ACOS Technology Platform
The Advanced Core Operating System (ACOS)
A10 Networks ACOS is increasingly the platform of choice for enterprises, service providers, Web giants and government organizations seeking to optimize the performance and security of their data center applications and networks.
Best-in-Class Price/Performance: ACOS Shared Memory Architecture and Flexible Traffic Accelerator
Conventional application networking vendors have been challenged to scale system performance with rapidly escalating network backbone speeds, given the expensive nature of general-purpose computing platforms that are required to analyze unstructured application data within data network packets. The A10 Networks ACOS platform leverages our Shared Memory Architecture and Flexible Traffic Accelerator to efficiently utilize multi-core processors and scale performance linearly with increasing CPU/processor density, allowing customers to meet rapidly growing application networking performance scalability requirements. By taking full advantage of the capabilities of multi-core processors, our ACOS application networking platform provides end-customers with superior price/performance benefits. For example, ACOS can deliver two-to-five times more Web transactions (measured by Layer 4 connections per second) in certain head-to-head product comparisons per unit of computing and memory resources, power, rack space or list price.
ACOS Advanced Core Architecture
The ACOS platform was built to maximize the performance of the most expensive system resources, processors and memory in order to deliver the highest performance Application Services Gateways (ASGs). To improve the performance of our multi-core processor architecture, we developed the Shared Memory Architecture, which allows all processors to share common memory and system state simultaneously.
As a result, we avoid the overhead associated with inter-processor communication (IPC) architectures deployed in first-generation approaches. We optimize memory to be visible to all cores simultaneously, while minimizing communication overhead and contention among processors for allocated memory space. This efficient design achieves significantly greater performance and throughput from comparable levels of memory and processor resources, eliminating memory duplication and repetitive inter-processor communication tasks. Similarly, ACOS leverages a Flexible Traffic Accelerator, which is a high-performance intelligent network I/O technology that can balance application traffic flows equitably across processor cores.
Unlike competitor platforms that inefficiently dedicate processor resources to certain limited tasks, our system maximizes the utility of all processor cores by distributing traffic equally and making the shared memory available to all cores. Our Flexible Traffic Accelerator logic can be implemented either as software running within a standard x86 processor or a Field Programmable Gate Array (FPGA) semiconductor.
Our Flexible Traffic Accelerator also performs certain hardware-based security checks for each packet and can discard suspicious traffic before it can impact system performance. The combined effect of our Shared Memory Architecture and Flexible Traffic Accelerator provides customers with superior price performance in application networking products that scale their application and security services with the speed of their network backbone.
Agile Software-Based Platform
An increasing percentage of networking industry challenges can only be resolved with a high-performance, application-aware network infrastructure. For example, accelerating the performance of Web server farms, conducting protocol translation between IPv4 and IPv6, and detecting distributed botnet cyber attacks all require an application networking platform that can look deeply into application content, modify content for performance optimization or security purposes, and forward the traffic at rapidly escalating network data rates. The inherently flexible, software-based nature of ACOS allows us to constantly evolve and expand the feature set into these and other applications, which we have successfully demonstrated through expansion into our Carrier Grade NAT (CGN) product line and our new Threat Protection System (TPS) product line that detect and mitigate DDoS attacks. Similarly, the ACOS platform has been expanded to support a growing range of management and cloud orchestration features that support the evolving migration towards cloud computing architectures. The inherently software-based, agile ACOS architecture allows us to rapidly introduce new features and products to address changing market needs. And because ACOS is entirely software-based, it can be rapidly adapted to new use cases and customer needs.
Data Center Efficient Designs
Efficient design is the hallmark of the modern data center. Hyperscale and cloud-based data centers are being built with rapidly increasing capacity, giving rise to a critical need to squeeze more from the same power, cooling and rack space to achieve world-class cost per workload. The ACOS platform has the industry’s most efficient data center design, delivering significantly more throughput per unit of power and size. These efficiency benefits are a direct result of the high-performance design described above, which requires dramatically fewer CPU cores and memory to achieve a comparable capacity relative to conventional designs. In recent large data center wins, A10 has delivered comparable functionality and performance as competitive offerings at one-third the power and one-seventh the rack space requirements. These are real and demonstrable ROI results addressing critical efficiency requirements of modern efficient data center designs.