[SOLVED] Is ryzen 9 waste of money for 4k gaming?

[gemart]

Hello,

im building new pc for 4k 60fps maximum graphics settings gaming. The gpu will be rtx 2080 ti (since that is the only capable option for my needs).
I have samsung 65 inch 4k tv.

I havent seen online any 4k gaming benchmarks for new ryzen 9. Wonder why?

What CPU should i buy? What would be "best bang for buck" cpu for 4k max settings gaming?

Thanks in advance

 

[sizzling]

The reason you don't see 4k benchmarks is you won't see any difference than a much lessor cpu, the gpu is limiting long before the cpu. At 4k the fps is much lower than 1080p/1440p, cpu load is more correlated to fps and resolution makes little difference. 60fps @ 1080p is similar cpu workload to 60fps @ 4k. Something like the 3600 would be more than enough for 4k.

 

[TJ Hooker]

Another vote for a 3600. With a B450 motherboard and some ~3200 MHz RAM.

 

[drea.drechsler]

Another vote for a 3600. With a B450 motherboard and some ~3200 MHz RAM.

But what GPU to keep with the budget-friendly 4K gaming theme?

 

[rigg42]

While the CPU won't matter that much for average frame rates you can still get stuttering even at 4k in some games. The charts don't reflect this. If you are going to spend $1200 on the GPU don't skimp on the CPU IMO. If going ITX with limited cooling than you could make an argument for the 3600/3700x. Otherwise I'd be looking at a 3800x/3900x/9900k with this kind of budget. The 3800x would only be worth it if you want to overclock. I've been hearing a lot of reports of people hitting 4.4ghz+ at 1.35v on the 3800x CPU's. $200-300 more spent on the CPU isn't much in the context of a build that's easily going to come out in the $2500 range when all is said and done. That's my 2 cents anyways.

 

[ConanLock]

RTX 2080 Ti matched with whatever CPU you can get is the best. Sure Ryzen 9 won't make much difference as, like said, it's not going to make much increase in FPS. However, that doesn't mean a Ryzen 3 is going to be just as good. The 3600 would probably be about right, but remember that you will also need dual channel memory (2 sticks) as this doubles the bandwidth of data to the CPU. If you can get a faster CPU, like Ryzen 7 or Ryzen 9, then it future-proofs the PC in that department, and it means if you were to upgrade the GPU for even more FPS, then the CPU will be able to keep up.

 

[gemart]

RTX 2080 Ti matched with whatever CPU you can get is the best. Sure Ryzen 9 won't make much difference as, like said, it's not going to make much increase in FPS. However, that doesn't mean a Ryzen 3 is going to be just as good. The 3600 would probably be about right, but remember that you will also need dual channel memory (2 sticks) as this doubles the bandwidth of data to the CPU. If you can get a faster CPU, like Ryzen 7 or Ryzen 9, then it future-proofs the PC in that department, and it means if you were to upgrade the GPU for even more FPS, then the CPU will be able to keep up.

Yeah, well the futureproof will be the idea, so that I could jump over next gpu generation. Its true that money doesnt matter if I can buy 1200 dollars gpu.

I dont want to overclock but I think bying motherboard that has OC- ability wont cost much more than one with no possibility to overclock.

Im going to build "ultimate silent" atx-build. What case do you recommend?

 

[drea.drechsler]

..... I've been hearing a lot of reports of people hitting 4.4ghz+ at 1.35v on the 3800x CPU's....

I'd be very leery of such claims.

Not that they didn't achieve the voltage & clock...but that doing so improved performance especially in balanced useage scenarios including lightly threaded loads such as games.

Watch GN's videos where Steve goes through Zen2 boosting in agonizing detail. Lowering voltage, in particular, actually degrades performance as the processor will then limits it's boost speeds. And driving voltage so low for fixed high all-core freq's will render the processor unstable in heavily threaded workloads.

I'd definitely want to know system configuration...like what cooling setup was used to do this or whether any kind of serious stress testing was done. It might have been done with a chilled system: like a big CLC... maybe even with the radiator in ice batch.

 

[TJ Hooker]

As said above, the 3700X does generally do a bit better in 1% frame rates compared to the 3600, even if the averages are nearly the same. https://www.techspot.com/review/1871-amd-ryzen-3600/

But I'd say the 3600 still holds up quite well. And keep in mind that at 4K the differences will be even smaller.

 

[rigg42]

I'd be very leery of such claims.

Not that they didn't achieve the voltage & clock...but that doing so improved performance especially in balanced useage scenarios including lightly threaded loads such as games.

Watch GN's videos where Steve goes through Zen2 boosting in agonizing detail. Lowering voltage, in particular, actually degrades performance as the processor will then limits it's boost speeds. And driving voltage so low for fixed high all-core freq's will render the processor unstable in heavily threaded workloads.

I'd definitely want to know system configuration...like what cooling setup was used to do this or whether any kind of serious stress testing was done. It might have been done with a chilled system: like a big CLC... maybe even with the radiator in ice batch.

I am somewhat leery of these claims. I believe Wendell from level 1 techs made this claim. I'd trust that he stress tested it. Obviously hitting stable 4.4ghz at that voltage isn't guaranteed and will vary from chip to chip. If you can hit that it's absolutely worth it. If you are that close to the max single core then the difference in single thread is small enough that the massive advantage you'll get in all core workloads makes it a winning strategy. Not all games are heavily single core dependent anyway.

I've watched the GN content. I'm well aware of the undervolt behavior on stock clock as well. I've been posting about it since launch day when I bought 3 of the Ryzen 3000 CPU's. I even started a thread about it after the debacle optimum tech had with his original undervolting video.

https://forums.tomshardware.com/thr...-3000-if-using-auto-clock-multiplier.3501169/

1.35v doesn't require exotic cooling on these CPU's. The CPU's will hit that at full load on stock settings.

 

[drea.drechsler]

....
1.35v doesn't require exotic cooling on these CPU's. The CPU's will hit that at full load on stock settings.

That's true, at least I've seen it on several youtubes (I'm still waiting to get my 3600 for my mortar). I do mean voltage at idle. Voltage readings on Ry3K seem to be even flakier than they were on 1000/2000 chips. Many people are manually lowering voltage it idle because they're seeing something like 1.45 volts or even more. There's a number of reasons...including BIOS behaviour (AGESA versions seem to affect it differently).

 

[rigg42]

That's true, at least I've seen it on several youtubes (I'm still waiting to get my 3600 for my mortar). I do mean voltage at idle. Voltage readings on Ry3K seem to be even flakier than they were on 1000/2000 chips. Many people are manually lowering voltage it idle because they're seeing something like 1.45 volts or even more. There's a number of reasons...including BIOS behaviour (AGESA versions seem to affect it differently).

The idle voltage on all of mine is less than a volt. While there probably are some bios issues with voltage, a lot of what people are seeing is due to the monitoring software and how it interacts with the new ryzen balanced power plan.

View: https://www.reddit.com/r/Amd/comments/cbls9g/the_final_word_on_idle_voltages_for_3rd_gen_ryzen/?ref=share&ref_source=embed&utm_content=title&utm_medium=post_embed&utm_name=0ee1138e018c4ef191f695648fd1b83c&utm_source=embedly&utm_term=cbls9g

 

[drea.drechsler]

The idle voltage on all of mine is less than a volt. While there probably are some bios issues with voltage, a lot of what people are seeing is due to the monitoring software and how it interacts with the new ryzen balanced power plan.

View: https://www.reddit.com/r/Amd/comments/cbls9g/the_final_word_on_idle_voltages_for_3rd_gen_ryzen/?ref=share&ref_source=embed&utm_content=title&utm_medium=post_embed&utm_name=0ee1138e018c4ef191f695648fd1b83c&utm_source=embedly&utm_term=cbls9g

Oh yes, and that too. I think most of the BIOS issues come from pre-release AGESA code that's out there... many mfr's have yet to get to the 1.0.0.3ab release.