F5F Stay Refreshed Power Users Overclocking Ryzen 1600 stability

Ryzen 1600 stability

Ryzen 1600 stability

T
TheKawaiiMelon
Junior Member
46
11-19-2017, 06:32 AM
#1
for more than ten days i kept running my Ryzen 1600 @3.9Ghz with 1.32v, which is really good. Aida64 reports stress tests run smoothly without crashes, reaching a max of 72°C, but Prime95 (slow FTT) crashes after five minutes at around 82°C, though it remains stable at 3.9 at 1.35v and 83°C. I actually played a lot of BF1 with 3.9 at 1.32v without any issues. Which test should I focus on? The differences are significant and confusing. One more question: what is the NB CPU/SoC voltage? The default is 0.9v, but when my CPU + RAM overclock it to 1.15 or 1.2... should I adjust that manually or keep it at 1.0v?
T
TheKawaiiMelon
11-19-2017, 06:32 AM #1

for more than ten days i kept running my Ryzen 1600 @3.9Ghz with 1.32v, which is really good. Aida64 reports stress tests run smoothly without crashes, reaching a max of 72°C, but Prime95 (slow FTT) crashes after five minutes at around 82°C, though it remains stable at 3.9 at 1.35v and 83°C. I actually played a lot of BF1 with 3.9 at 1.32v without any issues. Which test should I focus on? The differences are significant and confusing. One more question: what is the NB CPU/SoC voltage? The default is 0.9v, but when my CPU + RAM overclock it to 1.15 or 1.2... should I adjust that manually or keep it at 1.0v?

I
ItsDrAxel
Member
113
11-24-2017, 08:41 PM
#2
There isn't one definitive test to confirm stability. The most effective approach combines several stress tests with practical real-world usage. This is important since various programs utilize the CPU in distinct manners, and the CPU itself contains numerous components that function differently.
I
ItsDrAxel
11-24-2017, 08:41 PM #2

There isn't one definitive test to confirm stability. The most effective approach combines several stress tests with practical real-world usage. This is important since various programs utilize the CPU in distinct manners, and the CPU itself contains numerous components that function differently.

R
RadicalTitan
Member
65
12-09-2017, 01:39 AM
#3
my regular usage is mostly gaming, with a small part on the web, not too demanding.
if there are various specs for testing, what should i focus on?
R
RadicalTitan
12-09-2017, 01:39 AM #3

my regular usage is mostly gaming, with a small part on the web, not too demanding.
if there are various specs for testing, what should i focus on?

H
horselover328
Member
148
12-12-2017, 10:19 AM
#4
for certain individuals - stability means consistency in all areas.
others rely on those partially stable conditions that support gaming, yet encounter issues elsewhere.
H
horselover328
12-12-2017, 10:19 AM #4

for certain individuals - stability means consistency in all areas.
others rely on those partially stable conditions that support gaming, yet encounter issues elsewhere.

1
111carys111
Posting Freak
832
12-12-2017, 05:59 PM
#5
the nb cpu/soc voltage is typically 0.9v by default, but when your cpu and ram are active it can increase to 1.15 or 1.2v. you should adjust it accordingly rather than leaving it manually at 1.0v.
1
111carys111
12-12-2017, 05:59 PM #5

the nb cpu/soc voltage is typically 0.9v by default, but when your cpu and ram are active it can increase to 1.15 or 1.2v. you should adjust it accordingly rather than leaving it manually at 1.0v.

S
seomilk77
Member
181
12-12-2017, 07:11 PM
#6
NB mainly handles memory controller, which is something to consider.
The SoC is mainly dedicated to the iGPU, unlike XD.
S
seomilk77
12-12-2017, 07:11 PM #6

NB mainly handles memory controller, which is something to consider.
The SoC is mainly dedicated to the iGPU, unlike XD.

O
OwnerOfHive
Member
105
12-13-2017, 08:22 AM
#7
In my mind, stability tests should only be performed if instability occurs with typical and demanding programs; otherwise, they're just unnecessary work to fit into normal parameters. OCCT is my preferred tool, providing graphs during execution and allowing me to report crashes promptly. I believe a maximum of one hour should suffice to identify any issues.
O
OwnerOfHive
12-13-2017, 08:22 AM #7

In my mind, stability tests should only be performed if instability occurs with typical and demanding programs; otherwise, they're just unnecessary work to fit into normal parameters. OCCT is my preferred tool, providing graphs during execution and allowing me to report crashes promptly. I believe a maximum of one hour should suffice to identify any issues.

C
Crinklekitty
Member
61
12-15-2017, 07:27 AM
#8
after extensive testing, i explored various ram speeds combined with different cpu frequencies and voltages. i found that using the CINEBENCH R15 provided the most reliable stability assessment. although it seems a bit awkward, i consistently rendered 8 out of 10 cpus. this method lets me monitor maximum temperatures after numerous runs, and if there are no crashes, it confirms stability. i also use hwinfo to analyze performance data, which helps identify the optimal cpu voltage for that load—like setting it at 1.32v when i found max usage at 3.85c with 1.35v. this adjustment achieved full stability.
C
Crinklekitty
12-15-2017, 07:27 AM #8

after extensive testing, i explored various ram speeds combined with different cpu frequencies and voltages. i found that using the CINEBENCH R15 provided the most reliable stability assessment. although it seems a bit awkward, i consistently rendered 8 out of 10 cpus. this method lets me monitor maximum temperatures after numerous runs, and if there are no crashes, it confirms stability. i also use hwinfo to analyze performance data, which helps identify the optimal cpu voltage for that load—like setting it at 1.32v when i found max usage at 3.85c with 1.35v. this adjustment achieved full stability.