Early look at Alder Lake performance in MSFS

[ChaoticBeauty 6,692]

Reviews of Alder Lake-S have started going up ahead of launch. Gaming performance is very solid overall, however Microsoft Flight Simulator seems to be one of the few games where Zen 3 is on par or doing better on average.

For those who want to go with an Alder Lake build, it might be best to wait and see whether Asobo can optimise further for the big.LITTLE architecture.

[cavaricooper 2,613]

I just upgraded to a 11900k and a 3080Ti and so will sit tight, but it’s good to see you posting news… I’ve missed that 🙂

C

Edited November 4, 2021 by cavaricooper

[ryanbatc 12,775]

1080?  Is this 2008 lol?

Let's see 2k and 4k?

[Av8ing1 46]

Benchmarking at the lower resolution pushes the CPU and not the GPU. Higher res would show you GPU performance. This is a CPU we are looking at. 

[bbsmitz 132]

Anyone else find it weird that the 5950x is getting beaten by both the 5900x and the 5600x?  Even if MSFS 2020 is more heavily single core speed dependent, the 5950x has a higher max clock than the 5600x.

Edit: Also good to see Intel getting back on the horse. They've had a rough couple of years.

Edited November 4, 2021 by bbsmitz

[Flying-Hare 218]

I'm ready to upgrade my 8700K😊

[DylanC 181]

3 hours ago, ChaoticBeauty said:

Reviews of Alder Lake-S have started going up ahead of launch. Gaming performance is very solid overall, however Microsoft Flight Simulator seems to be one of the few games where Zen 3 is on par or doing better on average.

For those who want to go with an Alder Lake build, it might be best to wait and see whether Asobo can optimise further for the big.LITTLE architecture.

For many who are running older hardware (I'm running a 6700K), this test doesn't matter.   The benefits far outweigh the negatives.  I bought it (12900K) regardless.  In fact many of the reviewers are already saying if you have anything less than a 9900K, this upgrade is worth it.  In fact the I5 12600K looks like a sweet spot for budget gaming.

 

[AnkH 2,233]

50 minutes ago, ryanbatcund said:

1080?  Is this 2008 lol?

Let's see 2k and 4k?

CPU or GPU Limit something you are aware of? Makes no sense to run CPU comparison tests in 4K, because you will constantly run into a GPU limit and your results will be worth nothing. CPU tests are always done in the lowest possible resolution, with the lowest possible graphics settings to eliminate as much of the GPU limit as possible. 

[ryanbatc 12,775]

27 minutes ago, AnkH said:

CPU or GPU Limit something you are aware of? Makes no sense to run CPU comparison tests in 4K, because you will constantly run into a GPU limit and your results will be worth nothing. CPU tests are always done in the lowest possible resolution, with the lowest possible graphics settings to eliminate as much of the GPU limit as possible. 

I understand that.  However, running at 4k would certainly affect the CPU.  I can't just upgrade to a Rtx 3080 and expect my fps to increase running my current CPU.  Upgrading to a 2k monitor slightly decreased my fps - and I'm still CPU limited.

[Maxis 1,849]

1 hour ago, bbsmitz said:

Anyone else find it weird that the 5950x is getting beaten by both the 5900x and the 5600x?  Even if MSFS 2020 is more heavily single core speed dependent, the 5950x has a higher max clock than the 5600x.

Edit: Also good to see Intel getting back on the horse. They've had a rough couple of years.

No .. Don't find it weird. The 5950x has x2 ccx with 8 cores and a slower base clock of 3.4ghz  but a theoretically higher boost clock of 4.9ghz/ both the 5900x and 5600x have higher base clocks of 3.7ghz with the 5900x having 6 active cores in its double ccx layout and a boost clock of 4.8ghz and with the 5600x having only 1ccx of 6 cores so no inter ccx latency to deal with and a boost clock of 4.6ghz.

However 

Based on the boosting algorithm and power management heat has to be factored in and after enabling PBO and +.2 ghz boost overclock you have situations where the 5950x is not as efficient as the other two counterparts because more cores + its accompanying heat are factored in. Example .. After tuning my chip 5900x can reliably operate at 5 to 5.1ghz. I have tuned 5600x chips to operate at 4.825ghz with no issues. 

Have to choose the correct chip for the job .. Especially when it comes to AMD cpus .. More isnt necessarily better. Most review sites acknowledge that for all practical purposes the 5900x is AMD's fastest gaming chip. 

Good job on intel for catching up. However from these benches i guess there is no reason for me to sell and switch .. (Especially with those motherboard prices)

Waiting on the 3dvcache 5900x equivalent.

Cheers

Edited November 4, 2021 by Maxis

[Ianrivaldosmith 5,317]

I’m on the bench, unsure if I’d see much difference from a 9900ks at 4K… 

[ryanbatc 12,775]

Just now, Ianrivaldosmith said:

I’m on the bench, unsure if I’d see much difference from a 9900ks at 4K… 

I agree - if only they'd show a res that people would run in 2021.  They could so 4k and still easily render a scene that is CPU limited.

[bean_sprout 1,201]

Glad to see my 5900x is still relevant. 

I gotta pass on Intel for the near future.

bs

[ChaoticBeauty 6,692]

3 minutes ago, ryanbatcund said:

However, running at 4k would certainly affect the CPU.  I can't just upgrade to a Rtx 3080 and expect my fps to increase running my current CPU.  Upgrading to a 2k monitor slightly decreased my fps - and I'm still CPU limited.

Very few games nowadays are still CPU-limited at resolutions above 1080p, so it doesn't make sense to make benchmarks at higher resolutions for a CPU review. Since tech sites don't have much time to make benchmarks in time for the embargo lift, such tests are usually reserved for future articles.

Also 1080p is still by far the most popular gaming resolution, 66,5% according to the Steam survey. Not really a thing of the past.

42 minutes ago, DylanC said:

For many who are running older hardware (I'm running a 6700K), this test doesn't matter.

In a recent thread though people were contemplating upgrading from CPUs like Comet Lake and Rocket Lake because it was generally expected that Alder Lake would consistently beat Zen 3 in performance, which is not the case after all. For older hardware any modern CPU would be a good upgrade.

12 minutes ago, Maxis said:

The 5950x has x2 ccx with 8 cores and a slower base clock of 3.4ghz  but a theoretically higher boost clock of 4.9ghz/ both the 5900x and 5600x have higher base clocks of 3.7ghz with the 5900x having 6 active cores in its double ccx layout and a boost clock of 4.8ghz and with the 5600x having only 1ccx of 6 cores so no inter ccx latency to deal with and a boost clock of 4.6ghz.

Correct, and it's also worth noting that these tests were made on Windows 11 which still had some Ryzen bugs at the time. The CCPC2 fix only started getting deployed today.

[ryanbatc 12,775]

I'm just not understanding how accurate that 1080P bench can be....    there's a miniscule difference between the flagship intel cpu and the 12600K.  I believe at higher resolutions that difference would be greater.  

Maybe it is time to invest in AMD - though I'm not sure how well AMD would work on next gen XP platform either....traditionally AMD wasn't great with XP11.  Seems to be totally fine in MSFS.