9600K vs Ryzen 5600X

[Dustbag1969]

Upgrade 9600K vs Ryzen 5600X

 

Current system:-

ROG Z390-H wiith I5 9600K @5.0Ghz

2080Ti

32Gb Memory

1Tb M2 drive

Rift S

Soundblaster Recon Z

 

If I do update my current system will be used to upgrade my sons PC so it will not be wasted and he is happy as he gets a free update.

However does anyone have experience of changing from a 9600K @5Ghz to a Ryzen 5600X and if so what improvement if any can I expect to see?

 

Thanks

Dustbag

Edited April 25, 2021 by Dustbag1969

[xoxen]

Hi Dustbag,

what is the resotution you are playing with? The higher this is the less impact has a CPU upgrade based on my knowledge. Check with MSI Afterburner how good the GPU and CPU is utilized. If the GPU does not run at 98-100%, you are CPU bottlenecked the the new CPU might bring a benefit. But from a i 5 9600K to a 5600X I do not expect that much gain.

Just my 2 cents.

Xoxen

[Dustbag1969]

Hi,

 

I play in VR on Rift S and it runs pretty well it is just that lately I have got into playing the Liberation Dynamic Campaign and finding that even on the small scale maps I get a very low FPS.

I realise there is a lot going on in the Liberation campaign with many AI units so not expecting miracles but just wanted to know if it is worth upgrading from a 9600 to a 5600X

 

Thanks

[Skwirl865]

On 4/25/2021 at 11:25 AM, Dustbag1969 said:

Hi,

 

I play in VR on Rift S and it runs pretty well it is just that lately I have got into playing the Liberation Dynamic Campaign and finding that even on the small scale maps I get a very low FPS.

I realise there is a lot going on in the Liberation campaign with many AI units so not expecting miracles but just wanted to know if it is worth upgrading from a 9600 to a 5600X

 

Thanks

Have u tried disabling multi threading? 
I don’t know about Intel but on amd, it makes dcs use all cores pretty much evenly instead of just 2 cores.  Makes it much easier for the cpu to run the game.  
I’m running a 2700x with a 2080 super and my gpu stays between 98-100% and my cpu stays between 40-60% on all cores.  
I have a rift s and a quest 2. 

[Andrei]

I can't say for 9600K vs 5600X. I moved from 7700K to 5800X. There was a massive improvement with DCS Liberation. From barely playable it became very playable  

[AlpineGTA]

Have you considered an i5-11600K? It's cheaper and appears to outperform the 5600X. I was considering a 5600X but was swayed towards the intel i5 after seeing this:

 

https://cpu.userbenchmark.com/Compare/AMD-Ryzen-5-5600X-vs-Intel-Core-i5-11600K/4084vs4113

 

And I'm glad I did. I upgraded to the i5-11600K from a relatively ancient i7-3770 which was severely bottlenecking my system on missions with lots of AI. After the upgrade I went from a slide show to nicely smooth and playable.

 

I also upgraded my motherboard and RAM I might add (I had to, for the new CPU socket), which also helped, but I'm pretty sure most of the bottlenecking was caused by the old CPU as I retained the same GPU. I think the i5-11600K seems a really good choice at the moment for price/performance. 

 

Edit: Having said that, your 9600K is a lot better than my i7-3770 so you might not see as much improvement as I did.

Edited May 6, 2021 by AlpineGTA

[panton41]

5 hours ago, AlpineGTA said:

Have you considered an i5-11600K? It's cheaper and appears to outperform the 5600X. I was considering a 5600X but was swayed towards the intel i5 after seeing this:

 

https://cpu.userbenchmark.com/Compare/AMD-Ryzen-5-5600X-vs-Intel-Core-i5-11600K/4084vs4113

 

 

It gets there by burning at least twice as many watts, if not more. (Look at their TDP and then realize Intel is notorious for outright lying about real world power consumption (measured by a wattmeter) vs. claimed TDP. It might be burning 250+ watts to actually get its full performance.

 

There's also the simple fact that UserBenchmarks simply isn't a very good site for what it claims to do.

Edited May 6, 2021 by panton41

[AlpineGTA]

Yes, I've heard it does have a higher power consumption but still, the performance is undeniably good. I've seen it tested well for performance on some youtube channels too. It certainly works for me anyway, I'm happy with it. 

[Wali763]

Hi.

2D with 1440p monitor here.

I upgraded from a 5,0 GHz 9600k to a stock 9900k (gave it back) and than to a 5600x in February. Also added additional 32GB Ram. In some scenarios I was often having stutters (MP or the SC). Not low FPS but stutters, where eg the antenna on the island would not smoothly rotate but often briefly stop. 9900k did nothing at all, 5600x and RAM only a bit. What cured the stutters was going from a 2070S to a 6700XT. Very interesting considering the stutters were clearly CPU-related. Could be SAM or the supposed less CPU demands of Radeon-GPUs....

 

Anyway, Im happy now (if DCS does not crash of course), even the 6800 would have been perfect but I was unable to get one at a not completly insane price.

 

In hindsight Id say much financial  effort for very little gain.........As always in DCS....

 

Still, I could not recommand any way. Depends too much upon the personal situation. Still the IPC-advantage for sure would support going AMD.

 

Also considered the 10700 or 10600K but since the 9900k helped zero I went for AMD. With Intels 10th vor 11th gen Id get a non-K CPU, since differences in stock are tiny and OC requires a massive cooling solution, still offering very little more perf.

 

 

EDIT 2 Hours later:

 

I attached a pic to demonstrate the change the 6700 made. The problem was mostly with the Tomcat (A and B more or less the same) and not with the Hornet, that had a straigt frametime-curve. Every large spike in that curve meant a stutter. In the last line (with the 6700XT) one can see, that those spikes from above (then with the 9600K) were gone; with the 9900K there was no change and with 5600X they seemed less (not conclusive though) and still some stutters. Even though I made the last curve with 2.7 and the other curves are from 2,5.6, I witnessed the same with 2.5.6 on the 6700XT.

 

Edited May 7, 2021 by Wali763

[AlpineGTA]

Yes, with hindsight, an i5-11400 is probably the CPU I should have bought rather than the 11600K. The 11400 draws a lot less power and runs cooler for barely any noticeable difference in FPS so it's even better value. There's a good comparison video here. This guys thinks overclocking CPUs is dead.

 

 

 

[Andrei]

5 hours ago, Wali763 said:

What cured the stutters was going from a 2070S to a 6700XT. Very interesting considering the stutters were clearly CPU-related.

Interesting, thanks for the observation. I think it were guys from Hardware Unboxed who made a study of how different GPUs work with weak CPUs. Radeon had a very notable advantage there. Something related to nVidia drivers if I recall correctly. None of the CPUs you tested can be considered "weak", but DCS being a resource hog that it is... You might have very well seen the same effect.

[Wali763]

Actually I was not expecting this, since HWUnboxed stated that this phenomenon (older Radeon being faster than eg 3080s when coupeled with weaker CPUs) only applys to games, where the CPU is completly used, as of meaing ALL threads at alomost 100%. But since DCS is requiring mostly singlethreaded performance and most of the cores of our CPUs actually dont have to do anything, such a situation should not arise.

 

Not sure exactly, what it is, but it just works visibly better then before with the 2070S in my situation. YMMV of course.

[Dustbag1969]

In the end I went for an AMD 5600X and and MSI X570 motherboard.

 

I must admit I am quite disappointed. I wasnt expecting major improvements but I am seeing much worse frame rates with the 5600X than I was with my 9600K @5GHz

 

I have not owned an AMD processor for many years so maybe I dont have it setup correctly.

 

I have installed Ryzen Master but not sure what to do with it. I have also read about PBS but again not sure what to do with it either.

 

Does it sound normal to have worse performance after switching from a 5Ghz 9600K to a 5600X and if so is there anything I can do (Apart from selling the 5600X and going back to Intel)

 

Thanks

[Lurker]

1 hour ago, Dustbag1969 said:

In the end I went for an AMD 5600X and and MSI X570 motherboard.

 

I must admit I am quite disappointed. I wasnt expecting major improvements but I am seeing much worse frame rates with the 5600X than I was with my 9600K @5GHz

 

I have not owned an AMD processor for many years so maybe I dont have it setup correctly.

 

I have installed Ryzen Master but not sure what to do with it. I have also read about PBS but again not sure what to do with it either.

 

Does it sound normal to have worse performance after switching from a 5Ghz 9600K to a 5600X and if so is there anything I can do (Apart from selling the 5600X and going back to Intel)

 

Thanks

 

There is something definitely wrong with your system or with the DCS World install. The 5600X should be outperforming the i5-9600k even at stock speeds all other things being equal. Did you do a clean install, did you update your motherboard drivers and graphics card drivers? What memory (RAM) are you using? 

[Dustbag1969]

Hi,

 

It was a fresh install and all latest drivers installed.

Memory is 32Gb of Patriot Viper 3600 using XMP profile

 

It runs OK in single player missions but as soon as there is a lot of AI units (e.g. Liberation DCE) it becomes very jerky and much worse than it was with the 9600K

 

 

[Svsmokey]

XMP ? Have you verified that memory is AMD-compatatible ?

[Dustbag1969]

How do I verify it is AMD compatible?

 

It is Patriot Viper 4 Blackout Series part number PVB432G360C8K

[Lurker]

On 5/10/2021 at 2:25 PM, Dustbag1969 said:

Hi,

 

It was a fresh install and all latest drivers installed.

Memory is 32Gb of Patriot Viper 3600 using XMP profile

 

It runs OK in single player missions but as soon as there is a lot of AI units (e.g. Liberation DCE) it becomes very jerky and much worse than it was with the 9600K

 

 

 

Was it a direct 2.7 to 2.7 comparison? What I'm asking is basically whether you already ran the 2.7 patch on your i5-9600k system? Because if it's not a direct comparison and you ran 2.5 on the i5, and are now running 2.7 on your new system then maybe this might be the reason for the discrepancy in performance?

[Dustbag1969]

Yes it was a direct comparison as I was running 2.7 on my 9600K for a few weeks before changing to the 5600X.

 

 

[Thinder]

On 5/13/2021 at 1:44 PM, Dustbag1969 said:

Yes it was a direct comparison as I was running 2.7 on my 9600K for a few weeks before changing to the 5600X.

 

 

 

You need to be aware of the need for the 5600X and other Ryzen of its generation for a good RAM bounding.

 

It's not about frequencies as AMD recommands Up to 3200MHz but latency; the Ryzen 5 are optimized for low latency so you could gain (as I did) a lot by using  4 X 8GB CL14 single bank 3200MHz RAM.

 

The kit I use now completely eliminated any RAM bootlenecks and the CPU doesn't throttle back because its controller struggles to cope with more ranks than it can handle when mixed with higher frequencies.

 

Single rank also allows for rank interleaving, something dual ranks doesn't, instead they cause the CPU controller to meet its limits much earlier than it could, causing a bootleneck.

 

That's where the issue with Ryzen 5 is, RAM controller vs RAM rank and high frequencies, if you want to use 2 ranks modules, do not use more than two modules, especially if you use frequencies higher than 3200MHz, that's where your CPU controller will throttle back and it affects RAM to CPU but also RAM to GPU channels speeds.

 

In short, I gained 6.04% in 3D Karck CPU speed using this kit over a CL16 dual rank, same frequency...

 

G.SKILL TridentZ RGB Series 32GB (4 x 8GB) 288-Pin DDR4 SDRAM DDR4 3200 (PC4 25600) Desktop Memory Model F4-3200C14Q-32GTZR

 

Here are the results (Crucial RAM first. G.GKILL second)

Fire Strike Ultra 4K gaming with DirectX 11

3840 x 2160 (16:9) (4K) MSAA x 2

MSI Afterburner ON. AMD Ryzen Master ON; Game mode.

>>>>>>

Graphics score: 6 496 vs 6 583. +1.33%
 
Physics score: 23 894 vs 25 339. + 6.04%

Combined score: 3 605 vs 3 654. + 1.35%.

 

I passed on my results to MSI for info as my motherboard doesn't have this RAM kit listed, here is their reply:

 

Quote

These are solid gains indeed! We are glad to see you are happy with our product.

 

 

 

Edited May 26, 2021 by Thinder

[Bossco82]

Hi there,

 

I have a 5600x and a MSi X570 motherboard and have no issues running DCS DCE campaigns on high/ultra detail. For example landing a F-16 at Dubai with other AI around I get a solid 60fps, checked with vsync off and its around 80-90fps. I am running a 2080ti GPU.

 

Maybe its not boosting properly for some reason. Try downloading Ryzen Master and set a profile to the following.

Manual

4600 set for each core

1.25vcore

 

See how it runs with that enabled. Run DCS with full screen disabled and alt tab out to check the cpu speeds in Ryzen Master.

 

Also check how your CPU is actually performing, a quick test with CPUZ. Run a benchmark. My scores are single thread 638 and multi thread 5002, just ran that while typing this.

 

I use to have a 9900k @ 5Ghz on a Z390 system, my 5600x performs better in flight sims because of the single thread performance. 

 

Best of luck I know from experience chasing these gremlins is a PITA. You should not be seeing such a performance hit between a 9600k and a 5600x.

[Sn8ke_iis]

I'm sorry you are getting such bad advice.

 

The 5600X is a good CPU for gaming. When Gamer's Nexus tested it, it performed really well even at stock. Thing is to get the measurable differences in framerate for all his graphs he had to use benchmark software under specific test situations like 1080p medium. An even then it was different for different games. Under normal gaming conditions with higher graphic settings especially at 1440p or 4K there isn't going to be a difference. Especially one that is noticable to the human eye. People who are telling you differently have some kind of bias that isn't backed up by data. Which is unfortunately very common with this issue. Less so on this forum but it's really bad on a lot of PC build forums. 

 

On games like CounterStrike and Rainbow 6 there might be a higher framerate delta but whether your eye will actually perceive any difference is debateable even if you had a 240Hz or 360Hz monitor.

 

People really need to stop making asinine power usage comparisons if they don't know what they are talking about. To get my CPU above 200W you have to stress test it in Prime95 small FFTs with AVX on. Right now my entire computer is drawing 72W while running Windows and having a few browser tabs open. When playing DCS, per MSI Afterburner my CPU draws a whopping 67.2W.  DCS only uses 3 cores. There will not be any significant difference in power draw while gaming between CPUs of Intel or AMD.

 

TDP has absolutely nothing to do with claimed power consumption or total potential power draw of a CPU. It means thermal design power and is a metric for cooler manufacturers to design their coolers for typical use cases not a metric of how much power it is using.  

 

All the anecdotal evidence I've seen from qualified people seems consistent. AMD CPUs are sold with very little headroom for overclocking so it's not really worth breaking the warranty unless your prepared to buy a new chip. They do respond well to memory tuning and tightening up timings to lower latencies. Thing is, if you have to buy expensive specialized B-die memory to get the full potential out of the chip, that's no different than having to pay for cooling to overclock Intel chips.

 

My general advice between a 5600X or a 10600K/11600K is get whatever is on sale. The only real difference you will actually perceive with your senses will be the price. 

 

 

[Thinder]

On 5/31/2021 at 7:01 AM, Sn8ke_iis said:

Thing is, if you have to buy expensive specialized B-die memory to get the full potential out of the chip, that's no different than having to pay for cooling to overclock Intel chips.

 

 

There us a huge difference bewteen the two approaches:

 

In one you fit RAM kits which are optimized for the Ryzen CPU, take full advantage of their design when it comes to low latency, stay within AMD recommandation for max frequencies, take advantage of rank interleaving etc. B-Die doesn't play a role there, apart for the fact that it allows for tighter frequencies settings.

 

On the "cooling for overclock" solution, you spend the money on cooling which the Ryzen 5600X doesn't need when coupled with the RAM in question as it already runs cooler and has a lower cosumption than the equivalent Intel CPU, and you'll gain 6.04% in speed at 4K without overclocking, you can achieve this with the stock cooler.

 

The whole point being that overclocking doesn't remove the CPU limits in terms of what its controller can cope with when it comes to RAM number of ranks and frequencies, a point that was made very clear by AMD and MSI in their replies to me when I submitted my results to them.

 

The reason why AMD recommend 3200MHz is the CPU controller and its ability to cope with higher frequencies and number of ranks per RAM stick, 3200MHz being the maximum for a 4 X ranks, passed that, be it ranks or frequency, the CPU is going to throttle down when under load and it does affect every single buses it is connected to including CPU-to-GPU.

 

So at the end of the day, spending a premium in this kind of RAM kit is way better that pushing the CPU with overclocking regardless of the cooling solution and speed gain, simply because the bottleneck is not removed and the controller limits are still there.

 

For info:

 

The Crucial Kit I replaced with the G.Skill cost me £155.99, I paid £238.99 for the G.Skill kit, that's a difference of £83 for the resuls I got at 4K;

Graphics score: 6 496 vs 6 583. +1.33%, Physics score: 23 894 vs 25 339. + 6.04%, Combined score: 3 605 vs 3 654. + 1.35%.

 

If you take into account the fact that my CPU is running at a frequency where it is under guarantee, it is more than worth the extra £83.

 

Intel Core i5-9600K @ 3.70GHz vs AMD Ryzen 5 5600X

 

Intel Core i5-9600K vs. AMD Ryzen 5 5600X

 

Intel Core i5 9600K vs AMD Ryzen 5 5600X

 

UserBenchmark

Edited June 6, 2021 by Thinder