Beskrivelse
Getting started guide: https://www.youtube.com/watch?v=2l90-AyoXls
Rack-mountable, PoE-powered carrier board for Raspberry Pi Compute Module 4/5 and compatible modules with all the necessary interfaces
The Compute Blade, an Uptime Industries product, is a enterprise grade rack-mountable, PoE-powered carrier board for Raspberry Pi Compute Module 4/5 with all the necessary interfaces. With Compute Blade, you can create a high-density, low-power-consuming, plug-and-play blade server for home or data-center use. The Compute Blades were designed to work 24/7.
Create your high-density server with Compute Blade: You can fit up to 20 Compute Blades into a 1U 19-inch rack space, which can add up to 80 ARM cores, 320GB RAM, and 10TB of NVMe flash storage depending on your configuration.
Key Features
- Raspberry Pi CM4 and CM5 support
- support for fan using PWM fan connector
- supports NVMe SSDs up to 22110 (2230, 2242, 2260, 2280)
- Power over Ethernet IEEE 802.3at (PoE+)
Interfaces
- Gigabit Ethernet
- HDMI (up to 4k60) (Compute Blade TPM and Compute Blade Dev)
- UART0 on the front
- Uptime high speed module support
- USB-A, USB-C switchable USB input
Additional features:
- TPM 2.0 secure boot (Compute Blade TPM and Compute Blade Dev)
- WiFi, BT and EEPROM write-protection (WiFi + BT require CM4/5 with WiFi / BT support, only on Compute Blade Dev)
- Controllable RGB LEDs
A Compute Module (CM4) or Compute Module 5 (CM5) is required to use this device, and is sold separately. The compute modules and other accessories shown in the images are shown for illustration purposes.
We offer a variety of models in our shop, for the operation of the Compute Blade we recommend a Compute Module 4 with the maximum amount of RAM (8GB), for example CM4108032.
Specifications
| Compatibility | Raspberry Pi Compute Module 4 / 4 Lite Raspberry Pi Compute Module 5 / 5 Lite Raspberry Pi Compute Module 4 Compatible Boards Raspberry Pi Compute Module 5 Compatible Boards |
|---|---|
| Storage | M.2 NVMe up to 22110 Micro SD |
| Networking | Gigabit Ethernet (Wi-Fi on some CM4/CM5 models, not recommended in rack usage) |
| GPIO | 2 x UART 7 x User Controllable GPIO including I2C Fan Unit Connector Programmable Button 2 x Controllable RGB LEDs |
| Operating Voltage | 5V – USB C 5.1V – Power over Ethernet IEEE 802.3at (PoE+) up to 30W @5.1V |
| Input Power (Normal Operation) | 2 – 8W |
| Ports | HDMI (4K60) USB A USB C |
| Security | TPM 2.0 Hardware switchable Wi-Fi, Bluetooth and EEPROM write-protection Support for ZYMKEY 4i, Security Keys |
| Dimensions | 42.5 x 250 x 17.5 mm |
Accessories for the Compute Blade platform
The Compute Blade platform is a product family, consisting of products optimized to work with each other:
- Compute Blade (this product), compatible with Raspberry Pi Compute Module CM4 / CM5
- Compute Blade Fan Unit (Basic Fan Unit & Smart Fan Unit versions) – one fan unit cools two compute blades.
- Compute Blade Heatsink – available in black, red, blue
- BladeRunner Compute Blade Enclosures – available in a 19”-rack mountable version, and a 4-unit housing version
- (coming soon): AI accelerator module, adds two Coral TPUs to the compute blade
- (coming soon): BladeCluster Pro and BladeCluster ProSW – fully assembled 19” 1U metal racks, including compute modules and software (ProSW only)
Advantages of the Compute Blade platform
Unlike conventional blade servers, the Compute Blade system eliminates a single point of failure as it doesn’t rely on a server platform. While it still uses a PoE network switch (also necessary for blade servers), switches are simpler to replace or keep as spares compared to entire server platforms. Network equipment typically boasts high reliability.
Individual blades can be rebooted or power-cycled by momentarily disabling their switch port power. Thanks to the CM4/5’s network boot feature, reprovisioning and system rescue are straightforward. Each blade is compact and devoid of moving parts.
Note: the optional fans which are recommended for high loads of course include moving parts
Versions
The Compute Blade is available in three different versions:
- The Compute Blade Basic is intended for scaling your cluster environment. It contains only the necessary components for an enterprise grade worker node. This is also the most affordable version of the Compute Blade.
- Compute Blade TPM: Adds a TPM 2.0 for advanced security
- Compute Blade Dev: Recommended as master node, and for development and homelab environments.
| features | Compute Blade Basic | Compute Blade TPM | Compute Blade Dev |
| Raspberry Pi CM4 and CM5 support | yes | yes | yes |
| M.2 Key M up to 22110 for NVMe disks or high-speed modules | yes | yes | yes |
| 1Gbit Ethernet with PoE up to 30 W | yes | yes | yes |
| UART | yes | yes | yes |
| Compute Blade headers port | yes | yes | yes |
| Button | yes | yes | yes |
| Stealth mode | yes | yes | yes |
| Digital LEDs | yes | yes | yes |
| USB-A | yes | yes | yes |
| TPM 2.0 onboard | yes | yes | |
| HDMI port | yes | ||
| USB-C for bootloader update | yes | ||
| microSD card slot | yes | ||
| Wi-Fi, BT, nRPIBOOT | locked | locked | switchable |
The image above shows the different options and extensions which the Compute Blade will support, amongst others, opening up a world of possibilities for your application.
Applications
Edge servers
Reduce latency and extend compute power to make your processes leaner, more efficient, and more cost-effective.
CI/CD systems and Automated tests
Perform performance tests on dedicated hardware for far more stable results than running tests on virtual machines.
Stateless Computing
Seamlessly deploy specialized stateless computing platforms, such as OpenFaaS, onto the blades to enhance functionality and performance.
High Availability Computing
Deploy robust, high-availability computing to maintain critical operations and services with minimal downtime.
Education
Enhance educational experiences with practical, hands-on tech learning, facilitating innovation and computing skills.
Private Cloud
Create your own secure, scalable private cloud for efficient data management and tailored IT services.
Home Labs
An enterprise-level homelab experience that you can use to build, play with, and explore new technologies
Hosting provider
Provide dedicated resources and isolate customer machines on a physical layer, to protect against modern CPU/hypervisor exploits.
Smart Homes
Make your home server highly available and increase the possibilities e.g by adding more modules like the Uptime AI-Module.
Documentation
FAQ
What is a blade server?
A blade server is a compact, self-contained computer server, specifically designed to minimize the use of physical space and energy. Unlike traditional rack-mount servers, which are standalone servers that require individual components such as network and storage interfaces, blade servers are housed in a chassis that provides these resources collectively. This setup allows for a high density of servers, which can share power, cooling, networking, and storage facilities provided by the chassis.
Blade servers are particularly suitable for environments requiring high compute density without the need for a lot of individual storage, like web hosting, virtualization, and cloud computing services.
Where can I find out more about Uptime Industries, what is their story?
Uptime Industries was founded in April 2023 by Ivan Kuleshov and Sascha Bührle following a successful Kickstarter campaign “Compute Blade: Your Rack-Mountable ARM Cluster,” where the first 12,000 blades were sold and the community expanded.
Based on community requests, Uptime Industries decided in early 2024 to extend the Compute Blade platform with multiple modules and accessories, besides working on its own software products for a holistic Compute Blade experience, innovating in the cluster computing space.
During the same time, a well-recognized VC, TrueVentures, joined forces to expand and scale the Compute Blade platform rapidly besides developing an upcoming edgeAI focused product line.
You can find out more about Uptime Industries on their homepage:
How does the Compute Blade contribute to energy savings and sustainability?
The Compute Blade is designed with energy efficiency at its core. Its low-power ARM cores and PoE+ capabilities allow you to scale your computing needs without the high energy costs typically associated with traditional servers. This not only results in operational cost savings but also reduces your carbon footprint, contributing to a greener planet.
Can the Compute Blade be used for educational purposes?
Absolutely! The Compute Blade is an excellent tool for educational environments. It provides a tangible, hands-on experience for students learning about computing, networking, and hardware. With its modular design and safe, low-power operation, it’s perfect for classroom settings.
What support options are available for Compute Blade users?
Uptime Industries offers comprehensive support for all Compute Blade users. This includes an extensive online knowledge base, a user forum for community assistance, and direct customer service for more complex inquiries. Please refer to the following links to get started:
How does Compute Blade ensure data security and privacy?
Security is paramount for Uptime Industries. The Compute Blade comes equipped with TPM 2.0 for hardware-based security (in the TPM and Dev versions), including secure boot processes and encryption key generation.
What are the upgrade options for the Compute Blade?
The Compute Blade is designed with future-proofing in mind. It supports various add-ons and modules, allowing you to expand storage, add AI co-processors, or enhance security features. You also have the option to choose / upgrade your compute modules to other CM4/CM5 versions, in case, for example, your RAM requirements change.
Please note: at this stage, CM5 is not publicly available, therefore CM5 compatibility can’t be guaranteed.
What is a TPM?
TPM stands for Trusted Platform Module. It is an industry standard module for secure storage and generation of encryption keys. Whereas software offers attack surfaces through algorithm changes (e.g. allowing a savvy attacker to brute-force a password), the TPM is designed to start increasing lock-out times under attack, thereby making it physically impossible to brute-force the password. It can also be used to verify that you are indeed talking to the device you want to be talking to, this way Man in the middle attacks can be mitigated.
Note that these features also require software implementation, please check if a software implementation for your use case exists.
What advantages do PoE-powered computing solutions like the Compute Blade offer?
PoE-powered computing solutions, such as the Compute Blade, streamline power management and cable clutter, leading to cleaner installations and easier maintenance. This also translates to lower setup costs and increased reliability for both home and enterprise networks.
Are there specific Raspberry Pi server solutions that the Compute Blade supports? The Compute Blade is optimized for the Raspberry Pi Compute Module 4 and 5, supporting a range of configurations that cater to various applications, from educational platforms to robust data center innovations.
Anmeldelser
Der er endnu ikke nogle anmeldelser.