable at 4.1ghz.  Annoying but that's the way it goes I suppose

[Brother Ben]

10 hours ago, Gee_Simpson said:

After comparing the other 3 Benchmarks to the Ryzens, my i9 9900k mostly keeps up or sometimes betters them up until the last Benchmark. I don't ever run full detail anyway, so it doesn't bother me much really. Only in Bench D the Ryzens have a considerable advantage.

This is very much what I expected going in to this but the difference is this year that I think the gap is getting very small indeed

[ScrambledEgg]

33 minutes ago, Brother Ben said:

I however lost out badly in the silicon lottery on my 4670k and can only get it stable at 4.1ghz.  Annoying but that's the way it goes I suppose

Yeah it's probably the first CPU I've been able to get anything meaningful out of, I did forget to mention the AIO pump and fan ramp up a bit when overclocked and under load though, a bit louder than I can put up with if I'm having a marathon FM session.

[dylan.gangat]

Type: Custom Build

Model: Desktop

CPU Model: AMD Ryzen 5 3500X 

CPU Base Frequency: 3.6GHz

CPU Turbo Frequency: 4.1GHz

RAM: 16GB

RAM Clockspeed: 3200 mhz

GPU: Nvidia GeForce RTX 2060

Graphics Level in 3D: Very High

Storage Type: SSD

 

Benchmark A: 02:12

Benchmark B:  09:23

Benchmark C: 13:56

Benchmark D: 

[Brother Ben]

1 hour ago, ScrambledEgg said:

Yeah it's probably the first CPU I've been able to get anything meaningful out of, I did forget to mention the AIO pump and fan ramp up a bit when overclocked and under load though, a bit louder than I can put up with if I'm having a marathon FM session.

It's a balancing act

At the moment I play mostly on my commute with a Windows tablet, so I've basically got 4 systems but I have to tailor my set up to my weakest system - the tablet but i think i'm going to have to maybe have 2 saves on the go as i'm really getting the itch to build another rig

[JamieTC13]

I am going to run the test again with an all core overclock of 5ghz from 4.8ghz just to see if it will be any better I will post the results as a comparison  

[Brother Ben]

Going to start the spreadsheet tomorrow folks, thanks for all the results so far  

[ScrambledEgg]

This was a fun little experiment, result nothing to write home about though...clock speed is just too low on these server CPUs.

[Brother Ben]

3 hours ago, ScrambledEgg said:

This was a fun little experiment, result nothing to write home about though...clock speed is just too low on these server CPUs.

Can't wait to see benchmark D on a server CPU

190Gb of RAM though   Might need its own section on my spreadsheet

Edited November 30, 2020 by Brother Ben

[judicious]

Type: Laptop

Model: MSI GS66 Stealth

CPU Model: i7 - 10750H

CPU Base Frequency: 2.6GHz

CPU Turbo Frequency: 5.00 GHz

RAM: 16GB

RAM Clockspeed: 2666Mhz

GPU: NVIDIA RTX 2060 - 6GB

Graphics Level in 3D: High

Storage Type: SSD

OS: Windows 10 Home 64-bit

Benchmark A: 03 min 00 Sec

Benchmark B: 13 min 44 Sec

Benchmark C: 18 min 56 Sec

 

I must admit I was expecting better results for this laptop, especially since having browsed through the thread I found SBII's results below which are a lot better.

I know the GPU is different and that on paper the RTX 2070 Max-Q is better than the RTX 2060 but my understanding is there is very little in it so any ideas as to the difference in results? Also my (admittedly limited) understanding of how this works is that the GPU should have minimal influence on holidaying in game. 

 

On 27/11/2020 at 16:37, SBII said:

Type: Laptop

Model: MSI GS66

CPU Model: i7-10750H

CPU Base Frequency: 2.6 GHz

CPU Turbo Frequency: 5.0 GHz

RAM: 16 GB

RAM Clockspeed: 2666 Mhz

GPU: Intel Iris 6100 - RTX 2070 Max-Q

Graphics Level in 3D: Very High

Storage Type: SSD

 

Benchmark A: 2 min 8 sec

Benchmark B: 9 min 25 sec

Benchmark C: 13 min 9 sec

 

Edited December 1, 2020 by judicious
Typo

[EdL]

7 minutes ago, judicious said:

Type: Laptop

Model: MSI GS66 Stealth

CPU Model: i7 - 10750H

CPU Base Frequency: 2.6GHz

CPU Turbo Frequency: 5.00 GHz

RAM: 16GB

RAM Clockspeed: 2666Mhz

GPU: NVIDIA RTX 2060 - 6GB

Graphics Level in 3D: High

Storage Type: SSD

OS: Windows 10 Home 64-bit

Benchmark A: 03 min 00 Sec

Benchmark B: 13 min 44 Sec

Benchmark C: 18 min 56 Sec

 

I must admit I was expecting better results for this laptop, especially since having browsed through the thread I found SBII's results below which are a lot better.

I know the GPU is different and that on paper the RTX 2070 Max-Q is better than the RTX 2060 but my understanding is there is very little in it so any ideas as to the difference in results? Also my (admittedly limited) understanding of how this works is that the GPU should have minimal influence on holidaying in game. 

 

 

GPU has zero effect.

I'd say if an equivalent CPU is besting you things to look at overheating (i.e. do your fans need cleaning out), are you in the correct power plan to try and make syure the CPU is getting maximum juice and things like chipset drivers being upto date

[judicious]

10 minutes ago, EdL said:

GPU has zero effect.

I'd say if an equivalent CPU is besting you things to look at overheating (i.e. do your fans need cleaning out), are you in the correct power plan to try and make syure the CPU is getting maximum juice and things like chipset drivers being upto date

 That's what I thought re the GPU.

My fans are definitely clean as this is quite a new laptop and I've barely used it and I only use it on my laptop tray.

Also using the extreme performance power plan and all chipset drivers are up to date so completely stumped as to what could be causing the issue!

[SBII]

2 hours ago, judicious said:

 That's what I thought re the GPU.

My fans are definitely clean as this is quite a new laptop and I've barely used it and I only use it on my laptop tray.

Also using the extreme performance power plan and all chipset drivers are up to date so completely stumped as to what could be causing the issue!

Hmm. Have you switched to "Ultimate Performance" in Power Options?

 

EDIT: 

So i have hust redone test A and this time it was a little slower at 2:17 minutes. but still somewhat faster that your gs66.

This is how i has set up my laptop for the test:

User scenario: Extreme performance (with the GPU overclocked, but it shouldn't matter?)

Power plan: Ultimate Performance

Gaming Mode: ON

Other than that i have limited the framerate to 60 fps. i dont know if this help with the temperature and leads to a little less thermal throtteling?

 

 

Edited December 1, 2020 by SBII

[Gee_Simpson]

14 hours ago, Brother Ben said:

This is very much what I expected going in to this but the difference is this year that I think the gap is getting very small indeed

I agree. Intel is getting harder and harder to recommend now, Ryzens are great value for money as things stand. Intel need to get their act together.

Edited December 1, 2020 by Gee_Simpson

[Gee_Simpson]

I've done my own internal tests, and despite FM21 having faster saving and load times, FM21 has no advantage over FM20 in processing speed which was  a little disappointing.

 

[ScrambledEgg]

4 hours ago, judicious said:

 That's what I thought re the GPU.

My fans are definitely clean as this is quite a new laptop and I've barely used it and I only use it on my laptop tray.

Also using the extreme performance power plan and all chipset drivers are up to date so completely stumped as to what could be causing the issue!

Look for any background apps that may be using resources - strangely my OneDrive is always using loads of CPU at times.

[ScrambledEgg]

1 hour ago, Gee_Simpson said:

I agree. Intel is getting harder and harder to recommend now, Ryzens are great value for money as things stand. Intel need to get their act together.

Just the availability isn’t great though otherwise I’d have built a Ryzen system, Intel 11th gen desktop CPUs are released early in 2021 so we’ll probably see them back at the top of the performance charts for real world usage (ie not just synthetic benchmarks). Really good for consumers that the competition between AMD and Intel is so good though finally.