CPU scaling @4k when gaming w/gtx 1080

[Gallbert]

I've been looking around and I'm unable to find a good benchmark or a solid answer.

I have an i5 3570k clocked at 4.5ghz, and have been considering upgrading to a Rysen 2700x.

 

I've noticed that in most benchmarks the average frame rate in a given game levels out between modern processors as the resolution increases, but I can only find benchmarks up to 2k, and none of them include my processor, so it's hard to judge if it would actually be better.

 

I have a 3440x1440 ultrawide with a 100hz refresh rate, and a 4k monitor with 60hz, this is why I've been trying to find out.

 

If anyone has the answer, or could point me to a video or article that I might have missed that would be amazing.

 

Thank you. Let me know if I need to give more information, this is my first post. 

[Princess Luna]

If gaming is the only thin you do at all, you'll hardly feel that much of a impact in frame-rates.

[Motifator]

You might want to wait for the 9900K.

[MedievalMatt]

1 minute ago, Gallstafff01 said:

I've been looking around and I'm unable to find a good benchmark or a solid answer.

I have an i5 3570k clocked at 4.5ghz, and have been considering upgrading to a Rysen 2700x.

 

I've noticed that in most benchmarks the average frame rate in a given game levels out between modern processors as the resolution increases, but I can only find benchmarks up to 2k, and none of them include my processor, so it's hard to judge if it would actually be better.

 

I have a 3440x1440 ultrawide with a 100hz refresh rate, and a 4k monitor with 60hz, this is why I've been trying to find out.

 

If anyone has the answer, or could point me to a video or article that I might have missed that would be amazing.

 

Thank you. Let me know if I need to give more information, this is my first post. 

Most people arent benching 4K because its still not widely used for gaming.

 

On the whole.  as resolution increases, the CPU matter less and less and the GPU matter more and more.

 

Think of it like this.  your CPU feeds your GPU the data it needs to make frames (among other things like audio, NPC, etc.  These are variable with the game and dont scale from resolution to resolution).  If your GPU can make 1000 frames at 1080p, and only 100 at 4K.  the CPU has to do far less work to feed the GPU data at 4K because it takes the GPU longer to crunch out a frame.

[agent_x007]

CPU thinks in points (for triangles) in space. And does AI, phys sim, etc. stuff in spare time
GPU thinks in triangles to pixels exchange.
CPU doesn't see pixels (more or less), it only does driver usage and other stuff mentioned earlier.
In short : 

IQ = Image Quality
^This is the general idea everyone should have.
And Yes, I know there are graphics options that can impact both (like render distsance).
You can also add RAM speed (latency/bandwidth), next to CPU on Y-axis.

[Mira Yurizaki]

If you can't find benchmarks for your CPU, it's best to find a CPU that performs roughly the same. You can reference UserBenchmark (maybe) and PassMark to get a general idea of where your processor is roughly performing.

[Gallbert]

Hmm. Maybe i've been thinking about it wrong. I know my CPU will beat a Rysen CPU in games that don't fully utilize multi-threading no matter what, but my main question is whether that difference will be less noticeable or level out as resolution increases and the CPU becomes less of a bottleneck and it shifts toward the GPU. Again, maybe I'm thinking about it wrong and didn't put much thought into it before posting, and maybe it would be better to save for a newer generation intel CPU if the processor speed really concerns me more than the amount of cores.

 

It's just proving difficult to find the benchmarks I need to satisfy my curiosity..

[Motifator]

5 minutes ago, Gallstafff01 said:

I know my CPU will beat a Rysen CPU in games that don't fully utilize multi-threading no matter what

Your CPU doesn't really beat my 2700X at 4.35ghz all core stable, even in single threaded tasks. This CPU has IPC similar to that of Broadwell chips.

[Gallbert]

2 minutes ago, Motifator said:

Your CPU doesn't really beat my 2700X at 4.35ghz all core stable, even in single threaded tasks. This CPU has IPC similar to that of Broadwell chips.

Thank you, that answers one of my questions. Really wanted to know if i'd have to sacrifice any performance.

[bcredeur97]

i went from a 3570k that would only do 4.2GHz to a 3960X that will only do 4.3

other than being good at losing the silicon lottery; the two extra cores have helped in really demanding games like BF1 where the 3570k would make things stuttery sometimes.

 

I did get the 3960X + a mobo for $150 tho. just was a great deal hence why i went for it. But someone like you may be able to make a better choice.

[Gallbert]

2 minutes ago, bcredeur97 said:

i went from a 3570k that would only do 4.2GHz to a 3960X that will only do 4.3

other than being good at losing the silicon lottery; the two extra cores have helped in really demanding games like BF1 where the 3570k would make things stuttery sometimes.

Hmm, that's good. Along with normal games, I'm also concerned about emulators such as Dolphin and Cemu.

[Gallbert]

On passmark the single threaded scores of the 3570k and a ryzen 2700x seems to be very similar, with the 2700x pulling out ahead. I'm going to run my own test to see what my score is with my overclock.

[Gallbert]

My score came to an overall 9419 with at single threaded score of 2526, beating the 2700x recorded on the Passmark charts which was 2190.

I honestly didn't think mine would be that much higher.

[Motifator]

You're comparing an OC'ed CPU to stock. Also, PassMark is garbage. This is what I get with my chip on UserBenchmark:

http://www.userbenchmark.com/UserRun/10369957

[Gallbert]

52 minutes ago, Motifator said:

You're comparing an OC'ed CPU to stock. Also, PassMark is garbage. This is what I get with my chip on UserBenchmark:

http://www.userbenchmark.com/UserRun/10369957

Stock vs stock would be pointless if i'm looking to see the difference in performance of my current vs. what I can get.

Just ran that benchmark, http://www.userbenchmark.com/UserRun/10426358

It's about even, with your 2700x being a bit better which is nice. Pretty much exactly what I wanted to see, thank you.

This pretty much solved my problem.