![VersaLogic VL-EPU-3311 Скачать руководство пользователя страница 59](http://html.mh-extra.com/html/versalogic/vl-epu-3311/vl-epu-3311_hardware-reference-manual_1006769059.webp)
Thermal Considerations
VL-EPU-3311 Reference Manual
59
EPU-3311 Thermal Characterization
The EPU-3311 board underwent the following thermal characterization tests:
Test Scenario 1: Single core EPU-3311-EAP + HDW-406 heat sink
Test Scenario 2: Dual core EPU-3311-EBP + HDW-406 heat sink, with/without HDW-411
fan
Test Scenario 3: Quad core EPU-3311-EDP + HDW-406 heat sink, with/without HDW-411
fan
Test Scenario 4: Quad core EPU-3311-EDP + HDW-406 heat sink + HDW-408 heat pipe
block, with/without HDW-411 fan
Table 19 describes the thermal testing setup for the board.
Table 19: EPU-3311 Thermal Testing Setup
Hardware configuration
EPU-3311 (Osprey) single/dual/quad core CPU with:
4 GB of DDR3L DRAM (2 GB for the single- and dual-core board models)
HDW-406 (passive heat sink)
HDW-408 (heat pipe block)
HDW-411 (heat sink fan)
One VGA display device (connected through the LVDS interface)
One SATA hard disk drive
Two RS-232 ports in loopback configuration
One VersaLogic VL-MPEe-E3 Mini PCIe Gigabit Ethernet module
Two active Ethernet ports in loopback configuration
Two USB 2.0 ports in loopback configuration
(Note)
USB keyboard and mouse
(Note)
BIOS
ID string: Osprey_3.1.0.334.r1.101
Passive thermal trip point setting: 105 ºC
Critical thermal trip point setting: 110 ºC
Operating system
Microsoft Windows 7, SP1
Test software
Passmark BurnIn Test v7.1 b1017
- CPU utilization ~90%
Intel Thermal Analysis Tool (TAT) v5.0.1014
- Primarily used to read the CPU core temperature
Test environment
Thermal chamber
Note:
This device is connected through a VersaLogic VL-CBR-4005B paddleboard.
The test results reflect the test environment within the temperature chamber used. This particular
chamber has an airflow of about 0.5 linear meters per second (~100 linear feet per minute).
Thermal performance can be greatly enhanced by increasing the airflow beyond 0.5 linear meters
per second.
The system power dissipation is primarily dependent on the application program; that is, its use
of computing or I/O resources. The stress levels used in this testing are considered to be at the
top of the range of a typical user’s needs.