Intel Core i5 9600k worth to upgrade to a i9 9900k in these AMD days?

[Vayra86]

That's what happens when you test the AC but your brakes are the issue.

Actually being serious, I hate IBT.. that is one of the more useless stress tests around.

The only pro IBT really has is that you can get an idea of relative performance between settings/runs because it gives you a simple number, and heat. Lots of heat.

its a bit like a Furmark with a bench score.

Personally still partial to OCCT for quick testing, and then on to real workloads until the OC falls apart, basically I do still think a synthetic test is essential because you need some sort of similar testing/starting situation to analyze stuff. And the graphs, OCCTs graphs are great.

Ok, I'm going full off-topic here, but do people really run their CPUs at max overclock? I mean, if a stress tool told me the CPU is stable at, say, 5GHz, I'd dial that back 5% or so if I was going for stability.

I don't... I'll even happily back down another 100-200mhz if it means a big reduction in voltage/heat/noise. Given the actual clock, 200mhz is negligible anyway.

 

[bug]

Ok, I'm going full off-topic here, but do people really run their CPUs at max overclock? I mean, if a stress tool told me the CPU is stable at, say, 5GHz, I'd dial that back 5% or so if I was going for stability.

 

[EarthDog]

Ok, I'm going full off-topic here, but do people really run their CPUs at max overclock? I mean, if a stress tool told me the CPU is stable at, say, 5GHz, I'd dial that back 5% or so if I was going for stability.

Short answer, Yes (at least I do).


Otherwise, why not? You just tested it to be stable... Unless voltage/temperature was higher than one prefers, then there isn't a reason to.

Why test at X.xGHz when you are only going to lower it? That makes little sense to me assuming one has proper stress tests and methods (length of time running it for their needs and using a test that works out all instruction sets).

 

[windwhirl]

Ok, I'm going full off-topic here, but do people really run their CPUs at max overclock? I mean, if a stress tool told me the CPU is stable at, say, 5GHz, I'd dial that back 5% or so if I was going for stability.

Personally, no way. I'm not drowning in money, so I try to keep all my gear working for at least 3 to 5 years before upgrading. Right now, I have my Ryzen 3600 CPU running in "Eco-Mode", and my RX 580 stays at default clocks. No extra OC.

 

[bug]

Short answer, Yes (at least I do).


Otherwise, why not? You just tested it to be stable... Unless voltage/temperature was higher than one prefers, then there isn't a reason to.

Why test at X.xGHz when you are only going to lower it? That makes little sense to me assuming one has proper stress tests and methods (length of time running it for their needs and using a test that works out all instruction sets).

But that's just it, how do you test for proper stress tests? As elaborate as a tool may be, I highly doubt it will exercise all combinations of CPU load that you'll see irl.

 

[EarthDog]

But that's just it, how do you test for proper stress tests? As elaborate as a tool may be, I highly doubt it will exercise all combinations of CPU load that you'll see irl.

This is something one figures out when testing as it varies depending on use. If you are overclocking, you should at least figure out which works best for YOU instead of running one thing. It is trial and error. While none are perfect, some simply work better for users depending on their workload. But most work fine for most people. Ive never felt the need to do so, ever.

But yeah, I don't follow the logic in lowering clocks on a stress tested stable and otherwise stable system unless temps/voltage are too high. Makes no sense to me to test higher and lower after.

 

[bug]

This is something one figures out when testing as it varies depending on use. If you are overclocking, you should at least figure out which works best for YOU instead of running one thing. It is trial and error. While none are perfect, some simply work better for users depending on their workload. But most work fine for most people. Ive never felt the need to do so, ever.

But yeah, I don't follow the logic in lowering clocks on a stress tested stable and otherwise stable system unless temps/voltage are too high. Makes no sense to me to test higher and lower after.

I just gave you the logic: it leaves a little room for scenarios the tool doesn't stress

 

[dirtyferret]

Gaming wise, I'd agree with EarthDog. Most of the load is GPU-dependent. And as far as we know, the PS5 will have 8 Zen 2 cores (which will probably be not as highly clocked as their PC counterparts), so it's not an outrageous difference.

They are not Zen 2 cores but rather "Zen 2 based cores" so something is getting cut down (probably cache levels at least) so that the CPU draws the typical 20w that console CPU historically use. Digital foundry tested a Ryzen 3700x (since the console CPUs are 8c/16t) at the rumored 1.8ghz and 3.2ghz levels that seem to be the low and high frequency out on the internet. At 1.8ghz the Ryzen 3700x performed similar to an FX-8350 while at 3.2ghz it performed similar to a Ryzen 1500x. Hard to say how efficient console game developers can make the CPU and remember the 3700x doesn't use cut down cores.

 

[Deleted member 171912]

I have a very good Z370 mini ITX motherboard a Asrock Fatal1ty z370 Gaming-ITX/ac, I use my i5 9600k at 5ghz with a NZXT Kraken x62

And my graphics card is RTX 2080 Super.

1440p gaming at 75hz

I have a youtube channel, where I want stream, but I wasn't ready to start yet and also I edit my videos (1080p@60fps) with this CPU, sometimes very long with high bitrate (12000).

For gaming in your case it is not needed but it can help with streaming and video editing. i9 9900K is not so hot and you have NZXT Kraken X62 already.

 

[cucker tarlson]

Ok, I'm going full off-topic here, but do people really run their CPUs at max overclock? I mean, if a stress tool told me the CPU is stable at, say, 5GHz, I'd dial that back 5% or so if I was going for stability.

I'm going for max overclock but only for cpu intensive gaming.I'm using a balls to the walls XTU profile in odyssey,bf1,division and other cpu dependent games but a standard oc in others.
I like software oc more than bios,I can use profiles like in afterburner.got two overclocked profiles,default one and undervolted one.

 

[EarthDog]

I just gave you the logic: it leaves a little room for scenarios the tool doesn't stress

illogical when there are tests that work.

You also specifically stated STABLE which means a stress test works for the user to be stable. So I answer why, if I am actually stable, would I lower my clocks??? I wouldn't because I am stable and that isn't logical (to me) to do so. You are saying its stable on one side of your mouth but trying to make a point about 'just in case it isnt' out of the other side. It doesn't work like that. Its stable for your uses, or it isnt. And if I am stable, I do not bother lowering the clocks, no. I don't test and show stability, and then lower it again for giggles. Get a more applicable stress test for your situation if you are not stable doing you daily tasks.

Anyway, OT, so I digress.

 

[cucker tarlson]

what I do is find my stable oc settigs and then add +0.010-0.015 on top of that just to be sure.
my 5775c laughs at 1.41v anyway so I can afford it,barely hits 80W package TDP in IBT max.My D15S doesn't even go full 1500rpm I think it's not a bad idea to add a little bit of voltage if you have temperature headroom.

 

[lexluthermiester]

I have a very good Z370 mini ITX motherboard a Asrock Fatal1ty z370 Gaming-ITX/ac, I use my i5 9600k at 5ghz with a NZXT Kraken x62, and I know today is the AMD the king of desktop CPUs, but as I counted the Ryzen 9 3900x would be very expensive in my country to change and I would need to buy better memory, I'm using a 16GB Corsair Vengance 3000mhz CL16 kit, which perfect for Intel, but not for AMD. What do you think, i9 9900k would be futureproof, I would use about 2 years. And my graphics card is RTX 2080 Super.

Given that you already have a solid platform, upgrading your CPU to 9900k would bring you the best value and it will last you for several years. This presumes you need that much CPU power. If most of what you do is gaming, your 9600k will be good for a few years also. It just depends on your needs and usage model.

 

[Vario]

Ok, I'm going full off-topic here, but do people really run their CPUs at max overclock? I mean, if a stress tool told me the CPU is stable at, say, 5GHz, I'd dial that back 5% or so if I was going for stability.

I run it well under, lately just run as far as the stock voltage takes me and leave it at that. There isn't any reason to run it at max the speed is negligible. I've gone to 5.1 before and it didn't make any difference that I could tell except generate a lot of heat.

 

[Cheeseball]

1440p gaming at 75hz, because my monitor only handle that refresh rate (it's an UltraWide monitor) and I have a youtube channel, where I want stream, but I wasn't ready to start yet and also I edit my videos (1080p@60fps) with this CPU, sometimes very long with high bitrate (12000).

Your 9600K @ 5 GHz on a Z370 is more than enough for what you are currently doing with it (which is normal gaming and light video editing).

Once you start streaming though, you may want to (1) offload the streaming to another dedicated PC, (2) use NVENC (if you're planning to go Twitch.tv) or (3) upgrade the CPU to the 9900K if you're planning to use software encoding.

 

[ERazer]

the question is how quick can you sell the 9600k? think about that before you upgrade. I cant sure afford 200$ worth CPU collecting dust.

 

[Cheeseball]

the question is how quick can you sell the 9600k? think about that before you upgrade. I cant sure afford 200$ worth CPU collecting dust.

If the 9600K is $199 brand new at NewEgg.com, selling it would probably be around $150 or $130 depending on how new it is and if it still has a running warranty.

I still would recommend keeping it and only upgrading when needed, as it can handle all current games at 5 GHz with ease (it can also do it at stock speeds as well).

 

[PeterPaul0808]

the question is how quick can you sell the 9600k? think about that before you upgrade. I cant sure afford 200$ worth CPU collecting dust.

I started advertise my i5 9600k on used market place in my countrey at reasonable price. But noone needs today a Core i5, because everybody is in Ryzen frenzy. I will keep it, and thank you everyone who answered me.
P.S.: With NVENC I tried out the streaming, it's okay enjoyable.

 

[droopyRO]

I have a similar dilemma. Should i upgrade from an i5 8600k to a 9700k or a Ryzen 3700X. I made this small test, 4 cores/4 threads got me bellow 30 fps in this battle, 6c/6t to about 30 fps, the question is would another 2 cores and 12 threads get me to 40 fps ?

 

[EarthDog]

I have a similar dilemma. Should i upgrade from an i5 8600k to a 9700k or a Ryzen 3700X. I made this small test, 4 cores/4 threads got me bellow 30 fps in this battle, 6c/6t to about 30 fps, the question is would another 2 cores get me to 40 fps ?

Look up some reviews on CPU scaling for this titles and see...

Total War: Warhammer benchmarks strike fear into CPUs

An upcoming DX12 patch will help.

www.pcgamer.com

 

[droopyRO]

Those are irrelevant, the in-game benchmark is no way representative of an actual 40 vs 40 units like i posted above.

 

[EarthDog]

Those are irrelevant, the in-game benchmark is no way representative of an actual 40 vs 40 units like i posted above.

I mean, who knew unless you mentioned it initially...

It should still be relevant and scale... the more things on a screen that the CPU has to deal with...........right?

 

[dirtyferret]

I have a similar dilemma. Should i upgrade from an i5 8600k to a 9700k or a Ryzen 3700X. I made this small test, 4 cores/4 threads got me bellow 30 fps in this battle, 6c/6t to about 30 fps, the question is would another 2 cores and 12 threads get me to 40 fps ?

Your best bet would be the total war forums as they would have far more experience with the game on various hardware.

It could be your CPU or it could be your GPU causing the low FPS. Have you tried lowering graphic settings and does it help much?

From personal experience I went from a 8600k to a 9700k and it wasn't a drastic change in performance since my 8600k handled all my games with ease already but I didn't play total war games. Since you already have the mobo "upgrading" to the 9700k would be the most economical move and that CPU generally beats the Ryzen 3700x in gaming benchmarks. OC the 9700k will further the separation.

AMD Ryzen 7 3700X review: gaming performance tests

The buzz surrounding the third generation Ryzen has been extraordinary, with AMD's new range of processors taking the f…

www.eurogamer.net