Ryzen 5 3600 very high temps

[Deleted member 185158]

I know about boost because the FX uses it but I disabled mine along with power states and CnQ, its why I can run 5.0 all day lol.

Indeed, that was mainly for the original poster of the thread to help him understand it a little about how it works.

However, FX chips use P-states for boost frequencies while Ryzen chips do not.
So a max boost could be IE: 2700x 4ghz on auto, PBO OC 4.1ghz and 4.2ghz respectively.

 

[tabascosauz]

Thermal design point is a figure that has around 5% give and take. The 65w number means nothing besides the amount of BTU per hour it can produce under load.

I'm sorry, I'm going to have to call you out on this one. TDP does matter for the 65W Ryzen 3000 parts, and stock boost algorithm is strict about not exceeding roughly 75-80W power draw with said TDP. That's not up for debate; it's plain to see with the CPU scaling back the clocks to compensate in very heavy stress tests.

Also, Ryzen 3000 definitely does not wait until 95c to start throttling. That's simply untrue.

 

[Deleted member 185158]

The thermal block contains all the features related to temperature sensing, control, and reporting. It includes:
• Temperature collection and calculation using TCON (digital control logic) and TMON and Remote Diode Interface macros.
• Fan speed control for off-chip fans.
• Temperature reporting through the SMBUS interface

So from your angle, 65W TDP (thermal design point) needs a cooler capable of removing about 222 BTU/hr.
1v at 65 amps is 65W. Obviously this figure can be arranged a bunch of different ways.

AMDs angle for the term TDP is not exactly meaning power draw or usage while that is based on algorithms monitoring temperatures, but rather the amount of heat it can produce and is why it's called THERMAL DESIGN POWER.... not wattage design power.

 

[cucker tarlson]

Voltages are high, but amperage/wattage is low - so its irrelevant.

fantastic.
now I'm off to put 1.6v through my 65w cpu.the wattage will stay okay so really what can happen.

 

[Deleted member 185158]

fantastic.
now I'm off to put 1.6v through my 65w cpu.the wattage will stay okay so really what can happen.

lol, that's not exactly what he means dude....

You can use 1.6v if you had the ability to manually lower the AMPs. Just because you raise voltage, does not mean amps increase.
You can run 1.5v at 30 amps or 60 amps.... but it's still 1.5v.
The issue is that we cannot change the AMP input manually, the voltage regulators do most of that work. increasing and decreasing according to logic registers.

So 32 AMPs at 2v is only 65W

 

[cucker tarlson]

lol, that's not exactly what he means dude....

You can use 1.6v if you had the ability to manually lower the AMPs. Just because you raise voltage, does not mean amps increase.
You can run 1.5v at 30 amps or 60 amps.... but it's still 1.5v.
The issue is that we cannot change the AMP input manually, the voltage regulators do most of that work. increasing and decreasing according to logic registers.

So 32 AMPs at 2v is only 65W

well then he should be more specific

what is the amp input on ryzen 3000 then ? 40A ?

 

[Deleted member 185158]

well then he should be more specific.

what is the amp input on ryzen 3000 then ? 40A ?

I understood it, but he's right the AMPs don't really matter.... let me try to explain why.

So lets say you run your chip at 1.300v at 3.9ghz. If you increase ONLY the speed say to 4.1Ghz.... the chip runs hotter...

The real question is why does the chip run hotter even though the Voltage input and reading are the same for both frequencies.....

Why does AVX instructions create more heat than programs that do not use AVX instructions?

The answers are pretty simple.

A. transistors may as well be viewed as mechanical parts. The faster they move, the more heat created. No extra voltage needed but increases Thermal wattage output also known as BTU.

B. AVX instruction sets load cpu cache which also produces more heat.... even at the sustained voltage.

 

[cucker tarlson]

I understood it, but he's right the AMPs don't really matter.... let me try to explain why.

So lets say you run your chip at 1.300v at 3.9ghz. If you increase ONLY the speed say to 4.1Ghz.... the chip runs hotter...

The real question is why does the chip run hotter even though the Voltage input and reading are the same for both frequencies.....

Why does AVX instructions create more heat than programs that do not use AVX instructions?

The answers are pretty simple.

A. transistors may as well be viewed as mechanical parts. The faster they move, the more heat created. No extra voltage needed but increases Thermal wattage output also known as BTU.

B. AVX instruction sets load cpu cache which also produces more heat.... even at the sustained voltage.

okay,but what about the op's question though ?
is running 3600 on a puny heatsink with voltages around 1.4v and really high temps okay as long as he doesn't throttle?

 

[Deleted member 185158]

Hopefully next we could address leakage characteristics of processors.

okay,but what about the op's question though ?
is running 3600 on a puny heatsink with voltages around 1.4v and really high temps okay as long as he doesn't throttle?

Well, 95c is T-Case max, not Thermtrip. Yes he'd be ok.

 

[cucker tarlson]

Hopefully next we could address leakage characteristics of processors.

I just don't know why people are saying that.It was never okay to have a hot cpu running 1.4/1.5v but the boys say it's fine with 3000.what changed it's fine now ?

 

[Deleted member 185158]

That's not entirely true..... maybe for daily average users... the unknowing or those don't really care much about it.....

So it's always been a long debate with AMD processors running high voltage for extended periods of time. But past processors had a t-case of as low as 65c. That's quite tough to keep under control.

But what do you mean by ok? Degradation? Longevity of the cpu? ..... If your rig made it 10 years with OC.... then it's good?

However, the more enthusiast type would want cooler temps to reach a high max clock.

AMD took care of that for everyone. Everyone is overclocking even if they don't know it and it's now called a "boost".

Base clock for my chip is the Max P-state registered. You can view this by going to the cpu-z about tab, click save report txt and view it for yourself. You probably won't see any "boosting" p-states like you would with previous gen (AM3+ and earlier)

Leakage. Inquire on Leakage. Or binning. Maybe touch on that.

 

[cucker tarlson]

That's not entirely true..... maybe for daily average users... the unknowing or those don't really care much about it.....

So it's always been a long debate with AMD processors running high voltage for extended periods of time. But past processors had a t-case of as low as 65c. That's quite tough to keep under control.

But what do you mean by ok? Degradation? Longevity of the cpu? ..... If your rig made it 10 years with OC.... then it's good?

However, the more enthusiast type would want cooler temps to reach a high max clock.

AMD took care of that for everyone. Everyone is overclocking even if they don't know it and it's now called a "boost".

Base clock for my chip is the Max P-state registered. You can view this by going to the cpu-z about tab, click save report txt and view it for yourself. You probably won't see any "boosting" p-states like you would with previous gen (AM3+ and earlier)

Leakage. Inquire on Leakage. Or binning. Maybe touch on that.

I'm eyeing a 3700x for myself,but still have lots of doubts.
for example,why does it spike over 90-100w during gameplay ?

 

[Deleted member 185158]

I'm eyeing a 3700x for myself,but still have lots of doubts.
for example,why does it spike over 90-100w during gameplay ?

Well sure it can. Maybe not often but that is a power usage of electrical current. The BTU does not change THAT drastically. You would use an average of electrical wattage used to be converted to BTU/hr output hence Thermal Design Point

So yes, electrical wattage may fluctuate greatly while maintaining a 65w TDP or 222 BTU/hr.
\
Follow?

EDIT:
I will say that it drives me crazy that people get stuck between TDP and what HWMonitor tells them.
So for informational purposes. https://en.wikipedia.org/wiki/Thermal_design_power
Quoted from the wiki:
"sometimes called thermal design point, is the maximum amount of heat generated by a computer chip or component (often a CPU, GPU or system on a chip) that the cooling system in a computer is designed to dissipate under any workload. "

I really hope this information helps you guys/gals understand that a current wattage differs greatly from what TDP stands for.
Does TDP raise with overclocking = YES. Even if it's done automatically through SenseMi technologies or manually.
YES you can increase TDP by simply running a processor faster without changing the Wattage input of the processor.

 

[Mussels]

fantastic.
now I'm off to put 1.6v through my 65w cpu.the wattage will stay okay so really what can happen.

and heres someone who failed to understand. thanks for making my point.

*THIS IS NOT ABOUT A STATIC VOLTAGE FOR 24/7 ALL CORE LOAD. IT IS NOT THE SAME THING AT ALL. I HOPE ALL CAPS GETS THIS THROUGH TO YOU.*

 

[cucker tarlson]

and heres someone who failed to understand. thanks for making my point.

*THIS IS NOT ABOUT A STATIC VOLTAGE FOR 24/7 ALL CORE LOAD. IT IS NOT THE SAME THING AT ALL. I HOPE ALL CAPS GETS THIS THROUGH TO YOU.*

great,thanks for explaining it before lashing out.

 

[Agentak74u]

Damn guys this is too much, you guys are amazing .. I probably need few days to digest these info

 

[Chomiq]

I'm eyeing a 3700x for myself,but still have lots of doubts.
for example,why does it spike over 90-100w during gameplay ?

Might be the fault of default LLC set by the mobo.

 

[Xzibit]

I'm eyeing a 3700x for myself,but still have lots of doubts.
for example,why does it spike over 90-100w during gameplay ?

Read his Details

CPU: AMD Ryzen™ 5 3700X, 3.6 - 4.4GHz PBO+Auto OC

PBO depending on the settings hes using in PBO he can be allowing 35 watts more.

Could be this unreleased Ryzen 5 3700X uses more watts then actual Ryzen 7 3700X