Replace i7-5930k for GTX-1080ti with i9-9900k or waite ?

baha_timimi

Distinguished
Mar 8, 2010
156
1
18,715
Hi everyone.
I have desktop specification below:
CPU i7-5930k with good air cooler
GPU GTX1080ti MSI gaming X
corsair 850w gold power supply
32GB 2400 DDR3
Asus deluxe X99 2011v3
Team 256Gb nvme pci-express for system
WD 4TB HDD
MG278Q gaming monitor 1ms 144Hz 1440 pix

Is it worth to upgrade the cpu with a new one because its generation is old compared to the cpu's now in market.
Also, is it good idea to replace it with i9-9900k.
Note: My cpu (i7-5930k) situation is very good and it is cool where the temp. no exceed 66 C in full load.
But, could it can support games like Tom Clancy's Ghost Recon - Wildlands in high frame per seconds. because I tested the game at ultra setting but the FRS max was less than 60?
 
Solution
While your 5930k is a few years old, Intel has made very little progress in CPUs since its release. I would not make the change now. I would wait till next year and see what Intel's 10nm or AMD's 7nm chips do. Then they may be a better upgrade. But right now it wont help much. Especially at 1440p. When you move up in resolution, the bottleneck moves from the CPU to the GPU.

As for Wildlands, that game is just horribly optimized. I could only get it over 60fps with running 1080ti sli. But dropping it down to high settings still looks really good and your fps goes up fairly well.
While your 5930k is a few years old, Intel has made very little progress in CPUs since its release. I would not make the change now. I would wait till next year and see what Intel's 10nm or AMD's 7nm chips do. Then they may be a better upgrade. But right now it wont help much. Especially at 1440p. When you move up in resolution, the bottleneck moves from the CPU to the GPU.

As for Wildlands, that game is just horribly optimized. I could only get it over 60fps with running 1080ti sli. But dropping it down to high settings still looks really good and your fps goes up fairly well.
 
Solution
Many games are limited more by the single thread performance than by the number of threads.
To get some idea if this applies to your games, try these experiments:

1. REDUCE your cpu capability by perhaps 20%.
in windows cpu power management, set the max to 80%
If this shows up as significantly worse performance, it indicates that cpu core speed is an issue for your games.

2. In a similar vein, reduce the number of threads from the 12 you have to something less perhaps to 8.
You can do this in the windows msconfig boot advanced options option.
You will need to reboot for the change to take effect.
This will tell you how sensitive your games are to the benefits of many threads.
If you see little difference, your game does not need all the threads you have.

Your i7-5930K has a passmark rating of 13632. A very impressive number when all threads are fully busy.
The single thread rating is 2100.

For the i9-9900K with 16 threads the rating is 20224/2904. An impressive cpu boost for either single or multithreaded operation.

If budget is an issue, look at the 8 thread i7-9700K with a rating of 17368/2833

All that is nice but WORTH is something only YOU can determine.

And, you might test out your ability to run a stronger graphics card like the RTX2080ti.

Run YOUR games, but lower your resolution and eye candy.
If your FPS increases, it indicates that your cpu is strong enough to drive a better graphics configuration.
If your FPS stays the same, you are likely more cpu limited.


 

xxxlun4icexxx

Honorable
Jun 13, 2013
519
5
11,065


I'd maybe wait. You lose quad channel memory capability and the extra pci-e lanes to get some decent performance boosts but I don't know if it justifies the cost right now. The 5930k is a hidden gem monster. I ran mine into the ground and have 0 regrets about running it so long. Heck people are still buying them new where they can for upgrade.

Are you sure it's your cpu bottlenecking you? I feel as it shouldn't have any issue until you get into pretty high frame rate gaming.
 

baha_timimi

Distinguished
Mar 8, 2010
156
1
18,715


thanks so much for useful information. I afraid that just me have this FPS.
And to know, the CPU set at default speed without OC.
 

baha_timimi

Distinguished
Mar 8, 2010
156
1
18,715


Thanks for new information especially about (msconfig) option to set the nimbers of active cores.
I will try this and see where the bottleneck.
 
Nov 1, 2018
2
0
10
I have the same CPU... I did some research and I think* 17000 mark is minimum 4k encoding for live streams simultaniously , as a server with an Xeon with that mark was used a while back to broadcast multiple 4k streams, I read. To put into a better perspective, a recommended (not the minimum) spec for project cars 2 is 11000... I've seen just the slightest differences in performance of a GPU with changing out different CPUs.

wildlands says a 3770 i7 rather than say project cars 2 for a i7 6700.

Also, the asus bios is a little weird, 2400 ram is a strange number for that cpu as i have locked it up at that number 2133 runs fine... I was thinking it should be DDR 4 right?
 
i just moved up from a X99 mobo with a 5960x to a Z390 with an i9-9900K - i was extremely attracted to the higher clock freq as i saw the effect of tweaking my 5960x to 4.3 had on rendering videos, or the time to render

i've only had the 9900k running for less than a week and could not be happier. I'm only at 4.9GHz clock speed, because of temps - i'm running 66-75C rendering with 100% cpu load. Both CPUs have the same core/thread count, with the x99 & 5960x supporting quad channel memory and iirc 40 PCI lanes, where the z390 & 9900k only provide 16. So i was curious (and a little worried) my experiment wasn't going to work out.

Same video file went from 41 minute render time to 27 minutes. But there was another factor, the Z390 mobo (Gigabyte AORUS Master) has two M.2 ports that run full 32 Gb/ps that do not dis-able any Sata ports, but on the X99 i had 10 Sata ports, only six on the Z390 board. The X99 dis-abled two Sata ports so the difference drops some, and with that Asus m.2 to PCIe card that holds 4 NVMe PCIe SSDs, that difference is removed (not sure i'll be able to use all four as that card is meant for Asus X299 mobo.

But clock speed makes a helluva difference in my usage - i can't speak to gaming.

 

baha_timimi

Distinguished
Mar 8, 2010
156
1
18,715


Impressive information.
I am sure the game performance is higher than i7-5960x.
Do you mean that the max. temp. is 75 C at 4.9Gh?
is it safe with this temp?
Dose it effect the life span for CPU?
What type of cooler for both CPUs?
And at last. What GPU you used ?
Finally, thanks for you for good experimental informations.
 


Impressive information.
I am sure the game performance is higher than i7-5960x.
Do you mean that the max. temp. is 75 C at 4.9Gh? YES
is it safe with this temp? YES
Dose it effect the life span for CPU?
What type of cooler for both CPUs? In my signature below, Noctua NH-D15S
And at last. What GPU you used ? Using a Zotac GTX 1070, but most of the rendering /editing programs i use run / rely on the CPU & the software, not on the GPU
Finally, thanks for you for good experimental informations.
[/quotemsg]

Again i can't speak to gaming as i just don't game, but most of these guys are pretty sharp on what games need or want multi-cores and hyper threading

As far as heat degrading the CPU - the CPU has safety thermal limits built in to protect it, Most seem to throtte down at 100 or 105c. They work pretty well at protecting the CPU. When i built my first computer with a 4790(non-k), i was relying on ASUS's AI Suite III performance utility for temp monitoring.

I used the rig for video rendering and noticed at 98-100% load, it seemed to always plateau out at 67C. The two rendering programs i use seem to be core hogs and run the CPU full tilt.

One day i downloaded RealTemp and then intel's XTU and both showed temps were actually running 98-100C, ie it would hit 100C, fall back to 98C for a few seconds then go right back up to 100C and continually repeat the cycle. I corrected the issue, ie better cooler, and started using the benchmark test in XTU (Intel's Extreme Tuning Utility) to test my CPU. As that test offers the opportunity to upload your benchmark score to the internet and compare it to other users with same CPU and mobo., I would

- my scores are consistently in the top 5% - bear in mind, i rendered 1-2 video files a day, 2-3 hour jobs, 5 days a week, for over 3 months, 5-6 days/week unaware i was running it at 98-100C - that was almost 4 years ago. I'm pretty confident no permanent damage was done to the CPU

But i don't recommend that test on your CPU - i prefer to keep my temps as low as possible, at least under 80C
 

bobbybluz

Distinguished
Mar 6, 2010
32
1
18,545
I have the same mobo and speed RAM you do but 64GB. I was running a 5930K until last summer when I got a J code 5960X for $300 off Ebay. I also sold the 5930K on there for $275. Both CPU's were stable at 4.7GHZ (heavy duty liquid cooling) at default voltage setting. The 5930K was good for 17,451 on Passmark. The 5960X got 21768 with the same exact settings. I do video and audio production work. The extra two cores do indeed make a difference. The system idles at 80 degrees F and peaks around 150 F under heavy loads. It also is loud but I'm used to that. For $25 it was worth making bids on 5960X's until a seller said yes. That's far cheaper than a new mobo and CPU will cost you plus it drops right in.