The hyperconverged technology for the next generation datacenters

Kaleao has its roots in research and development. The company has set up longstanding collaborations with prestigious research centers and leverages its own R&D to design and manufacture innovative products that help data centers to become more efficient and effective.
 

True Convergence

KALEAO developed patented software and hardware technologies that enable a new approach to hyperconvergence. True convergence is the native convergence of compute, storage and network within a framework that makes the delivery of virtualization and software defined infrastructure much more efficient and effective.


With true convergence, hardware resources are organized in global pools of compute, storage and networking from which they can be dynamically associated to virtual machines through an intelligent and highly efficient “microvisor” which performs the I/O logic and management, without replicating hardware in software. With the true converged approach hardware stays in hardware, despite being fully software defined and controlled.


The net results is gaining agility in managing IT infrastructure, without compromising in performance, scalability and energy efficiency.

 



 

 
TRUE CONVERGENCE SCHEMAS

True Convergence Schemas

 

 

 
 

Innovation

KALEAO innovates by leveraging R&D to design hardware and software architectures that allow IT systems to be more efficient, compact and optimized. This gives KALEAO the possibility to propose to IT leaders more effective and efficient ways of managing their infrastructure by maximizing cost savings and streamlining operations. At the same time, it allows to build on existing or emerging technologies and environments to keep compatibility and standardization.


 
 


Software defined, hardware accelerated

At the heart of true convergence there is "physicalization" (KALEAO own neologism, coming from the mix of “physical” and “virtualization”) , the technology that allows the dynamic creation of I/O hardware interfaces, leveraging components like FPGAs. In this way, KALEAO elevates the idea of virtualization from a resource-consuming layer to an efficient combination of software and hardware.


When new hardware is added, it is exposed to global pools of physicalized resources spanning across the system. These resources can be allocated directly to VMs, maintaining bare metal performance and efficiency, but keeping the flexibility of being dynamically expanded, retracted and configured.


This approach is unique to KALEAO and only possible because of the company competences in mastering both the hardware and the software technologies, development and manufacturing.

 

 

Kaleao: an IP company

KALEAO owns patents to protect the IP behind the core parts of its technology. Giving customers the possibility to leverage unique solutions to perform better and distinguish themselves from competition.
 
 


Compatibility

KMAX is compatible with the vast ecosystem of software already supported by the ARM 64-bit architecture. This is a fast-growing universe of OS, middleware and applications, behind which there is an effort involving companies and organisations like ARM, Google, Linaro, Qualcomm together with the chip manufacturers and final users.


The supported operating systems are different Linux distros, like Ubuntu, CentOS, Red Hat…The combination of the small footprint “Microvisor” and “physicalization” enables a high performance and ultra-efficient virtualization layer.


Most of the middleware software, especially open source, for cloud, web hosting, IoT, data analytics, big data … is nowadays compatible with ARM 64-bit.

 
 
ARM based intelligence
 
 
MORE

With the high demand for mobile services and the move of internet usage into smartphones, mobile technology has made remarkable progress in delivering the maximum performance possible within restrictive power budgets. KALEAO systems use ARM based processors, typically found in mobile devices, and leverages a design finalized to maximize efficiency without compromising performance for a large range of workloads. The architectural approach and the physical characteristics of the main components bear the additional advantage of allowing a very compact design, delivering the maximum density possible.


ARM also offered the openness KALEAO needed to engineer a more optimized and rationalized server design, which is enabling true convergence and a more efficient and effective software defined infrastructure.

ARM® big.LITTLE™
 
 
MORE
KALEAO is the first server platform to leverage ARM® big.LITTLE™ processing power-optimization technology. With this technology, high-performance ARM CPU cores are combined with the most efficient ARM CPU cores to deliver peak-performance capacity, higher sustained performance, and increased parallel processing performance, at significantly lower average power. The latest big.LITTLE software and platforms can save 75% of CPU energy in low to moderate performance scenarios, and can increase performance by 40% in highly threaded workloads.
 
Optimized hardware and software integration
 
 
MORE
KALEAO technology focuses on the concept of next generation micro servers, systems capable of exceptional performance while maintaining the control over energy consumption. A similar result is achieved by selecting the latest generation of ARM processors together with an optimized hardware and software design that exploits intelligent active components (like FPGAs), ad hoc low overhead virtualization, ultra-efficient software defined networking and storage.
Scalable compute unit architecture
 
MORE
KALEAO technology enables a unique scale-out hardware and software architecture that brings modularity and scalability as well as resource sharing capabilities. KALEAO systems preserve full out-of-the-box compatibility with the broad ecosystem of software applications and environments available from the open-source community and the ARM-based platform.
 


Distribute scale out approach

KMAX pools storage, memory and network resources making them available to computation. In this way, it is possible to create flexible aggregation of resources setting up physical servers of any size matching workloads of different nature and with different resource requirements.