AMD Ryzen For Simracing?

[RasmusP]

32 points single and 95 points multi for at least £305 extra cost.....ouch!
Though I think it'd be interesting to see the difference FPS wise (ok, the 2080ti will also help!) to see a price per frame cost of the CPU+GPU on a system like the 9900k above, vs say myself and @Martin Fiala on an R5 3600 and 1660ti on a few racing sims.

Absolutely!
The problem for us simracers is that VR adds more cpu load on the already limiting big main thread of the sims.
And you need constant 90 fps to not risk getting motion sick or put off by asw kicking in and dropping to 45 fps.

So how I interpret a lot of 9900k builds:
- I want 90 fps vr
- it's worth any prize below 1000€

So it's not about best bang for the buck, it's simply getting the nice vr experience no matter the costs within reasonable range.

Good thing about simracing though:
The already working sims won't increase their demands on the cpu at all in the future and newer sims use more threads so that 9900k should last a long time for simracing, streaming and some workstation stuff.

 

[RobertR1]

32 points single and 95 points multi for at least £305 extra cost.....ouch!
Though I think it'd be interesting to see the difference FPS wise (ok, the 2080ti will also help!) to see a price per frame cost of the CPU+GPU on a system like the 9900k above, vs say myself and @Martin Fiala on an R5 3600 and 1660ti on a few racing sims.

You can get a 6core, 10600k with high end mem, tune it and get the same performance as my 9900k.

Price to performance generally stops scaling in a linear fashion after a certain point like how a 100k car isn't twice as fast a 50k car.

 

[RobertR1]

@RasmusP that's exactly right. I'm running a locked 90fps on the Pimax5k+ with large fov in AC with most settings maxed out. So it's really use case specific. Not having stuttering or shimmering in VR is worth it to me.

 

[karoliens]

Just to add, i notice theres seems to be a disproportionate jump to my score vs the others for a small mhz bump, this may be due to having 3600mts RAM, whereas id assume most are on 3200?

I'm using 3200Mhz RAM with enabled XMP profile.

 

[Case_]

I have to finish that video...

Edit: Commentary script finally done, yay!

 

[robt100]

605 here 4.35ghz all core on a ryzen 3600.
203 on single core, so scaling quite nicely

Just as a quick update, did a little bit of memory tweaking, went from 18-22-22-22-42 3600 to 16-21-21-21-42 3600.
Scores moved a tiny bit..
607 3thread
204 single core

on R20 its
505 single core and
3944 multi

 

[RasmusP]

Just as a quick update, did a little bit of memory tweaking, went from 18-22-22-22-42 3600 to 16-21-21-21-42 3600.
Scores moved a tiny bit..
607 3thread
204 single core

on R20 its
505 single core and
3944 multi

Don't want to crush your success but 3 points on 3 thread bench are fluctuations
My results vary up to 10 points. If you want a real comparison, do it 5x for each tuning step.

But memory timing might not do anything for cinebench but maybe for other applications.

 

[Case_]

Well, I've finally finished it, in case anyone's interested. Apologies for the awkward and badly pronounced commentary as always.

 

[robt100]

Nice video @Martin Fiala
Just a quick question, your 1660ti looks identical to mine in everything but the name....I just wondered if you're running a 'silent' fan curve? As I noticed it was thermal throttling in the ACC benchmark. I shall screenshot my Fan and overclock settings in afterburner and post them below..

 

[robt100]

Hoping these might help for comparison? CPU settles to 2040mhz once at 'max' temp

 

[Case_]

I'm not sure if it's temperature that's affecting its clock. If I remember correctly, every time I've checked in the past, it was running at its power limit when throttling down, not at temperature limit (which I have set to 89). In any way, I haven't touched the fan curve at all, it's using whatever it has as stock.

 

[Durge Driven]

While burning DVDs

 

[Durge Driven]

AMD Ryzen RAM scaling - performance effect in games

The impact of memory timings and frequency on AMD Ryzen 3000 systems in games has been a topic of discussion. In this artilcle we'll zoom in on specifically that. See, AMD made a change in 3rd gener...

www.guru3d.com

.......but you need to keep in mind at all times that we tested it at 1080p with a GeForce RTX 2080 Ti, any other graphics card is far more GPU limited. Also, at higher resolutions, like 1440p/4K, the GPU will again become the limiting factor in most games, obliterating that memory performance differential

That is all I have ever said because in reality no one with this GPU runs sims @ 1080p

My 3600 C14 loses 1% to best results really splitting hairs and again only @1080p

I use 1440p and 4K monitors ...so I am going to guess this graph would be closer again

 

[Durge Driven]

XT reviews

AMD Ryzen 9 3900XT review

We review the new XT model Ryzen 9 3900XT. Released by AMD as a respin of the 3900X model the XT processors are optimized to bring a bit more bite in single-threaded and medium threaded workloads. ... Conclusion

www.guru3d.com

I know most don't agree but again with the 1080p ( utterly useless test )
3900XT is worth 2 FPS in F1 vs 3900X

So what is the gain at a sensible resolution ( for 3900XT - RTX2080Ti ) ....nadda !

 

[Bjarne Hansen]

I know most don't agree but again with the 1080p ( utterly useless test )
3900XT is worth 2 FPS in F1 vs 3900X

Actual it is losing 2 fps to to the 3900X, but that might be a mistake

 

[Durge Driven]

See this test is more meaningful to me because they include all resolutions that show how poor the gaming boost can be ....even overclocking !

AMD Ryzen 9 3900XT Review

The Ryzen 9 3900XT is the flagship of the new AMD Ryzen XT series. It comes with higher boost clocks and can sustain them better, which helps with single-threaded workloads. In our Ryzen 9 3900XT review, we also saw better overclocking and lower temperatures than on the original Ryzen 9 3900X.

www.techpowerup.com

1440p - 4K tests are there you be the judge ........

 

[RasmusP]

Still not sure why looking at 2% +/- differences in the gpu limit is meaningful for anything...
For me it's quite simple:
1. You check what's your gpu performance is in game X at resolution y
2. You check what's the cpu performance in 720p is for game X.

The lower fps number will be the maximum you'll see.
Then you can decide if you want a cpu that will be below the gpu limit or above.
How far above would make sense, financiallly, is down to your glass bowl, trying to estimate future gaming optimization...

ACC in VR shows us how important it is to check the 720p results.
People will upgrade to a 3080ti when it comes out. Plug and play on the same pc build.
(or the new amd gen ofc)

If they watched the 720p benchmarks a few years ago, many will be very happy to have a 8700k that still holds up very well.
Maybe they'll get some more case fans to overclock it even higher.
And many will bite their asses over their 8600k, which sowed identical performance at 1440p or 4k when it came out but really lacks performance now compared to the 8700k...

If you also do a lot of other stuff than gaming on your pc, re-installing windows and everything is a huge pita.
So keeping a cpu for 5 years or more while upgrading the graphics card 2-3 times makes sense.

 

[Alex Townsend]

See this test is more meaningful to me because they include all resolutions that show how poor the gaming boost can be ....even overclocking !

AMD Ryzen 9 3900XT Review

The Ryzen 9 3900XT is the flagship of the new AMD Ryzen XT series. It comes with higher boost clocks and can sustain them better, which helps with single-threaded workloads. In our Ryzen 9 3900XT review, we also saw better overclocking and lower temperatures than on the original Ryzen 9 3900X.

www.techpowerup.com

1440p - 4K tests are there you be the judge ........

Can't believe they don't have the 3800x in there... like it doesn't exist...

Anyways, I'm hoping my 3800x can serve me well for the next 5 or so years, hopefully even longer.

 

[RasmusP]

Can't believe they don't have the 3800x in there... like it doesn't exist...

Anyways, I'm hoping my 3800x can serve me well for the next 5 or so years, hopefully even longer.

Take the clocks of your 3800x, compare it to the 3700x and then just calculate the performance difference.
There are enough comparison benchmarks online

 

[Case_]

I know most don't agree but again with the 1080p ( utterly useless test )
3900XT is worth 2 FPS in F1 vs 3900X

So what is the gain at a sensible resolution ( for 3900XT - RTX2080Ti ) ....nadda !

And so we have found out that in such resolutions, the GPU is the limit, and we have no idea what the CPU's performance differences actually are.

What an exciting finding in a CPU benchmark it is not...