Comparison and Analysis of Four Switch Architectures

Processing performance is the most important point for users when choosing core switches. At the same time, due to the diverse network environment of users, it is always in constant development and change. How to deal with future development and changes? How to deal with the challenges of users in different environments? All these challenges the core switch to support various services flexibly, and the ability of service superposition on demand becomes inevitable. At the same time, the service superposition should not affect the processing performance of the core switch, that is, the service and performance should be considered in the design Pay equal attention to requirements.

In response to this situation, the industry adopted the NP + ASIC design method. This architecture perfectly satisfies the design requirements of modern core switches with strong processing capabilities, business stacking on demand, and equal emphasis on business and performance, becoming the most current core switch design. Important development direction.

The architecture of the switch largely determines its processing capabilities and business support capabilities. At present, there are several commonly used technologies in the industry:

1 The advantage of a general-purpose CPU is that the function is easily expandable. In theory, any network function can be realized, but the disadvantage is that its performance is low. Therefore, in the architecture design of the switch, the general-purpose CPU is generally only used for the control and management of network devices.

2 ASIC chips can use hardware to implement a variety of commonly used network functions with extremely high performance. A single chip can achieve a processing capacity of more than several hundred MPPS. However, once the ASIC chip is developed, it is difficult to continue to expand other applications, and the addition of new functions requires the chip development company to spend a long development cycle. Therefore, ASIC chips are most suitable for processing various mature traditional functions in the network.

3 FPGA is a gate array chip that can be repeatedly programmed, erased, used, and can achieve different functions with different software when the peripheral circuits are not moving. Unfortunately, due to technical limitations, FPGAs have so far had limited processing power (in June this year, the industry launched FPGAs with 10G processing power, but they have not yet been applied to large-scale mature applications). Handle multiple protocols at the same time, and cannot be satisfied with the expansion of complex services. Therefore, in the switch architecture, FPGA is generally only used for the expansion of a small number of simple protocols.

4 The NP network processor consists of several microcode processors and several hardware coprocessors. In recent years, NP technology has been greatly developed, so that NP retains the characteristics of ASIC high-performance data processing (in 2002, the industry appeared 10G NP products, and now the industry has also appeared a small number of 40G processing NP products). At the same time, NP can perform complex multi-service expansion through microcode programming through many microcode processors operating in parallel. The disadvantage of NP technology is that network manufacturers need to invest a lot of relevant developers when using NP for product design. The NP of each manufacturer needs a unified standard, and it cannot replicate the development experience. At the same time, there are still some gaps between the performance of NP and ASIC Therefore, the NP network processor is applied to complex multi-service expansion of high-end network products, but it is not used to implement traditional network functions.

Undoubtedly, through the analysis of several system design technologies, it can be seen that the system design method using NP + ASIC is the most perfect choice. Use ASIC chips to process various traditional services at high speed, such as Layer 2 switching, Layer 3 routing, ACL, QoS, and multicast processing, etc., to meet the core switch's demand for switch processing performance; Mature business, flexibly supporting various business functions such as IPV6, Load Balancing, VPN, NAT, IDS, policy routing, MPLS, firewall, etc., to meet the needs of core switches for business stacking on demand; at the same time, NP is close to the efficient features of ASIC The high performance provided by multiple services is guaranteed, and the core switch's demand for powerful processing capabilities is still maintained.

LED flexible neon is the latest and most popular LED breakthrough product, which makes up for the shortage of glass neon tube and fiber. LED flexible neon is a professional linear lighting decoration product. Although it looks like glass neon tube, it can be bent at will. It has shatterproof and waterproof functions. It can be used indoors and outdoors.

Neon Flex Rope Lights

Led Neon Flex Rope Light,Neon Flex Rope Lights,Neon Flex Sign,Led Neon Flex

Shenzhen Oleda Technology Co.,Ltd , https://www.baiyangsign.com