PhilE
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 2489
Joined: Mon Sep 29, 2014 1:07 pm
Location: Cambridge

Re: Dynamic profile VCore difference between RPi3 & CM3

Mon May 28, 2018 12:33 pm

In that case, let me make it clear that my primary concern is correct functioning - performance may suffer at higher temperatures but there should be no lockups or malfunctions.

John Westlake
Posts: 107
Joined: Thu Nov 09, 2017 4:34 am

Re: Dynamic profile VCore difference between RPi3 & CM3

Mon May 28, 2018 5:17 pm

Could we please have some official comment from RPi about what performance impact with CM3 verse RPI3 we should expect?

At the moment we see the CM3 almost permanently throttled while no thermal issues are encountered with the RPi3 with the same software.

Does RPi have some test suit we can use to evaluate the thermal performance so we can gain a better more contorlled understanding of the thermal operating conditions?

I also still do not understand why a naked RPI3 performs better then CM3 with heatsink - I just dont see the RPI3 larger PCB offering better heat dissipation then a decent heatsink....

I have thermal imaging in the lab, but I'm not sure what this will tell use about the total heat dissipation.....

gsh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 1445
Joined: Sat Sep 10, 2011 11:43 am

Re: Dynamic profile VCore difference between RPi3 & CM3

Mon May 28, 2018 6:04 pm

Are you saying there is no difference in the time it takes to reach 80 degrees with or without the heatsink? That sounds like poor thermal conduction, in comparison to the conduction into the PCB the thermal conduction

Page 13 of this is particularly interesting: Shows only 11% of the heat escapes through the top surface of the die, 87% through the bottom... Until you add a flow of air which still doesn't make a huge difference. Both CM3 and Pi 3 are wire bonded and therefore have no physical connection between the packaging and the top surface of the die suggesting that indeed a heatsink isn't particularly effective at sinking heat!

https://www.idt.com/document/apn/842-th ... -selection
--
Gordon Hollingworth PhD
Raspberry Pi - Director of Software Engineering

gsh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 1445
Joined: Sat Sep 10, 2011 11:43 am

Re: Dynamic profile VCore difference between RPi3 & CM3

Tue May 29, 2018 4:49 pm

https://drive.google.com/file/d/1blhvlt ... sp=sharing

The above link is a presentation showing our analysis of Pi3B and CM3 doing the same thing. The CM3 cannot dissipate as much energy into the air because of the reduced area, the addition of a heatsink doesn't make a huge difference to the dissipation without an active airflow over the heatsink.

Gordon
--
Gordon Hollingworth PhD
Raspberry Pi - Director of Software Engineering

Return to “Compute Module”