Discussion My System Performed 30% worse than identical systems

Eventine

Prominent
Mar 9, 2019
3
0
510
Hi everyone!
Yesterday I felt like my system wasn't performing well, so I decided to run a 3DMark test to confront to similar systems. I was shoked by the results!
https://www.3dmark.com/compare/spy/12333315/spy/7153068/spy/4764505/spy/1194463/spy/1282652# (my system is the first)
My Graphic card perfomed 30% worse than the same Card with the same OC. I want to note that I was recording the temp and neither CPU nor GPU temperature was above 67°C.
Also I decided to not OC my CPU over 4.3GHz because I already burnt a CPU (note that I followed all tutorials and I'm running a Water cooling AIO, and the CPU broke while at 4.5 GHz and a temp < 70 °C ).

Can someone enlighthen me on the results? Can you give me some advices? Can it be the PSU? (I have a 750W OCZ gaming PSU, not the best, but I bought it 9 years ago and I was a noobie, but it shouldn't be that bad, no?)

Thanks to everyone
Eventine
 

Phaaze88

Titan
Ambassador
You're comparing your score to Gaunknown and Marc-4362?

1)Did you disable all power saving features for your gpu? The average core and memory frequencies tell me you didn't, but perhaps something else is responsible for that.

2)The other 2 users went a little further on their gpu OCs than you did.
-Marc pretty much matches you on core clock, but Gaunknown went even further by almost another 100mhz.
-You didn't touch the memory frequency, but the other 2 bumped it by at least 100mhz.

3)The other 2 users also have a lead over you with their cpu overclocks as well.
4.3 VS 4.6 VS 4.5ghz.
Those OCs do allow for the cpu to feed a few more frames to the gpu to process.

4)3D Mark doesn't like to post ram speeds for some reason. They really should update that, as it can make a difference in some systems.
So for the time being, this is an unknown. They could also have an edge over you in this as well.

Conclusion: You simply didn't push your system as hard as they did theirs.
 
  • Like
Reactions: King_V

Eventine

Prominent
Mar 9, 2019
3
0
510
You're comparing your score to Gaunknown and Marc-4362?

1)Did you disable all power saving features for your gpu? The average core and memory frequencies tell me you didn't, but perhaps something else is responsible for that.

2)The other 2 users went a little further on their gpu OCs than you did.
-Marc pretty much matches you on core clock, but Gaunknown went even further by almost another 100mhz.
-You didn't touch the memory frequency, but the other 2 bumped it by at least 100mhz.

3)The other 2 users also have a lead over you with their cpu overclocks as well.
4.3 VS 4.6 VS 4.5ghz.
Those OCs do allow for the cpu to feed a few more frames to the gpu to process.

4)3D Mark doesn't like to post ram speeds for some reason. They really should update that, as it can make a difference in some systems.
So for the time being, this is an unknown. They could also have an edge over you in this as well.

Conclusion: You simply didn't push your system as hard as they did theirs.
Thanks for the answers.
Where can I disable the power savings settings? Only in the drivers or also windows?

I ran another test later, and got a similar result (above 6k score) to the ones I'm comparing with, then I ran the test again and I got back to 5k score.
 
How on earth you killed your last cpu, was it also i5-6600k? Ive been running systems overclocked over all limits past decade and more and only seen slight decrease in performance for 2 cpus out of 20, more vcore needed etc.

Maybve your PSU is giving wrong voltages have you checked 12v 5v 3.3v and vcore is stable in BIOS monitoring ?
 

Karadjgne

Titan
Ambassador
You got just over 4k cpu. The closest guy to you has 100MHz advantage, got 4300. The guy with a 500MHz advantage got 4777.

Your cpu is right on track. It's also going to affect gpu outputs since the gpu can only render what the cpu gives it, so the other two Vegas will naturally score higher running similar clocks.

Your cpu is wearing shackles on its ankles and you wonder why it comes in last in a 100meter dash.

That OCZ isn't doing you any favors, lousy outputs making for stability issues. And you cannot follow proscribed stuff for OC, every cpu is different, so just setting what someone else did doesn't work. It might seem ok, but it never really is. OC theory is right, it's the same across any of those cpus, regardless of motherboard, but that's theory. Practice and theory are not the same.
 
Last edited: