jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 23867
Joined: Sat Jul 30, 2011 7:41 pm

Re: Raspberry Pi 4 temperature

Tue Sep 10, 2019 9:22 pm

martinrowan wrote:
Tue Sep 10, 2019 9:15 pm
jamesh wrote:
Tue Sep 10, 2019 8:57 pm
Recent (today) change to bootloader has a tweak to SD RAM timings which will give a slight power saving at idle, which will help very slightly with temperatures.
James is that released, or just internal change being tested?
Thanks
Martin
https://github.com/raspberrypi/rpi-eepr ... e-notes.md
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed. Here's an example...
“I think it’s wrong that only one company makes the game Monopoly.” – Steven Wright

gsgs
Posts: 134
Joined: Sun Dec 13, 2015 6:48 pm
Location: Germany
Contact: Website

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 5:53 am

it should be possible (and easy) to change the temperature at which it throttles.
It could be default 50C or such in one special image offered to schools,classes

bensimmo wrote:
Tue Sep 10, 2019 3:54 pm
if it failed due to thermal stress, then it would be under warranty. why worry.

my only two (and a half) problems are the same from the start of this thread.
with children, schools.

the board is too hot when not cased and in use.

when in a case (e.g RPi case) is over heats and throttles a lot reducing it useful processing power. (not always a problem)

When you want to use the gpio in schools, it needs to be out of the case, you cannot use a fan as that uses the gpio pins and cover the area for Hats and plain gets in yhe way.

it would be really nice if there was a Pi4 Lite.
sort of cool and usable with HATs and out of the case.
the processing power can be lower, as long as it can still run the usual stuff. just without the heat.

If the A72s are not capable of running cool under use, no matter what, then that's a shame.

No doubt that's a small use case as actual use within Education or CodeClubs etc is no where, i am assuming, near the general user/enthusiast and maker community size (ignoring industry).

Which begs the question i don't and haven't seen the answer to.
How slow can the Pi4 go before it becomes unusable?
Webbrowsing and email, office.
Scratch3 with a SenseHAT on.
Media

but not to worry, mines working away fine, two monitors now and is useful as a desktop with a fan gently blowing air over.
and i have my pi3 series for accesible gpio work.

Can't wait to see what they come up with next :-)

User avatar
Imperf3kt
Posts: 2862
Joined: Tue Jun 20, 2017 12:16 am
Location: Australia

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 7:06 am

gsgs wrote:
Wed Sep 11, 2019 5:53 am
it should be possible (and easy) to change the temperature at which it throttles.
It could be default 50C or such in one special image offered to schools,classes
This was addressed - its entirely pointless to run it at a cool enough temperature if as a trade off you lose ~75% of the processing power.
And throttling won't reduce the temperature that much. While idle at a desktop (600MHz - the same clock speed as when throttled), my Pi4b exceeds 60°C, when under reasonably heavy load and running at stock 1.5GHz, it reaches 68°C, a difference of only 8°C - not anywhere near enough to bring it below 50°C and allow it out of throttle mode.
55:55:44:44:4C
52:4C:52:42:41

ProDigit
Posts: 374
Joined: Tue Aug 30, 2011 1:24 am

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 7:29 am

I think people are disappointed at the prospect of using either a large heat sink, or active cooling.
The active cooling mainly as idea about the impact on the environment, not taking in consideration that a huge heatsink might actually be worse to the environment than a small fan,

But also they're disappointed because any rotating device makes noise, can potentially break down, and things (finger, wires, dust) can get stuck in there.

If the pi foundation will come out in the future with a pi 5, they'll have to find a way to reduce the heat, while improving the performance, and the only way this will happen, is by getting a chip on a much smaller die (10nm); and If possible, with built in ram and GPU, to get a larger footprint for cooling the CPU. (Without it, the CPU would be too small to mount a fan or heatsink on, and cooling the CPU evenly will be harder).

I'm only speculating here.

fruitoftheloom
Posts: 20900
Joined: Tue Mar 25, 2014 12:40 pm
Location: Delightful Dorset

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 7:35 am

ProDigit wrote:
Wed Sep 11, 2019 7:29 am
I think people are disappointed at the prospect of using either a large heat sink, or active cooling.
The active cooling mainly as idea about the impact on the environment, not taking in consideration that a huge heatsink might actually be worse to the environment than a small fan, in

But also they're disappointed because any rotating device makes noise, can potentially break down, and things (finger, wires, dust) can get stuck in there.

If the pi foundation will come out in the future with a pi 5, they'll have to find a way to reduce the heat, while improving the performance, and the only way this will happen, is by getting a chip on a much smaller die (10nm); and If possible, with built in ram and GPU, to get a larger footprint for cooling the CPU. (Without it, the CPU would be too small to mount a fan or heatsink on, and cooling the CPU evenly will be harder).

I'm only speculating here.

I am 100% sure RPT / RPF are aware of the situation, suggesting otherwise is disrespectful.
Retired disgracefully.....

User avatar
PeterO
Posts: 5066
Joined: Sun Jul 22, 2012 4:14 pm

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 7:47 am

fruitoftheloom wrote:
Wed Sep 11, 2019 7:35 am
I am 100% sure RPT / RPF are aware of the situation, suggesting otherwise is disrespectful.
I wonder if he has the same issues with the Intel/AMD CPUs in every Desktop PC ? Or is it a case of "out of sight , out of mind " ?
PeterO
Discoverer of the PI2 XENON DEATH FLASH!
Interests: C,Python,PIC,Electronics,Ham Radio (G0DZB),1960s British Computers.
"The primary requirement (as we've always seen in your examples) is that the code is readable. " Dougie Lawson

ProDigit
Posts: 374
Joined: Tue Aug 30, 2011 1:24 am

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 7:52 am

PeterO wrote:
Wed Sep 11, 2019 7:47 am
fruitoftheloom wrote:
Wed Sep 11, 2019 7:35 am
I am 100% sure RPT / RPF are aware of the situation, suggesting otherwise is disrespectful.
I wonder if he has the same issues with the Intel/AMD CPUs in every Desktop PC ? Or is it a case of "out of sight , out of mind " ?
PeterO
I wonder why you need to respond to your alter ego profile on this forum all the time?


And no, suggestions are NOT disrespectful.
Being disrespectful to members is.

User avatar
PeterO
Posts: 5066
Joined: Sun Jul 22, 2012 4:14 pm

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 8:03 am

ProDigit wrote:
Wed Sep 11, 2019 7:52 am
I wonder why you need to respond to your alter ego profile on this forum all the time?
Do you know what "alter ego" means ?
Here, I'll google it for you...
a person's secondary or alternative personality.
an intimate and trusted friend.
I doubt that's what you mean ! :lol:
PeterO
Discoverer of the PI2 XENON DEATH FLASH!
Interests: C,Python,PIC,Electronics,Ham Radio (G0DZB),1960s British Computers.
"The primary requirement (as we've always seen in your examples) is that the code is readable. " Dougie Lawson

fruitoftheloom
Posts: 20900
Joined: Tue Mar 25, 2014 12:40 pm
Location: Delightful Dorset

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 8:22 am

PeterO wrote:
Wed Sep 11, 2019 8:03 am
ProDigit wrote:
Wed Sep 11, 2019 7:52 am
I wonder why you need to respond to your alter ego profile on this forum all the time?
Do you know what "alter ego" means ?
Here, I'll google it for you...
a person's secondary or alternative personality.
an intimate and trusted friend.
I doubt that's what you mean ! :lol:
PeterO

ROFL :lol:


ProDigit you have a bad fud attitude towards the Raspberry Pi and still surprised you keep posting.


As for me my Asus ChromeBit is not ready to be retired and replaced by a RPi 4B, though despite all the "perceived" issues would by one tomorrow if need arose.....
Retired disgracefully.....

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 23867
Joined: Sat Jul 30, 2011 7:41 pm

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 8:55 am

ProDigit wrote:
Wed Sep 11, 2019 7:29 am
I think people are disappointed at the prospect of using either a large heat sink, or active cooling.
The active cooling mainly as idea about the impact on the environment, not taking in consideration that a huge heatsink might actually be worse to the environment than a small fan,

But also they're disappointed because any rotating device makes noise, can potentially break down, and things (finger, wires, dust) can get stuck in there.

If the pi foundation will come out in the future with a pi 5, they'll have to find a way to reduce the heat, while improving the performance, and the only way this will happen, is by getting a chip on a much smaller die (10nm); and If possible, with built in ram and GPU, to get a larger footprint for cooling the CPU. (Without it, the CPU would be too small to mount a fan or heatsink on, and cooling the CPU evenly will be harder).

I'm only speculating here.
Not sure why people think that you can get more speed out of a CPU but still use less power. There are gains to be made in process node changes, but there are physical limits that generally mean that more CPU power = more heat. For example, the CPU performance per die area is about the same for ARM and x86. ARM is generally less speedy, so uses less power for the same die area - so if ARM gave the same performance as x86, it would use about the same die area. So the only improvement to get use less power is to throw features away, or to move to a smaller process node. BUT, there are disadvantages with that as well, which means it's not a linear thing (leakage etc). And generally, as you move to a smaller process node, you actually add more features (no one wants a new Pi to JUST have lower power consumption - they want more features!), so actually, the power improvements of moving to a newer node are negated by providing more features.

And of course the 10nm (or anything in that area) process is incredibly expensive compared with the 28 we use now, so you would certainly expect much higher cost Pi's.

Fans are the way to go IMO. Keep on a sane process, just provide cooling. Small fans are quiet, reliable, cheap. Case style heatsinks are even quieter, cheap and reliable, but get hot.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed. Here's an example...
“I think it’s wrong that only one company makes the game Monopoly.” – Steven Wright

martinrowan
Posts: 71
Joined: Wed Aug 26, 2015 11:00 am
Location: Berkshire, UK
Contact: Website Twitter

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 9:43 am

jamesh wrote:
Wed Sep 11, 2019 8:55 am
Fans are the way to go IMO. Keep on a sane process, just provide cooling. Small fans are quiet, reliable, cheap. Case style heatsinks are even quieter, cheap and reliable, but get hot.
James I know people are coming at this from different angles. I agree with what you are saying, but the issue for me at least is that the Raspberry Pi Official Pi 4 case doesn't have a fan, or any ventilation and thus isn't really suited to the current thermal load of the RPi4 as it just causes it to throttle even when idle. I honestly dont' think it should be being sold as a suitable case. I hope the software tweaks to reduce power and thus heat can make the case usable again.

User avatar
Gavinmc42
Posts: 3874
Joined: Wed Aug 28, 2013 3:31 am

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 10:11 am

honestly dont' think it should be being sold as a suitable case
How about as a suitable "carry" case?
Do you know anything better?

I pop the top back on, unplug and throw it in my carry bag.
Take it out, plug it in and pop the top to let air circulate around the heat sink.
It takes longer to plug the cables in than popping the top.
I forgot the other day it nearly melted when it got to 70C? Nope never noticed a thing.

Perhaps RPF need an official fan so everyone can say "told you so"?
Can we have one with LEDs?
I'm dancing on Rainbows.
Raspberries are not Apples or Oranges

User avatar
davidcoton
Posts: 4179
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 10:15 am

Gavinmc42 wrote:
Wed Sep 11, 2019 10:11 am
Perhaps RPF need an official fan so everyone can say "told you so"?
Can we have one with LEDs?
Only if they are pink LEDS in the shape of a pony. :lol:
Signature retired

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 23867
Joined: Sat Jul 30, 2011 7:41 pm

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 10:55 am

martinrowan wrote:
Wed Sep 11, 2019 9:43 am
jamesh wrote:
Wed Sep 11, 2019 8:55 am
Fans are the way to go IMO. Keep on a sane process, just provide cooling. Small fans are quiet, reliable, cheap. Case style heatsinks are even quieter, cheap and reliable, but get hot.
James I know people are coming at this from different angles. I agree with what you are saying, but the issue for me at least is that the Raspberry Pi Official Pi 4 case doesn't have a fan, or any ventilation and thus isn't really suited to the current thermal load of the RPi4 as it just causes it to throttle even when idle. I honestly dont' think it should be being sold as a suitable case. I hope the software tweaks to reduce power and thus heat can make the case usable again.
I cannot comment on the case, as I don't use it. If you don't like the heat build up, I'd suggest not using it too.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed. Here's an example...
“I think it’s wrong that only one company makes the game Monopoly.” – Steven Wright

User avatar
bensimmo
Posts: 4182
Joined: Sun Dec 28, 2014 3:02 pm
Location: East Yorkshire

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 4:11 pm

Fans and thing are the way to go, it's what I use. It doesn't take much to keep it happy.
(probably repeating myself ;-) now)

But Fans and HATs * don't * go.
The Pi4 is not edit overly /edit HAT friendly as heat build up under it of course.
It's is very much a Basic Desktop Computer friendly though.


Though... I will test the SenseHAT and other full or half size HATs and see what happens with the latest updates. Of course with no airflow as I would expect that setup to be used in clubs and school.
No case either.
...

Did the SenseHAT always had the blinding full power white flash of light at power on, just blinded myself.
(Before the rainbow).

I take it back, things have improved a lot.
Update as of yesterday and vl805ab
Simple SenseHAT datalogging rapidly and flashing an led on it when it reads.
Using Thonny, browser and LibreCalc

It's reaching just over 72C with recovery as needed. (The room not cold).

*Lots of improvements have been made and I take most of it back.*

I did pick it up to play with the giro etc and I would need to advise caution, the backs
blooming hot as is the uSD card slot.

You can watch instructional videos, more than that and it tries to self destruct.. not really but it does rapidly move to a thermometer icon.
Can't play any more than 720p YouTube, which is fine, it does keep it smooth until it reaches 84C which was mostly enough for a short video.
86C was the max reached.
It then just sits there playing a slightly jerky video, no buffering like 1080p

As a side, the SenseHAT temperature (onboard) started at 40C and only ramped to 45C staying at that.
Humidity from about 50% dropped to 37%

Still quite usable at 85C.
Just keep it away from users ;-)
It doesn't cool down very much or fast once video has stopped.
Probably as expected as the hot air is trapped.

Well done that is much better than day one :-)

(Pi4 4GB, two 1080 monitors and a Logitech k270 wireless dongle SanDisk A1 card,WiFi 5G @~24Mbps download on Googles speedtest ).

Just a typical school or club usage there. Ok maybe not dual monitors but it does make programming easier

ProDigit
Posts: 374
Joined: Tue Aug 30, 2011 1:24 am

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 8:44 pm

jamesh wrote:
Wed Sep 11, 2019 8:55 am
ProDigit wrote:
Wed Sep 11, 2019 7:29 am
I think people are disappointed at the prospect of using either a large heat sink, or active cooling.
The active cooling mainly as idea about the impact on the environment, not taking in consideration that a huge heatsink might actually be worse to the environment than a small fan,

But also they're disappointed because any rotating device makes noise, can potentially break down, and things (finger, wires, dust) can get stuck in there.

If the pi foundation will come out in the future with a pi 5, they'll have to find a way to reduce the heat, while improving the performance, and the only way this will happen, is by getting a chip on a much smaller die (10nm); and If possible, with built in ram and GPU, to get a larger footprint for cooling the CPU. (Without it, the CPU would be too small to mount a fan or heatsink on, and cooling the CPU evenly will be harder).

I'm only speculating here.
Not sure why people think that you can get more speed out of a CPU but still use less power. There are gains to be made in process node changes, but there are physical limits that generally mean that more CPU power = more heat. For example, the CPU performance per die area is about the same for ARM and x86. ARM is generally less speedy, so uses less power for the same die area - so if ARM gave the same performance as x86, it would use about the same die area. So the only improvement to get use less power is to throw features away, or to move to a smaller process node. BUT, there are disadvantages with that as well, which means it's not a linear thing (leakage etc). And generally, as you move to a smaller process node, you actually add more features (no one wants a new Pi to JUST have lower power consumption - they want more features!), so actually, the power improvements of moving to a newer node are negated by providing more features.

And of course the 10nm (or anything in that area) process is incredibly expensive compared with the 28 we use now, so you would certainly expect much higher cost Pi's.

Fans are the way to go IMO. Keep on a sane process, just provide cooling. Small fans are quiet, reliable, cheap. Case style heatsinks are even quieter, cheap and reliable, but get hot.
Well, in my opinion, if the pi 4 successor will be better, faster, it'll have to be either the same price, or more expensive.

Only the same price if prices of smaller node CPUs drop in price. 10nm is currently extremely expensive indeed, and 14nm would make more sense. But who knows what it'll be 2-3years down the road, when a new pi will be ready to hit the market?

The zero on the other hand, can improve by only a small performance boost, while having a potentially large cpu power consumption decrease. It won't need a lot more for a next model to be financially viable (at $25?).
That's a meal for 2 in a fastfood place.

I would think the zero wouldn't need a lot of improvements, other than the CPU (add an extra core and +100Mhz boost) and ram (add 256-512MB), which is the current bottleneck of the system, as well as a better wifi connection (from pi 3B+).
Add some cost there, remove some from decreasing the board size to a smaller form factor, and perhaps a few less gpio pins, and mostly unused options like rca, reset pins, etc...


If they can be built Qualcomm quad/octa core chips on a 10-14nm node into $100 cellphones, then it can be built into the next $100 pi.

To generate the same heat while dropping from 28nm to 14nm, you'd have to have a doubling in cpu cores, as well as a frequency boost, as a similar newer node CPU will be 4x smaller in size, and about 66-75% more efficient.

Edit: the situation of heat build up with the original case (V2) perhaps a modification can be made to the design, to fit a standard 10mm fan (2 screw mounting holes), few holes for circulation on top and side of the case.
A redesign of the official pi 4 case will solve the issue.
A 10mm, 5v fan is common and cheap to get, and will allow for a small heat sink, and eliminate thermal throttling under most situation, as well as run <50c idle.

User avatar
davidcoton
Posts: 4179
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 10:19 pm

ProDigit wrote:
Wed Sep 11, 2019 8:44 pm
The zero on the other hand, can improve by only a small performance boost, while having a potentially large cpu power consumption decrease. It won't need a lot more for a next model to be financially viable (at $25?).
...
If they can be built Qualcomm quad/octa core chips on a 10-14nm node into $100 cellphones, then it can be built into the next $100 pi.
I think the next Pi should include a live Pink Pony. It might have to sell at £1000, but that will be financially viable. :o :shock: :? :lol: :roll: :twisted:

Until there's a world shortage of Pink Pony Food. :ugeek:
Signature retired

ProDigit
Posts: 374
Joined: Tue Aug 30, 2011 1:24 am

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 10:28 pm

davidcoton wrote:
Wed Sep 11, 2019 10:19 pm
ProDigit wrote:
Wed Sep 11, 2019 8:44 pm
The zero on the other hand, can improve by only a small performance boost, while having a potentially large cpu power consumption decrease. It won't need a lot more for a next model to be financially viable (at $25?).
...
If they can be built Qualcomm quad/octa core chips on a 10-14nm node into $100 cellphones, then it can be built into the next $100 pi.
I think the next Pi should include a live Pink Pony. It might have to sell at £1000, but that will be financially viable. :o :shock: :? :lol: :roll: :twisted:

Until there's a world shortage of Pink Pony Food. :ugeek:
You cut and edited my response.
The $100 would be viable for a pi 4B successor. Not an updated zero

Gerbertus
Posts: 12
Joined: Tue Apr 17, 2018 3:39 pm

Re: Raspberry Pi 4 temperature

Wed Sep 11, 2019 10:49 pm

Most people using ARM cores this recent are using smaller processes. If things were things going a little more normally in the fab space I think it would have been right for RPi to jump all the way down to 14nm. Of course the fiascos people have had with sub-14nm processes mean that 14nm is still in heavy demand and relatively expensive even though it's 5 years old, so using 28nm is understandable for now. I just hope you don't stick with 28nm for another seven years!

I don't think pushing the performance envelope need be the overriding priority for that next process jump. It was more important to get up to a 'good enough' level and narrow the gap between the RPi and other low-end computing gear.

But even if a future 14/10/etc nm Pi manages to substantially reduce power draw compared to the Pi4, it may not be easy to avoid the need for better cooling, because thermal density may not go down. Smaller chip and proportionally smaller power draw -> same thermal density.
jamesh wrote:
Wed Sep 11, 2019 8:55 am
Fans are the way to go IMO. Keep on a sane process, just provide cooling. Small fans are quiet, reliable, cheap. Case style heatsinks are even quieter, cheap and reliable, but get hot.
This has never been my experience. Large fans are quiet and reliable but small fans are whiny at best and then start developing bearing problems or whatnot and get noisier and noisier till they fail entirely. I'm glad the FLIRC seems to solve the problem well enough for the Pi 4, and it's not getting so hot as to be a deal breaker. I hope that the RPi foundation takes a cue from that case for future official cases; the present official Pi4 case is kind of disastrous.

User avatar
davidcoton
Posts: 4179
Joined: Mon Sep 01, 2014 2:37 pm
Location: Cambridge, UK

Re: Raspberry Pi 4 temperature

Thu Sep 12, 2019 12:00 am

ProDigit wrote:
Wed Sep 11, 2019 10:28 pm
You cut and edited my response.
The $100 would be viable for a pi 4B successor. Not an updated zero
I clearly left your indicated price for an updated Zero. I did not misrepresent your message in my edit.

A Zero at $25 is not viable. That's Pi3A territory now. And a price increase destroys the USP of the Zero.
BTW, have you any idea how much a new SOC design costs?

A Pi5 at $100 is not viable. My crystal ball says a base model will be $35, with options at (some of) $45, $55, $65 or even $75, but not $100.

If your suggested prices were viable, don't you think RPT would have filled the gap with Pi4 models?
Do other makes of SBC sell like hot Pi(e)s at $100?
I invite you to prove me wrong -- tell us when your Kickstarter is ready to launch.
Signature retired

ProDigit
Posts: 374
Joined: Tue Aug 30, 2011 1:24 am

Re: Raspberry Pi 4 temperature

Thu Sep 12, 2019 1:04 am

When you look at the sales of the 4B, you can clearly see there's a market for a pi desktop replacement, and for many, the 4B isn't that one yet.
It would make sense to continue the trend of more power, more options, less heat, to satisfy that market.

But if you ask me,
I think the pi is popular because of low power consumption, not because of exceptional performance.
I think devices like the pi 1, 2 and zero became popular (taking the GPIO out of the equation) in part because they were cheap, in part because it's a new thing, ARM (not Intel), and in part because they were more energy efficient than x86 alternatives.

No one bought Pi's because of their performance, but a lot bought it for their performance per watt.

As far as a smaller die goes, smaller dies allow for a reduction in power consumption, as well as in heat.
Heat is directly proportional to power consumption.
So when power consumption goes down, so does heat.
In order to get a better heat dissipation with a die shrink, it's contact area needs to increase with the heat sink, which is why a smaller die CPU, is best paired with more cores.
This can also be done by combining CPU, GPU, and potentially RAM on one socket.
The CPU can then dissipate heat to the GPU and ram, which have a much lower TDP and heat generation, and which in turn dissipate it to the heat sink.
It would actually benefit arm to put 128-256MB of Vram right on the CPU/GPU block, so the chipset no longer needs to use regular ram, and has more direct Access to the memory.
But even if not, a smaller size CPU can be doubled or quadrupled, to increase surface size with the cooling solution. Meaning a single core at load will dissipate the heat through the other cores, and heat might only become an issue if all 4 cores on the smaller (14nm) die are operating at full load; thus reaching slightly higher temperatures as a 28nm single core CPU.

User avatar
jcyr
Posts: 418
Joined: Sun Apr 23, 2017 1:31 pm
Location: Atlanta

Re: Raspberry Pi 4 temperature

Thu Sep 12, 2019 1:57 am

ProDigit wrote:
Thu Sep 12, 2019 1:04 am
When you look at the sales of the 4B, you can clearly see there's a market for a pi desktop replacement, and for many, the 4B isn't that one yet.
Nor will it ever be (a desktop replacement) as long as it sticks to the Linux desktop. Linux has taken over cloud services and servers more generally, but for mass adoption on the desktop... not so much! Though usable, the Linux desktop presentation is kind of crude compared to what the masses are expecting from a Windows or Mac PC. Also, many of the tools used in various professions are not available for Linux, and when they are it's usually X86 Linux only.

It's not about the hardware, it's all about the software.

As to what drives Pi sales, I'm pretty sure it's pricing, not whether is runs an ARM, or an Atom core.
Last edited by jcyr on Thu Sep 12, 2019 1:36 pm, edited 1 time in total.
It's um...uh...well it's kinda like...and it's got a bit of...

User avatar
Gavinmc42
Posts: 3874
Joined: Wed Aug 28, 2013 3:31 am

Re: Raspberry Pi 4 temperature

Thu Sep 12, 2019 6:13 am

No one bought Pi's because of their performance,
I did, faster, cheaper and much easy to use than Microchip Arduino mega PCB's with Ethernet hat.
If I had known how much trouble Linux would cause me, I might have chosen differently :lol:
As to what drives Pi sales,
Cost, Qualitiy, Performance.
It is very good at meeting all three requirements.
https://en.wikipedia.org/wiki/Project_m ... t_triangle

For the extra performance of the PI4 we get some heat, big deal NOT.
I'm dancing on Rainbows.
Raspberries are not Apples or Oranges

gkreidl
Posts: 6108
Joined: Thu Jan 26, 2012 1:07 pm
Location: Germany

Re: Raspberry Pi 4 temperature

Thu Sep 12, 2019 6:32 am

@ProDigit: Your wildly speculative nonsense is completely off-topic in this thread. A mod might decide to close it because of this, although the thread topic is quite important.
You told us, that you don't want to buy a RPi 4. That's your business, but then you should keep your mouth shut about the real RPi 4 world.
Minimal Kiosk Browser (kweb)
Slim, fast webkit browser with support for audio+video+playlists+youtube+pdf+download
Optional fullscreen kiosk mode and command interface for embedded applications
Includes omxplayerGUI, an X front end for omxplayer

jdb
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 2115
Joined: Thu Jul 11, 2013 2:37 pm

Re: Raspberry Pi 4 temperature

Thu Sep 12, 2019 6:46 am

ProDigit wrote:
Thu Sep 12, 2019 1:04 am
When you look at the sales of the 4B, you can clearly see there's a market for a pi desktop replacement, and for many, the 4B isn't that one yet.
It would make sense to continue the trend of more power, more options, less heat, to satisfy that market.

But if you ask me,
I think the pi is popular because of low power consumption, not because of exceptional performance.
I think devices like the pi 1, 2 and zero became popular (taking the GPIO out of the equation) in part because they were cheap, in part because it's a new thing, ARM (not Intel), and in part because they were more energy efficient than x86 alternatives.

No one bought Pi's because of their performance, but a lot bought it for their performance per watt.

As far as a smaller die goes, smaller dies allow for a reduction in power consumption, as well as in heat.
Heat is directly proportional to power consumption.
So when power consumption goes down, so does heat.
In order to get a better heat dissipation with a die shrink, it's contact area needs to increase with the heat sink, which is why a smaller die CPU, is best paired with more cores.
This can also be done by combining CPU, GPU, and potentially RAM on one socket.
The CPU can then dissipate heat to the GPU and ram, which have a much lower TDP and heat generation, and which in turn dissipate it to the heat sink.
It would actually benefit arm to put 128-256MB of Vram right on the CPU/GPU block, so the chipset no longer needs to use regular ram, and has more direct Access to the memory.
But even if not, a smaller size CPU can be doubled or quadrupled, to increase surface size with the cooling solution. Meaning a single core at load will dissipate the heat through the other cores, and heat might only become an issue if all 4 cores on the smaller (14nm) die are operating at full load; thus reaching slightly higher temperatures as a 28nm single core CPU.
This post is a perfect example of the adage "a little knowledge is a dangerous thing".
I suggest people ignore it.

Prodigit, stop posting half-formed opinions as fact.
Rockets are loud.
https://astro-pi.org

Return to “Troubleshooting”