cryptohunter
Posts: 39
Joined: Wed Apr 05, 2017 5:56 am

Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 8:20 am

Hi there,

I've been thinking about the implication of these words from the
wikipedia page on RP1, RP2 and RP3

quote:
"The CPU chips of the first and second generation Raspberry Pi board did not require cooling, such as a heat sink, unless the chip was overclocked...."
(source: https://en.wikipedia.org/wiki/Raspberry_Pi )

So.... RP1 and RP2 don't need cooling... (presumably, they don't overheat).
RP3... by implication.... requires an (aluminium) heat sink; it gets hot.

My question is: Is it true that the RP3 needs a heat sink ?
Is it true that RP2 does not need a heat sink ?

If the answer the both of those questions is YES, then i wonder if that
suggests that the lifespan of a RP2 is greater than the lifespan of a RP3.

Thank you to anyone that replies... I appreciate your help :-)

cryptohunter :)

jahboater
Posts: 5756
Joined: Wed Feb 04, 2015 6:38 pm
Location: West Dorset

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 8:45 am

The new Pi2 (v1.2) has the same ARMv8 based SoC as the Pi3 but is clocked at 900Mhz instead of 1.2GHz for the Pi3. You are free to downclock the Pi3 to any speed you like.

In any case the Pi's lifetime is measured in decades.

The Pi will protect itself by progressively throttling the clock speed back when the temperature (see vcgencmd measure_temp) exceeds 80C. So if you workload is heavy, then adding a heatsink will improve performance.

By default (on Raspbian) the clock speed is controlled by the "on-demand" scaling governor, and when idle runs at 600Mhz.

cryptohunter
Posts: 39
Joined: Wed Apr 05, 2017 5:56 am

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 10:14 am

jahboater wrote:The new Pi2 (v1.2) has the same ARMv8 based SoC as the Pi3 but is clocked at 900Mhz instead of 1.2GHz for the Pi3. You are free to downclock the Pi3 to any speed you like.

In any case the Pi's lifetime is measured in decades.

The Pi will protect itself by progressively throttling the clock speed back when the temperature (see vcgencmd measure_temp) exceeds 80C. So if you workload is heavy, then adding a heatsink will improve performance.

By default (on Raspbian) the clock speed is controlled by the "on-demand" scaling governor, and when idle runs at 600Mhz.
Thank you jahboater :)
Great answer.

Nice suggestion... under-clocking the RP3. Inventive.

I'm interested by your suggestion that the life-span of the Pi might be measured in decades.
RP's are new devices... so there can't be empirical evidence.
Where did you read this idea ? Do you have a link to that ?
Thanks again for your help :)

regards

cryptohunter

jahboater
Posts: 5756
Joined: Wed Feb 04, 2015 6:38 pm
Location: West Dorset

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 11:00 am

cryptohunter wrote: I'm interested by your suggestion that the life-span of the Pi might be measured in decades.
RP's are new devices... so there can't be empirical evidence.
Where did you read this idea ? Do you have a link to that ?
Thanks again for your help :)
Various quotes from RPF engineers over the years.
Originally, for the first Pi, I remember the lifespan being given as 40 years, perhaps reduced to 35 years or so if over clocked. I have no idea how its calculated.

I personally would never down clock a Pi to increase its lifespan!

alphanumeric
Posts: 2546
Joined: Tue Jan 19, 2016 2:17 pm
Location: Sydney, Nova Scotia, Canada

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 11:24 am

My take on it is as follows. If you want maximum performance out of your PI 3B, add a heat sink so it doesn't get hot and down clock. I have heat sinks on almost all of my Pi's. They are inexpensive so I just add one to my order when buying a PI. I don't overclock my 2B's or under clock my 3B's. So far I'm only using passive cooling. I also drilled some cooling holes in my Pi foundation PI 3 case. A heat sinks not much good if the hot air can't get out and cool air in.

User avatar
solar3000
Posts: 1051
Joined: Sat May 18, 2013 12:14 am

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 1:38 pm

jahboater wrote:The new Pi2 (v1.2) has the same ARMv8 based SoC as the Pi3 but is clocked at 900Mhz instead of 1.2GHz for the Pi3. You are free to downclock the Pi3 to any speed you like.

In any case the Pi's lifetime is measured in decades.

The Pi will protect itself by progressively throttling the clock speed back when the temperature (see vcgencmd measure_temp) exceeds 80C. So if you workload is heavy, then adding a heatsink will improve performance.

By default (on Raspbian) the clock speed is controlled by the "on-demand" scaling governor, and when idle runs at 600Mhz.

what new pi2?
Antikythera

User avatar
allfox
Posts: 452
Joined: Sat Jun 22, 2013 1:36 pm
Location: Guang Dong, China

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 1:46 pm

I remove a piece of side wall on the official case to let the hot air out.

User avatar
solar3000
Posts: 1051
Joined: Sat May 18, 2013 12:14 am

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 1:52 pm

allfox wrote:I remove a piece of side wall on the official case to let the hot air out.

Or hang it so that the long ends are up and down. And remove two panels. This creates a [slow] wind tunnel.
Antikythera

cryptohunter
Posts: 39
Joined: Wed Apr 05, 2017 5:56 am

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 3:20 pm

How long before we see mini-fans for the RP3 .... i wonder.


jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 26655
Joined: Sat Jul 30, 2011 7:41 pm

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 7:32 pm

jahboater wrote:
cryptohunter wrote: I'm interested by your suggestion that the life-span of the Pi might be measured in decades.
RP's are new devices... so there can't be empirical evidence.
Where did you read this idea ? Do you have a link to that ?
Thanks again for your help :)
Various quotes from RPF engineers over the years.
Originally, for the first Pi, I remember the lifespan being given as 40 years, perhaps reduced to 35 years or so if over clocked. I have no idea how its calculated.

I personally would never down clock a Pi to increase its lifespan!
The chips are characterised by sticking them in an oven! By upping the temp to well over 100'c, you simulate lifetimes. Not sure of exact details, but I think 150'c is not out of order. The chips still run, but not for as long! That gives you some idea of MTBF at normal operating temperatures.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed.
I've been saying "Mucho" to my Spanish friend a lot more lately. It means a lot to him.

User avatar
solar3000
Posts: 1051
Joined: Sat May 18, 2013 12:14 am

Re: Overheating dangers between RP2 vs RP3

Wed Apr 05, 2017 8:37 pm


Oh. I just don't remember it ever being announced.
Antikythera

cryptohunter
Posts: 39
Joined: Wed Apr 05, 2017 5:56 am

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 10:05 am

Hello people,

It turns out there are lots and lots of videos on youtube where people
do their own experiments, using utilities like sysbench, to figure out how hot the RP3 gets.
The answer is... over 80 degrees in 5 to 10 minutes !

There are videos of people demonstrating that the 'official' heat sinks are inadequate
(because they only effect a tiny reduction in heat)
There are videos showing that with a preposterously oversized heat sink, you can
get really impressive heat dissipation (check out the channel: 'explainingComputers')
for example:
https://www.youtube.com/watch?v=1AYGnw6MwFM
and
https://www.youtube.com/watch?v=WfQMLInuwws

Having seen a number of these videos... and having seen how rapidly, and how
high temperatures soar.... I find it _very_ hard to believe that a RP3 will last decades.
The RP0 or RP2 might last many years... but i think the cat will outlive the RP3.

Time will tell.

Have a nice day

cryptohunter

jahboater
Posts: 5756
Joined: Wed Feb 04, 2015 6:38 pm
Location: West Dorset

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 10:12 am

cryptohunter wrote:It turns out there are lots and lots of videos on youtube where people
do their own experiments, using utilities like sysbench, to figure out how hot the RP3 gets.
The answer is... over 80 degrees in 5 to 10 minutes !
Try this ...

Code: Select all

wget https://raw.githubusercontent.com/ssvb/cpuburn-arm/master/cpuburn-a53.S
gcc -o cpuburn-a53 cpuburn-a53.S
./cpuburn-a53
cryptohunter wrote:There are videos of people demonstrating that the 'official' heat sinks are inadequate
What "official" heat sink? I didn't know there was one yet.
The "official" products in general are well engineered and work properly in my experience.

cryptohunter
Posts: 39
Joined: Wed Apr 05, 2017 5:56 am

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 10:44 am

jahboater wrote:
cryptohunter wrote:It turns out there are lots and lots of videos on youtube where people
do their own experiments, using utilities like sysbench, to figure out how hot the RP3 gets.
The answer is... over 80 degrees in 5 to 10 minutes !
Try this ...

Code: Select all

wget https://raw.githubusercontent.com/ssvb/cpuburn-arm/master/cpuburn-a53.S
gcc -o cpuburn-a53 cpuburn-a53.S
./cpuburn-a53
cryptohunter wrote:There are videos of people demonstrating that the 'official' heat sinks are inadequate
What "official" heat sink? I didn't know there was one yet.
The "official" products in general are well engineered and work properly in my experience.
----------

What does the cpuburn utility do, exactly ?

By official heat sink. I mean the heat sink that is normally sold with the RP3.
The Heat sink is aluminium and tiny, presumably because it has the same
footprint as the chip.

When you say "I didn't know there was one .... yet"
Is that suggesting there will be one soon ?

Maybe Raspberry should create a hugely oversized one, like
the video above from the 'explainingComputers' youtube channel.

jahboater
Posts: 5756
Joined: Wed Feb 04, 2015 6:38 pm
Location: West Dorset

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 10:54 am

cryptohunter wrote:What does the cpuburn utility do, exactly ?
It runs some NEON SIMD instructions on all four cores (I think its doing at least 16 operations per cycle maybe more because NEON is now quad issue). I have found it heats up the CPU quicker than anything else. You can examine the code in cpuburn-a53.S
cryptohunter wrote:When you say "I didn't know there was one .... yet"
Is that suggesting there will be one soon ?
I am guessing the Pi4 will have one (like the Odroid C2). Pure speculation. The Pi3 needs one, but there is no proper mounting, you normally have to rely on a small "stick on" sort of heat sink. My wish list for the Pi4 includes two small holes either side of the SoC so that a proper "northbridge" sized heat sink can be strongly attached (and a flat area around with SoC with low profile components nearby).

cryptohunter
Posts: 39
Joined: Wed Apr 05, 2017 5:56 am

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 11:28 am

jahboater wrote: I am guessing the Pi4 will have one (like the Odroid C2). Pure speculation. The Pi3 needs one, but there is no proper mounting, you normally have to rely on a small "stick on" sort of heat sink. My wish list for the Pi4 includes two small holes either side of the SoC so that a proper "northbridge" sized heat sink can be strongly attached (and a flat area around with SoC with low profile components nearby).
In my humble opinion, when the RP4 eventually does come out, it should come out in two versions,
just like the RP0 and RP0 W. (wireless and non-wireless varieties).

why ?
(1) because wifi is a security risk -- and more and more wifi hacking seems to be occurring.
(2) more and more evidence is accumulating that wifi really isn't good for you,
.. and especially isn't good for children
See here for example: http://www.eurekaselect.com/141391/article

And there was a recent news report of an experiment performed of Danish 9th grade school
kids showing that Wifi signals stunts the growth of seeds.

Not long now before parental concerns push the market towards 'wireless-free' gadgets.
Time will tell.

cryptohunter :)

jahboater
Posts: 5756
Joined: Wed Feb 04, 2015 6:38 pm
Location: West Dorset

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 11:37 am

Or simply provide a way of powering off the Wifi chip from within raspi-config.
Cheaper than two separate products.

I keep my seeds in the garden shed, so I'm OK ...

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 26655
Joined: Sat Jul 30, 2011 7:41 pm

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 12:16 pm

cryptohunter wrote:
jahboater wrote: I am guessing the Pi4 will have one (like the Odroid C2). Pure speculation. The Pi3 needs one, but there is no proper mounting, you normally have to rely on a small "stick on" sort of heat sink. My wish list for the Pi4 includes two small holes either side of the SoC so that a proper "northbridge" sized heat sink can be strongly attached (and a flat area around with SoC with low profile components nearby).
In my humble opinion, when the RP4 eventually does come out, it should come out in two versions,
just like the RP0 and RP0 W. (wireless and non-wireless varieties).

why ?
(1) because wifi is a security risk -- and more and more wifi hacking seems to be occurring.
(2) more and more evidence is accumulating that wifi really isn't good for you,
.. and especially isn't good for children
See here for example: http://www.eurekaselect.com/141391/article

And there was a recent news report of an experiment performed of Danish 9th grade school
kids showing that Wifi signals stunts the growth of seeds.

Not long now before parental concerns push the market towards 'wireless-free' gadgets.
Time will tell.

cryptohunter :)
Please don't regard 9th grade students work, good though it may appear to be, as valid. Especially when the facts of Wireless signals are well known.

Also, please ignore the usual conspiracy theories about Wifi 'radiation' being bad for you. It isn't. You get more ionising radiation from a human body due to the decay of C14 than you do from a wireless router.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed.
I've been saying "Mucho" to my Spanish friend a lot more lately. It means a lot to him.

cryptohunter
Posts: 39
Joined: Wed Apr 05, 2017 5:56 am

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 12:31 pm

jamesh wrote:
cryptohunter wrote:
jahboater wrote: I am guessing the Pi4 will have one (like the Odroid C2). Pure speculation. The Pi3 needs one, but there is no proper mounting, you normally have to rely on a small "stick on" sort of heat sink. My wish list for the Pi4 includes two small holes either side of the SoC so that a proper "northbridge" sized heat sink can be strongly attached (and a flat area around with SoC with low profile components nearby).
In my humble opinion, when the RP4 eventually does come out, it should come out in two versions,
just like the RP0 and RP0 W. (wireless and non-wireless varieties).

why ?
(1) because wifi is a security risk -- and more and more wifi hacking seems to be occurring.
(2) more and more evidence is accumulating that wifi really isn't good for you,
.. and especially isn't good for children
See here for example: http://www.eurekaselect.com/141391/article

And there was a recent news report of an experiment performed of Danish 9th grade school
kids showing that Wifi signals stunts the growth of seeds.

Not long now before parental concerns push the market towards 'wireless-free' gadgets.
Time will tell.

cryptohunter :)
Please don't regard 9th grade students work, good though it may appear to be, as valid. Especially when the facts of Wireless signals are well known.

Also, please ignore the usual conspiracy theories about Wifi 'radiation' being bad for you. It isn't. You get more ionising radiation from a human body due to the decay of C14 than you do from a wireless router.
There is accumulating evidence on the health dangers.
I presume you're completely ignorant of this data.

try googling for 'mobile phones and brain cancer'
-- there are massive statistical spikes across europe and US and litigation has been started.

or google for 'mobile phones and breast cancer'
(There's a fashion amongst young south American women for putting
their phones in their bikini top or bra -- which has resulted in alarming increases in breast cancer
in young women ... amongst age bands with a normally low incidence of breast cancer)

or google for 'sudden infant death syndrome and wireless router'
Don't put your wireless router next to your baby's cot.

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 26655
Joined: Sat Jul 30, 2011 7:41 pm

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 12:49 pm

There is no evidence, so nothing to accumulate. See here for a precis. http://www.wired.co.uk/article/noel-edm ... omagnetism

No further comment on this subject please. There are pseudoscience websites where this sort of post is more appropriate.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed.
I've been saying "Mucho" to my Spanish friend a lot more lately. It means a lot to him.

jahboater
Posts: 5756
Joined: Wed Feb 04, 2015 6:38 pm
Location: West Dorset

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 12:56 pm

jamesh wrote: Also, please ignore the usual conspiracy theories about Wifi 'radiation' being bad for you. It isn't. You get more ionising radiation from a human body due to the decay of C14 than you do from a wireless router.
Its probably the heating effect people (needlessly) worry about.
2.4GHz is close to the 2.45GHz that a domestic microwave oven operates at ;)
(Of course a router is a couple of orders of magnitude less powerful and not focused into a small cavity like a microwave oven).

Martin Frezman
Posts: 1009
Joined: Mon Oct 31, 2016 10:05 am

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 1:02 pm

All you have to do is consider smoking.

Think of how many decades it took to unravel that.
(And then how many decades after that it took to become common knowledge - that is, accepted by the man-on-the-street - which is to say, this has only happened within the last 5-ish years).

Think of how much accepted science said there was nothing to worry about.

That it was even good for you.

I always point to smoking as the classic instance of "just because everybody says there's nothing to worry about doesn't mean there's nothing to worry about".
If this post appears in the wrong forums category, my apologies.

cryptohunter
Posts: 39
Joined: Wed Apr 05, 2017 5:56 am

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 1:06 pm

Martin Frezman wrote:All you have to do is consider smoking.

Think of how many decades it took to unravel that.
(And then how many decades after that it took to become common knowledge - that is, accepted by the man-on-the-street - which is to say, this has only happened within the last 5-ish years).

Think of how much accepted science said there was nothing to worry about.

That it was even good for you.

I always point to smoking as the classic instance of "just because everybody says there's nothing to worry about doesn't mean there's nothing to worry about".
A good point, well made :-)

MarkTF
Posts: 318
Joined: Tue Mar 03, 2015 4:59 pm

Re: Overheating dangers between RP2 vs RP3

Thu Apr 06, 2017 1:47 pm

Martin Frezman wrote:All you have to do is consider smoking.

Think of how many decades it took to unravel that.
(And then how many decades after that it took to become common knowledge - that is, accepted by the man-on-the-street - which is to say, this has only happened within the last 5-ish years).

Think of how much accepted science said there was nothing to worry about.
In the US at least, we've had mandated warnings on cigarettes indicating that they are a known carcinogen for 50 years. The problem is that the public influence of "accepted science" was, for a long time, defeated by directed marketing. There is, to my knowledge, no scientific consensus that non-ionizing EM radiation at common exposure levels is a significant health hazard.

Return to “General discussion”