F5F Stay Refreshed Power Users Overclocking Silicone lottery, heat

Silicone lottery, heat

Silicone lottery, heat

I
iNaomiPlays
Senior Member
609
04-18-2018, 07:22 AM
#1
As far as I've grasped it, the silicone lottery concerns the amount of frequency you can extract from your CPU using minimal voltage. However, could it also depend on temperature? In other words, does higher heat affect performance beyond just voltage limits?

I'm using 1.36 volts through my 2600 with a crosshair vii main board. The heat sink is a 280mm kraken equipped with some aftermarket Vardar fans. I've managed to keep my 5820k under 70°C at all times while running at 1.3 4.5ghz.

This CPU reaches 80 degrees in an air-conditioned room during peak 95 tests. I've inspected the mounting and Re applied the paste (thermal grizzly). It usually stays in the 60s while playing blackout with spikes into the low 70s—all threads are getting decent use in that game, and it sits in the 50s in other games.

I'm fine with the temperatures for general use, but I was surprised by how hot it gets with a large radiator. Also, increasing the fans to 100% only makes it take longer to hit 80°C.
I
iNaomiPlays
04-18-2018, 07:22 AM #1

As far as I've grasped it, the silicone lottery concerns the amount of frequency you can extract from your CPU using minimal voltage. However, could it also depend on temperature? In other words, does higher heat affect performance beyond just voltage limits?

I'm using 1.36 volts through my 2600 with a crosshair vii main board. The heat sink is a 280mm kraken equipped with some aftermarket Vardar fans. I've managed to keep my 5820k under 70°C at all times while running at 1.3 4.5ghz.

This CPU reaches 80 degrees in an air-conditioned room during peak 95 tests. I've inspected the mounting and Re applied the paste (thermal grizzly). It usually stays in the 60s while playing blackout with spikes into the low 70s—all threads are getting decent use in that game, and it sits in the 50s in other games.

I'm fine with the temperatures for general use, but I was surprised by how hot it gets with a large radiator. Also, increasing the fans to 100% only makes it take longer to hit 80°C.

_
_lolGavin
Member
52
04-18-2018, 11:49 AM
#2
The "silicon lottery" reflects how effectively your actual CPU performs.
Within a specific model, they vary greatly. Some can be boosted to 5.2GHz, while others cap at 4.9GHz.
Regardless of cooling setup or voltage applied, the performance remains consistent.
At 80°C, thermal throttling does not occur.
Attempting higher GHz or voltage will cause temperatures to rise.
When you reach a point where this CPU can no longer operate at any higher GHz—no matter your cooling or voltage adjustments—you've reached its performance ceiling.
_
_lolGavin
04-18-2018, 11:49 AM #2

The "silicon lottery" reflects how effectively your actual CPU performs.
Within a specific model, they vary greatly. Some can be boosted to 5.2GHz, while others cap at 4.9GHz.
Regardless of cooling setup or voltage applied, the performance remains consistent.
At 80°C, thermal throttling does not occur.
Attempting higher GHz or voltage will cause temperatures to rise.
When you reach a point where this CPU can no longer operate at any higher GHz—no matter your cooling or voltage adjustments—you've reached its performance ceiling.

1
101pate
Member
73
04-18-2018, 03:51 PM
#3
Using the tools you have set up for monitoring temperatures.
1
101pate
04-18-2018, 03:51 PM #3

Using the tools you have set up for monitoring temperatures.

N
NRHuff1024
Member
126
04-21-2018, 03:42 PM
#4
Volkgren:
Some apps show increased temperatures with Ryzen because of a discrepancy. Which tools are you checking for temperature?
Hi,
I'm using Core Temp, MSI Afterburner, and NZXT CAM. They all give consistent readings.
N
NRHuff1024
04-21-2018, 03:42 PM #4

Volkgren:
Some apps show increased temperatures with Ryzen because of a discrepancy. Which tools are you checking for temperature?
Hi,
I'm using Core Temp, MSI Afterburner, and NZXT CAM. They all give consistent readings.

O
OoTheBearoO
Junior Member
11
05-04-2018, 11:48 PM
#5
I just purchased the Kraken X62 and I understand CAM works best at the right temperature without any offset. Make sure the pump is set to high in CAM for optimal results. Also, you should be able to operate at lower voltage. Running prime 95 with auto voltage on gives around 1.25v or close to it using the same motherboard and Ryzen 7 2700X.
O
OoTheBearoO
05-04-2018, 11:48 PM #5

I just purchased the Kraken X62 and I understand CAM works best at the right temperature without any offset. Make sure the pump is set to high in CAM for optimal results. Also, you should be able to operate at lower voltage. Running prime 95 with auto voltage on gives around 1.25v or close to it using the same motherboard and Ryzen 7 2700X.

W
Wediocre
Member
56
05-05-2018, 03:13 PM
#6
Volkgren:
I just purchased the Kraken X62 and I’m aware that CAM handles temperatures correctly without offsets. Make sure the pump is set to high in CAM for optimal performance. Also, you should be able to operate at a lower voltage. Running Prime 95 with auto voltage on will keep it around 1.25V or close to that using the same motherboard and Ryzen 7 2700X.

I own the X61.
It looks like you can only adjust fan speed through CAM?
Regarding the voltage, do you have an OC method applied? From what I understand, the gains are limited if you’re overclocking the x chips. So it doesn’t seem worth it, whereas non-X models definitely offer better results.
I’m set at 4.1GHz.
My stability errors in Prime 95 and ASUS Real Benchmark will drop significantly if voltage drops.
W
Wediocre
05-05-2018, 03:13 PM #6

Volkgren:
I just purchased the Kraken X62 and I’m aware that CAM handles temperatures correctly without offsets. Make sure the pump is set to high in CAM for optimal performance. Also, you should be able to operate at a lower voltage. Running Prime 95 with auto voltage on will keep it around 1.25V or close to that using the same motherboard and Ryzen 7 2700X.

I own the X61.
It looks like you can only adjust fan speed through CAM?
Regarding the voltage, do you have an OC method applied? From what I understand, the gains are limited if you’re overclocking the x chips. So it doesn’t seem worth it, whereas non-X models definitely offer better results.
I’m set at 4.1GHz.
My stability errors in Prime 95 and ASUS Real Benchmark will drop significantly if voltage drops.

F
Floris98
Junior Member
35
05-05-2018, 11:01 PM
#7
I've tried both OC and non-OC configurations. The first time I ran Prime95 it reached 100C. I'm unsure why, as I checked everything. I had an OC of 4.2GHz with auto voltage at 1.45v, but after that I looked up pump speed in CAM and set it to 100% – it hasn't changed since.
F
Floris98
05-05-2018, 11:01 PM #7

I've tried both OC and non-OC configurations. The first time I ran Prime95 it reached 100C. I'm unsure why, as I checked everything. I had an OC of 4.2GHz with auto voltage at 1.45v, but after that I looked up pump speed in CAM and set it to 100% – it hasn't changed since.

S
SnifePvP
Posting Freak
872
05-07-2018, 02:57 PM
#8
Volkgren:
I've tried both OC and non-OC configurations. The first time I used Prime95 it reached 100C. I'm unsure why, after checking everything. I had an OC of 4.2GHz with auto voltage at 1.45v, but once I adjusted the pump speed in CAM to 100% it hasn't worked since. Could you share a screenshot showing the pump settings? There have been many times when I found what I needed right in front of me. Maybe the X61 doesn't offer manual pump control—though I can only see fan control.
S
SnifePvP
05-07-2018, 02:57 PM #8

Volkgren:
I've tried both OC and non-OC configurations. The first time I used Prime95 it reached 100C. I'm unsure why, after checking everything. I had an OC of 4.2GHz with auto voltage at 1.45v, but once I adjusted the pump speed in CAM to 100% it hasn't worked since. Could you share a screenshot showing the pump settings? There have been many times when I found what I needed right in front of me. Maybe the X61 doesn't offer manual pump control—though I can only see fan control.