F5F Stay Refreshed Power Users Overclocking I just completed my first OverClock and have a few questions.

I just completed my first OverClock and have a few questions.

I just completed my first OverClock and have a few questions.

K
KingKylePwns
Junior Member
3
12-18-2017, 05:45 PM
#1
Just completed my first overclock on my i5-7600k, it’s currently running an EVO 212 and isn’t a great motherboard (MSI Z270-A Pro). I haven’t adjusted the overclock settings yet, but the 4.2ghz version seemed to be limiting my performance in some games (100% CPU usage, 60% GPU, stuttering).

After doing some research, I changed the BIOS setting to disable Intel Speed Step, tweaked the CPU ratio, and adjusted the voltage for each run during Prim95 blend tests. Most of the time, one of the cores would drop and reduce CPU usage by about 25%.

I finally passed all 15 tests at 4.8ghz with a voltage of 1.31 volts. During testing, the actual clock speed fluctuated often, rarely hitting the intended overclock speed, and temperatures reached around 84-85°C. However, I noticed an overall improvement when playing games like PUBG, and NZXT Cam kept the CPU steady at 4.8ghz with load temps near 71°C.

Would this approach work? Does everything seem to be in order?
K
KingKylePwns
12-18-2017, 05:45 PM #1

Just completed my first overclock on my i5-7600k, it’s currently running an EVO 212 and isn’t a great motherboard (MSI Z270-A Pro). I haven’t adjusted the overclock settings yet, but the 4.2ghz version seemed to be limiting my performance in some games (100% CPU usage, 60% GPU, stuttering).

After doing some research, I changed the BIOS setting to disable Intel Speed Step, tweaked the CPU ratio, and adjusted the voltage for each run during Prim95 blend tests. Most of the time, one of the cores would drop and reduce CPU usage by about 25%.

I finally passed all 15 tests at 4.8ghz with a voltage of 1.31 volts. During testing, the actual clock speed fluctuated often, rarely hitting the intended overclock speed, and temperatures reached around 84-85°C. However, I noticed an overall improvement when playing games like PUBG, and NZXT Cam kept the CPU steady at 4.8ghz with load temps near 71°C.

Would this approach work? Does everything seem to be in order?

G
211
12-18-2017, 07:04 PM
#2
yes anything above factory spec is regarded as an OC you did your research and that's a win
G
george_griveas
12-18-2017, 07:04 PM #2

yes anything above factory spec is regarded as an OC you did your research and that's a win

C
crazypotpie
Member
225
12-18-2017, 11:11 PM
#3
yes anything above factory spec is regarded as an OC you did your research and that's a win
C
crazypotpie
12-18-2017, 11:11 PM #3

yes anything above factory spec is regarded as an OC you did your research and that's a win

K
kc14867
Junior Member
20
12-22-2017, 01:42 PM
#4
it seems promising. you can always perform a R20 test to apply stress on all the cores, benchmark, and verify the clocks once more.
K
kc14867
12-22-2017, 01:42 PM #4

it seems promising. you can always perform a R20 test to apply stress on all the cores, benchmark, and verify the clocks once more.

Z
zCosmicCorvus
Member
212
12-29-2017, 02:49 AM
#5
What is the R20 test? It's something you'll likely encounter when considering overclocking, even if you haven't done it before.
Z
zCosmicCorvus
12-29-2017, 02:49 AM #5

What is the R20 test? It's something you'll likely encounter when considering overclocking, even if you haven't done it before.

K
KARLEISHAK
Member
197
01-07-2018, 07:28 AM
#6
Cinebench r20 evaluates your CPU performance and provides a score. Another benchmark, Geekbench 5, lets you see how others are doing with their overclocked systems.
K
KARLEISHAK
01-07-2018, 07:28 AM #6

Cinebench r20 evaluates your CPU performance and provides a score. Another benchmark, Geekbench 5, lets you see how others are doing with their overclocked systems.

L
60
01-09-2018, 12:25 AM
#7
Ah Cinebench. I haven't heard it referred to as R20 before, which surprised me. Yes, Geekbench works well too. A solid method for establishing a baseline and comparing others for reference.
L
lovelydisaster
01-09-2018, 12:25 AM #7

Ah Cinebench. I haven't heard it referred to as R20 before, which surprised me. Yes, Geekbench works well too. A solid method for establishing a baseline and comparing others for reference.

C
cheleen
Member
116
01-09-2018, 12:31 AM
#8
Why would, "the actual clock speed would go all over the place rarely hitting the actual oc speed, temps would hit 84-85C but it passed all the test. " Be normal? Normally if I manually overclock my cpu it stays at 4.8GHz in testing or whatever the AVX offset is under avx loads. So I would not expect the clock frequency to change by much.
The clock speed would be expected to change if say something was overheating. This could be the cpu or vrms.
C
cheleen
01-09-2018, 12:31 AM #8

Why would, "the actual clock speed would go all over the place rarely hitting the actual oc speed, temps would hit 84-85C but it passed all the test. " Be normal? Normally if I manually overclock my cpu it stays at 4.8GHz in testing or whatever the AVX offset is under avx loads. So I would not expect the clock frequency to change by much.
The clock speed would be expected to change if say something was overheating. This could be the cpu or vrms.

T
The_Gamer_932
Junior Member
12
01-15-2018, 05:28 AM
#9
Looks like room for 4.9, @ perhaps 1.35v ish, 71c is perfectly fine on gaming load, tbh up to 75 is also absolutelly fine.
also, i must say that this processor doesn't bottlenecks, even with a 2080ti, and even on stock clocks, but OC improvements if done right are clearly noticeable.
T
The_Gamer_932
01-15-2018, 05:28 AM #9

Looks like room for 4.9, @ perhaps 1.35v ish, 71c is perfectly fine on gaming load, tbh up to 75 is also absolutelly fine.
also, i must say that this processor doesn't bottlenecks, even with a 2080ti, and even on stock clocks, but OC improvements if done right are clearly noticeable.