I have the following specs.

i7-3770,
16GB DDR3 1600 RAM
GTX 1050 Ti

I am thinking of upgrading the CPU to i5-8400 (with B360 chipset, and budget CPU tower cooler, keep same GPU).

In terms of FPS (i get about 30 at moderately high settings and tweaks, etc)... how much should i expect it to scale? 

Thank in advance.

This depends on what you are running. Also, background programs will eat some memory.

What sim are you referring to?  More information will help someone help you.

 

Sorry, forgot to mention that. 

FSX Steam

3 hours ago, anshugiri said:

In terms of FPS (i get about 30 at moderately high settings and tweaks, etc)... how much should i expect it to scale?

By 28% - http://cpu.userbenchmark.com/Compare/Intel-Core-i7-3770-vs-Intel-Core-i5-8400/1979vs3939

18 hours ago, Jim Young said:

By 28% - http://cpu.userbenchmark.com/Compare/Intel-Core-i7-3770-vs-Intel-Core-i5-8400/1979vs3939

Ok. So it scales 1:1 (CPU single core performance to FPS % increase?)

The more cores, the better.  The above benchmark is about as close as you are going to get in guessing the fps on a newer system.

8 hours ago, Jim Young said:

The more cores, the better.  The above benchmark is about as close as you are going to get in guessing the fps on a newer system.

I'm not sure how you've arrived at this conclusion, but there is no supporting evidence to say that this theory applies to FSX/P3D.  In fact, we've known for literally decades now that MSFS simulator performance is primarily constrained by the performance of a single CPU core.  Additional cores can help load textures faster which has the effect of preventing "blurries", but the effect on FPS is zero, after a certain point.  This point being any modern CPU with at least 4 cores and 8 threads.  

That being said, an i5 8400 is a good CPU, and is somewhat faster than a 3770 in single thread-bound tasks like the Cinebench R15 single thread test, a far better analog for FSX/P3D performance than the generic aggregate performance test to which you've linked.  

Here is a site which serves a similar function to the one you've linked, but in this case they have collected the test results for Cinebench R15 single thread performance.  https://www.cpu-monkey.com/en/cpu_benchmark-cinebench_r15_single_core-7

Searching this list of results yields performance scores of 137 for the i7 3770k and 161 for the i5 8400 - a 17.5% increase.  Not huge, but it might be just enough to bump the minimum FPS up to desirable levels in the most demanding simulation scenarios.  

Edited September 25, 20187 yr by TechguyMaxC

31 minutes ago, TechguyMaxC said:

but there is no supporting evidence to say that this theory applies to FSX/P3D. 

Of course you are correct but the OP is going from the same sim to the same sim.  There is no differences there other than he wants to upgrade the CPU.  The OP wanted to know how much his fps would increase from an i7 3770 to an i5 8400.  I could go out and buy an i7 3770 and an i5 8400, run the same scenario on one of my flight simulators and possibly come up with what the exact fps would generate (that would be extremely difficult to do too as you change your views and it increases/decreases the fps but I guess I could keep my head looking straight ahead at all times).  You came up with a 17.5% increase based on a single core (my CPU uses all cores) and I came up with a 28% increase based on benchmarks conducted on the two systems by thousands.  I also think my percentage is better as the scores do not simply represent the CPU but also other components in the computer.  My method is the best guide for anyone considering upgrading their system as doing it any other way is a guess.  I use the link I posted above when I am looking for a new CPU. 

My method gives me a more accurate depiction of what 'might' be the increase as mine includes all components and not a single CPU and I highly recommend this method when someone is considering an upgrade.  For instance, a 25% increase or higher would probably be worth upgrading. 

Now, if the OP decided to buy the i5 8400K instead of the 8400, he could overclock his CPU and that would definitely increase his FPS even more. 

 

1 minute ago, Jim Young said:

Of course you are correct but the OP is going from the same sim to the same sim.  There is no differences there other than he wants to upgrade the CPU.  The OP wanted to know how much his fps would increase from an i7 3770 to an i5 8400.  I could go out and buy an i7 3770 and an i5 8400, run the same scenario on one of my flight simulators and possibly come up with what the exact fps would generate (that would be extremely difficult to do too as you change your views and it increases/decreases the fps but I guess I could keep my head looking straight ahead at all times).  You came up with a 17.5% increase based on a single core (my CPU uses all cores) and I came up with a 28% increase based on benchmarks conducted on the two systems by thousands.  I also think my percentage is better as the scores do not simply represent the CPU but also other components in the computer.  My method is the best guide for anyone considering upgrading their system as doing it any other way is a guess.  I use the link I posted above when I am looking for a new CPU. 

My method gives me a more accurate depiction of what 'might' be the increase as mine includes all components and not a single CPU and I highly recommend this method when someone is considering an upgrade.  For instance, a 25% increase or higher would probably be worth upgrading. 

Now, if the OP decided to buy the i5 8400K instead of the 8400, he could overclock his CPU and that would definitely increase his FPS even more. 

 

The reasoning you are using is flawed.  P3D is not a generic workload, it is a single application with specific bottlenecks and should not be compared to an average performance aggregated across numerous dissimilar workloads.  I say this to you both as a long time MSFS user and as someone that has performed application benchmarking in a professional capacity for many years, including in the PC 3D graphics space.

53 minutes ago, TechguyMaxC said:

The reasoning you are using is flawed.  P3D is not a generic workload, it is a single application with specific bottlenecks and should not be compared to an average performance aggregated across numerous dissimilar workloads.  I say this to you both as a long time MSFS user and as someone that has performed application benchmarking in a professional capacity for many years, including in the PC 3D graphics space. 

This thread is about FSX-SE, not P3D.  I think I have been a longer MSFS user than you but that really means nothing as we are not talking about 2007 or earlier CPUs.  We are talking about CPU's developed after 2016.  I know your background and capabilities.  You have done well providing expert advice to our membership and many members have come to you for expert advice!  There really is no rocket science anymore as it was back in the duo-core days.  You have an i7, i5, and i3.  Each has its own individual capabilities.  Websites have now been set up to benchmark these CPU's (back in the old days you had to do it).  Today there are algorithms that provide the required information to make a decision.  Should I upgrade or should I not upgrade.  Will a CPU provide more FPS or not.  This has already been determined on thousands of computers and benchmarks around the world.

I do not know why you are fighting with me by stating my reasoning is flawed.  I provided an assumption to the OP via a link and you provided your link and assumption.  There should be no reason to argue.  That's what forums are all about, to provide information to the OP to try to answer his question.  Again, the best CPU increase is 28% based on information from the website I provided.  Your link provides valuable information for the OP too.

Again, thanks for your past expertise to the FS community!

53 minutes ago, Jim Young said:

This thread is about FSX-SE, not P3D.  I think I have been a longer MSFS user than you but that really means nothing as we are not talking about 2007 or earlier CPUs.  We are talking about CPU's developed after 2016.  I know your background and capabilities.  You have done well providing expert advice to our membership and many members have come to you for expert advice!  There really is no rocket science anymore as it was back in the duo-core days.  You have an i7, i5, and i3.  Each has its own individual capabilities.  Websites have now been set up to benchmark these CPU's (back in the old days you had to do it).  Today there are algorithms that provide the required information to make a decision.  Should I upgrade or should I not upgrade.  Will a CPU provide more FPS or not.  This has already been determined on thousands of computers and benchmarks around the world.

I do not know why you are fighting with me by stating my reasoning is flawed.  I provided an assumption to the OP via a link and you provided your link and assumption.  There should be no reason to argue.  That's what forums are all about, to provide information to the OP to try to answer his question.  Again, the best CPU increase is 28% based on information from the website I provided.  Your link provides valuable information for the OP too.

Again, thanks for your past expertise to the FS community!

I'm not fighting with you Jim, I only want you to see that the information you are providing to the OP is inaccurate and could set him up for a potentially costly mistake.  If I were in his shoes, I would want the most accurate information possible.  The difference between 28% performance gain and 17.5% performance gain may be all the difference required to fulfill his desires.  If he goes into the upgrade expecting 28% higher FPS, and only sees 17.5%, he might feel cheated by the advice he's been given here.  I don't know his financial position nor his performance expectations so here I can only speculate on the actual outcome.  If 17.5% higher FPS is insufficient, then he should look at a different upgrade altogether such as the 8700k which has a higher turbo clock (therefore higher single thread performance). This CPU might be outside of his budget, in which case he has 2 choices:

1) buy a CPU that will provide an upgrade which does not meet his expectations 

2) wait and save to buy a faster CPU that will meet his expectations

 

On the subject of the simulator in use, the fact that the OP is using FSX rather than P3D only serves to reinforce my argument as FSX is even further constrained by single thread performance than P3D, being an extremely outdated engine which has seen no significant improvements after primary development ceased when SP2 was launched way back in December 2007.

21 minutes ago, TechguyMaxC said:

I'm not fighting with you Jim, I only want you to see that the information you are providing to the OP is inaccurate and could set him up for a potentially costly mistake.  If I were in his shoes, I would want the most accurate information possible.  The difference between 28% performance gain and 17.5% performance gain may be all the difference required to fulfill his desires.  If he goes into the upgrade expecting 28% higher FPS, and only sees 17.5%, he might feel cheated by the advice he's been given here.  I don't know his financial position nor his performance expectations so here I can only speculate on the actual outcome.  If 17.5% higher FPS is insufficient, then he should look at a different upgrade altogether such as the 8700k which has a higher turbo clock (therefore higher single thread performance). This CPU might be outside of his budget, in which case he has 2 choices:

1) buy a CPU that will provide an upgrade which does not meet his expectations 

2) wait and save to buy a faster CPU that will meet his expectations

 

On the subject of the simulator in use, the fact that the OP is using FSX rather than P3D only serves to reinforce my argument as FSX is even further constrained by single thread performance than P3D, being an extremely outdated engine which has seen no significant improvements after primary development ceased when SP2 was launched way back in December 2007.

Agree,  not give much for the cpu compare thing

I not post here much nowdays,  you understand why

I let real hardware guys play here

Just now, westman said:

Agree,  not give much for the cpu compare thing

I not post here much nowdays,  you understand why

I let real hardware guys play here

Hasse you're a funny guy.  You have an 8700k at 5.6GHz and you say you're not a real hardware guy?  

39 minutes ago, TechguyMaxC said:

Hasse you're a funny guy.  You have an 8700k at 5.6GHz and you say you're not a real hardware guy?  

His English is poor ...I think he said ‘I’m the only real hardware guy here'.

18 hours ago, FunknNasty said:

His English is poor ...I think he said ‘I’m the only real hardware guy here'.

He's just being humble.  Hasse is a world-class overclocker.  He consistently shows up on this site with higher clocks than anyone else, including me, and I've been doing it for 20 years.