Skip to main content

Overview

Dev MK4 Compute Blade with part descriptionDev MK4 Compute Blade with part description

Download the Datasheet

Features and Specifications

The Compute Blade comes in three different hardware variants.

Dev Compute Blade

MK4-k Dev Bladev1.0 Dev box label
CompatibilityRaspberry Pi Compute Module 4 / 4 Lite
Raspberry Pi Compute Module 4 Compatible Boards
StorageM.2 NVMe SSD up to 22110
Micro SD Card on lite CM4 Models
eMMC on non-lite CM4 models
GPIO2 x UART
7 x User Controllable GPIO including I2C
Fan Unit Connector
Programmable Button
2 x Controllable RGB LEDs
PoE+ Voltage40V-57V
PoE+ Current60 mA - 600 mA
3.3V Output5A
5V Output5.1A
Input Power (Normal Operation)2 - 8W
PortsHDMI (4K60)
USB Type-A/Type-C
SecurityTPM 2.0
Hardware switchable Wi-Fi, Bluetooth and EEPROM write-protection
Support for ZYMKEY 4i, Security Keys

Thermal Configuration

The Compute Blade is designed for deployment in enclosures, supporting configurations from a single pair in desktop enclosures to up to 10 pairs in a 19-inch 1U Rack BladeRunner rack-mounted chassis. Each pair is cooled using a specially designed Fan Unit equipped with a single 40x40x20mm fan.

The design of the Fan Units provides robustness and reliability: in the event of one Compute Blade being disconnected or failing, the Fan Unit will continue to operate, maintaining effective cooling for the remaining blade.

mk4 thermal flow
warning

The Compute blade has not been validated without the addition of the Fan Unit, BladeRunner™️, and Heat Sink

Cooling Configuration

Paired Blades: Pairs of Compute Blades are physically separated and each pair is cooled by its dedicated Fan Unit.

Fan Unit Design: The Fan Units are designed for easy replacement without interrupting operation. Each Fan Unit receives power from both Compute Blades in the pair, ensuring that the fan continues to run even if one of the blades is powered off.

Thermal Dissipation Capacity

This cooling configuration can dissipate up to 60W of heat per slot, which represents the maximum power consumption of two blades. This level of thermal dissipation corresponds to an extreme case scenario. On average, measured thermal dissipation for a pair of Compute Blades rarely exceeds 20W.

Block Scheme

This is the top part of the Raspberry Pi's GPIO connector. On the Compute Blade, PIN 1 corresponds to 3.3V, while on the Raspberry Pi's standard connector, it is labeled as 3.3V PWR.

Dev Blade block Scheme

Mechanical

Compute Blade dimensions