Intel Joule Developer Kit

Transcription

1 Intel Joule Developer Kit Thermal Management Overview Revision 1.0 November 2016

2 Disclaimers: INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL PRODUCTS. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN INTEL S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY WHATSOEVER AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO SALE AND/OR USE OF INTEL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. A "Mission Critical Application" is any application in which failure of the Intel Product could result, directly or indirectly, in personal injury or death. SHOULD YOU PURCHASE OR USE INTEL S PRODUCTS FOR ANY SUCH MISSION CRITICAL APPLICATION, YOU SHALL INDEMNIFY AND HOLD INTEL AND ITS SUBSIDIARIES, SUBCONTRACTORS AND AFFILIATES, AND THE DIRECTORS, OFFICERS, AND EMPLOYEES OF EACH, HARMLESS AGAINST ALL CLAIMS COSTS, DAMAGES, AND EXPENSES AND REASONABLE ATTORNEYS' FEES ARISING OUT OF, DIRECTLY OR INDIRECTLY, ANY CLAIM OF PRODUCT LIABILITY, PERSONAL INJURY, OR DEATH ARISING IN ANY WAY OUT OF SUCH MISSION CRITICAL APPLICATION, WHETHER OR NOT INTEL OR ITS SUBCONTRACTOR WAS NEGLIGENT IN THE DESIGN, MANUFACTURE, OR WARNING OF THE INTEL PRODUCT OR ANY OF ITS PARTS. Intel may make changes to specifications and product descriptions at any time, without notice. Designers must not rely on the absence or characteristics of any features or instructions marked reserved or undefined. Intel reserves these for future definition and shall have no responsibility whatsoever for conflicts or incompatibilities arising from future changes to them. The information here is subject to change without notice. Do not finalize a design with this information. The products described in this document may contain design defects or errors known as errata which may cause the product to deviate from published specifications. Current characterized errata are available on request. Contact your local Intel sales office or your distributor to obtain the latest specifications and before placing your product order. Copies of documents which have an order number and are referenced in this document, or other Intel literature may be obtained by calling or by visiting Intel s website at Intel processor numbers are not a measure of performance. Processor numbers differentiate features within each processor family, not across different processor families. See for details. Intel, the Intel logo, Intel Joule Module and Intel Atom are trademarks of Intel Corporation in the United States and other countries. * Other brands and names may be claimed as the property of others. Copyright 2016 Intel Corporation. All rights reserved. 2

3 Contents: 1. Thermal Management Objectives 2. Hardware Control and Firmware Settings 3. Workload Measurement and Setting Power Limit in BIOS 3

4 Section 1 Thermal Management Objectives

5 Overview Intent of thermal management The Intel Joule compute module integrates a powerful four-core, 64-bit Intel Atom processor, high definition graphics, and wireless communications into a compact system on module (SoM) form factor. The module generates thermal energy relative to workload; the goal of thermal management is to maintain balance between user safety, system performance, and platform capacity for heat dissipation. Successful thermal management techniques: Provide user safety through thermal ergonomics that control exposed surface temperatures. Prevent thermal runaway (uncontrolled overheating) that can result in platform instability. Thermal management solutions will vary significantly per use case: Device workloads and software decisions can have the greatest influence on thermal generation. Operating environments and enclosure designs are a critical part of thermal calculations. 5

6 Overview Thermal management basics The module provides two (BIOS) methods that work together to maintain a thermal target. It is the responsibility of the developer and validation team to ensure each is properly configured and enabled. Temperature Limit Setting: Programmable temperature when throttling to begins; module shutdown will happen at 105 C. Can be set lower than the programmed maximum of (75 C) to accommodate for safety ergonomics. Run-time Average Power Limit (RAPL): Module firmware can automatically adjust processor power to maintain temperature targets. The fixed maximum power limit helps avoids issues that can arise from rapid thermal transients. 6

7 Overview Thermal control methods There and numerous ways to remove thermal energy (heat) from the module. This overview outlines the reference heatsink and cooling fan assembly mounted to the Intel Joule module expansion board. Passive dissipation: Workloads with low heat generation can radiate thermal energy into their environment. Enclosures can be designed to encourage thermal convection (natural airflow). Active control: A powered fan can provide constant airflow across the thermal plate or heatsink. Module power can be varied to maintain module within an ideal thermal range. 7

8 Section 2 Hardware Control and Firmware Settings

9 Hardware Control Introduction The module provides hardware thermal controls that are intended to maintain operation within a set temperature range by adjusting processor speed or capping the maximum power consumption. These controls are configured within the BIOS; they operate independently of any operating system or software based thermal control loops and can influence system performance and power consumption. Thermal Control Circuit (TCC): Specified maximum temperature limit at which processor power will be automatically reduced. TCC can be set below the maximum default value, but not above it. The amount and rate of processor speed adjustment is hardcoded in the module and cannot be changed. Power Limit 1 (PL1): Processor Run-time Averaged Power Limit (RAPL). PL1 is a static limit of the maximum module power while TCC is below set point. When module temperature is above the TCC set point, the TCC function dynamically lowers PL1 until the TCC value returns below the set point. 9

10 Hardware Control Enter the module BIOS system Press the [F2] key during boot up splash to enter the BIOS system. Select Device Manager using arrow keys depress the [ENTER] key to reach the next menu level. Select System Setup and depress the [ENTER] key to reach the next menu level. 10

11 Hardware Control Select CPU Configuration menu and enable the TM1 and DTS sensors The TM1 and DTS (module thermal sensors) must be enabled for the TCC to function. Next select the CPU Power Management sub-menu. 11

12 Hardware Control Set PL1 attributes as needed for design intent, environment, and workload Power Limit 1 attributes: Power Limit 1 Function [Enable / Disable] Disable will allow module to operate without input power limit and prevents software thermal controls that utilize the PL1 function. Power Limit 1 Clamp Mode [Enable / Disable] Disable will prevent setting clock below base frequency for any domain. Enable allows firmware to set clock below base to meet thermal target. It is recommended that Clamp Mode always be enabled. Power Limit 1 Power [#] Range from 1 to 10 in whole numbers 4 is recommended max with reference heatsink and no fan. 5 and above should have active cooling solutions. 8 is only recommended for fully profiled and validated designs. Use the Go Back option to navigate and [F4] to commit changes. Power Limit 1 Time Window <#> Smaller time window values increase power sampling and response rate. Some applications will need a higher value if they have infrequent bursts of workload that are best if not quickly power limited. 12

13 Hardware Control Overview of the Thermal Configuration Parameters Hardware Control Parameters: Thermal Control Offset [#] Offset is specified relative to 88 C (Tj value). For example, Thermal Control Offset of 10 will result in TCC activation at (Tj) 88 C 10 C offset = 78 C TCC trip. Automatic Thermal Reporting [Enable / Disable] Enable will permit the BIOS to report ACPI trip points and active cooling gag to ACPI enabled operating systems. NOTE: The critical and passive trip points are for software packages that support ACPI functions: Critical Trip Point <### C> Temperature where module will perform an automatic (hard) shutdown without alerting the operating system. Passive Trip Point <### C> Temperature where the module will assert the TBD signal to inform operating system or other control structure that thermal management needs to activated before the module initiates a hard shutdown. 13

14 Section 3 Workload Measurement and Setting Power Limit in BIOS

15 Module Power Characterization Introduction to method The following elements will facilitate collecting power data that can be used to determine the cooling capacity required to maintain temperatures within a target range for the application workload: Active heatsink properly attached to the module. BIOS settings configured for maximum system performance. Stable and robust platform power supply that exceeds workload needs. All accessory devices connected and operational for the defined used case. The following slides will outline how to configure the platform, and collect the power measurement data that is used to determine the best BIOS power limit settling for the profiled application. 15

16 Module Power Characterization Configure BIOS to allow maximum module performance In the CPU settings / System Power Options menu: Enable Intel Turbo Boost Technology Configure Power Limit options as shown: In the Thermal menu: Set Thermal Control Offset to [0] Disable Automatic Thermal Reporting Set Critical and Passive trip points to <103 C> Use the Go Back option and depress [F4] to commit changes to BIOS 16

17 Module Power Characterization Theory of measuring processor power Processor power will be measured by accessing the RAPL driver s energy counter and recording periodic results to a log file for later analysis. The energy counter counts the amount of energy spent by the processor in micro ( ) joules. Therefore, processor power can be calculated by logging two instances of the energy counter and then dividing the difference by the measurement interval. PPPPPPPPPP = EEEEEEEEEEEE TTT EEEEEEEEEEEE TTT TTT TTT RAPL driver s energy counter can be accessed at /sys/class/powercap/intel-rapl/intel-rapl:0/energy_uj and an example script to collect data is shown on the following page. 17

18 Module Power Characterization Using a shell script to capture power data 1. Stop executing Thermal Daemon to halt operating system based thermal controls. sh-4.3# ps ~~~~~~~~~~~~~~~~ 580 root S /usr/sbin/connmand n 586 root 123m S /usr/sbin/thermald --no-daemon --dbus-enable --config-file 590 root S STUBINIT 601 root S /usr/sbin/wpa_supplicant u ~~~~~~~~~~~~~~~~ sh-4.3# kill Power measurement script example to collect power, in watts, every two seconds (adjust as appropriate). while true do P1=$(cat /sys/class/powercap/intel-rapl/intel-rapl\:0/energy_uj) T1=$(date +%s) sleep 2 P2=$(cat /sys/class/powercap/intel-rapl/intel-rapl\:0/energy_uj) T2=$(date +%s) PWR=$(echo scale=2;($p2-$p1)/($t2-$t1)/ bc) echo $PWR done 3. Run your application and the power measurement script simultaneously to collect workload data. Note: This script measures time averaged power, which is useful for thermal assessment. This method will not measure instantaneous spikes that are needed for power delivery validation. 18

19 Module Power Characterization Reading the log to determine peak power consumption After running the target application through all use cases and long enough to collect sufficient data, stop the logging script activity and open the log file. The long file contains Watts used expressed as integers. Module power limit (PL1) should be set at the next whole number above maximum power recorded. log data: The log files shows a peak power consumption of 2.50 Watts. PL1 is set in whole number increments, setting PL1 to 3 provides reasonable overhead capacity and device protection. 19

20 Heatsink Selection for Example Application This chart shows the ambient temperature (Y-axis) range for a known workload (X-axis) when using either the passive reference heatsink or the reference heatsink combined with a forced air fan. Our example 2.5 Watt application can run with the reference heatsink, as long as the ambient temperature remains below 30 C and above 30 C with the addition of a fan. With the reference heatsink and a forced air fan, a 7 Watt application could run in an ambient environment of up to 25 C. Results will vary per fan size, speed and placement location upon the platform. Reference heatsink without a fan 2.5 Watt Application 7 Watt Example Reference heatsink with a fan 20

21 Module Power Characterization Configure BIOS settings for characterized results In the CPU settings / System Power Options menu: Choose Intel Turbo Boost setting for your use case. Set PL1 to next whole number above the max power recorded in the log file; 3 for our 2.5W example. 3 Use the Go Back option and depress [F4] to commit changes to BIOS. 21

22