What is IP in the chip industry and what is its use

Foreword: In the chip industry, people often use the term IP, such as IP development, IP trading, IP multiplexing, IP manufacturers, IP providers, etc. Outsiders may feel unclear about this term, and do not know what to say. They may think it is an IP card, or they may think it is an IP address in Internet technology, but it is not. This article will focus on popular science and talk about what IP is in the chip industry. "IC, IP, IQ card, all tell me the password!" is a classic line from Fan Wei in the movie "A World Without Thieves". In reality, there is no IQ card, only IC card and IP card. IC card refers to integrated circuit card, also called chip card. For example, bank cards, telecommunication cards, transportation cards and social security cards are all implemented with chip cards. The IP card is a network phone card that uses the TCP/IP protocol, and the IP card has long been eliminated. However, the TCP/IP protocol has always been used as the underlying technology of Internet communication, so words such as IP and IP address can often be heard. This is the first type of IP, which is a technical vocabulary in the Internet field, specifically referring to the IP protocol or IP address.

The second type of IP is a legal term, which is the abbreviation of Intellectual Property (Intellectual Property) in English. In modern society, many countries attach great importance to the protection of intellectual property rights. The chip industry is a highly intellectual and capital-intensive industry, so the protection of intellectual property rights in this industry is even more important. Intellectual property in the chip industry can be called Silicon Intellectual Property, or SIP for short. SIP's protection forms mainly include patent protection (Patent Protection) and integrated circuit layout protection (IC Layout Protection).

This article will introduce the third type of IP, which is both a technical vocabulary, a thing, and a commodity. It embodies the wisdom of chip designers, has the attributes of commodities and intellectual property rights, and can be promoted, sold and applied. Therefore, the chip industry has technological and commercial activities such as IP development, IP trading, and IP multiplexing. Some companies are also referred to as IP manufacturers and IP providers.

1. What is the IP of the chip industry?

The IP referred to in the chip industry is generally also called an IP core. IP core refers to the mature design of circuit modules with independent functions in the chip. The circuit module design can be applied to other chip design projects containing the circuit module, thereby reducing the design workload, shortening the design cycle, and improving the success rate of the chip design. The mature design of the circuit module embodies the designer's wisdom and reflects the designer's intellectual property rights. Therefore, the chip industry uses the IP core (Intellectual Property Core) to represent the mature design of this circuit module. IP core can also be understood as an intermediate component of chip design.

Generally speaking, a complex chip is composed of a circuit part designed by the chip designer and connected to multiple purchased IP cores. As shown in Figure 2, to design a chip with such a structure, the design company can outsource all the IP cores (modules in different colors) in the chip, and only design the creative and self-designed parts of the chip (indicated by green) , And connect the parts together.

Figure 2. The self-designed part is integrated with multiple IP cores to form a complex chip design. The chip design process in Figure 2 is like the system circuit board development process in Figure 3. It uses an existing and mature IP core (or Chip) is the process of layout, placement, and signal connection. This process can be called the multiplexing of the IP core (or chip). The difference is that in addition to the chip and the connection line on the system circuit board, system developers rarely develop their own chips. In the chip design process, in addition to using the purchased IP core on the chip, generally speaking, the chip designer has to design a part of his own circuit and complete the signal connection between each part, and finally the entire The function and performance of the chip are repeatedly checked and verified before manufacturing.

Figure 3. Building a complex chip with an IP core is similar to building a complete system circuit board with a chip. If the above introduction is too professional, you can also use a puzzle drawing to compare the chip design. You can abstract the chip shown in Figure 2 The ground is understood as the puzzle picture shown in Figure 4. The IP cores with different functions purchased in the chip are represented by different colored blocks, and the self-designed circuit parts are represented by green blocks. The design process of a complex chip is like putting together this picture. The same is that the existing tiles (IP cores) are used to stitch beautiful pictures (complex chips). The difference is that the puzzle drawing only needs to consider the shape of the tiles, while the chip design needs to consider many parameters and indicators of the IP core, and correctly connect each IP core with the self-designed part to ensure that the function and performance of the entire chip are correct.

Figure 4. IP cores make it easier to design complex chips. IP multiplexing is similar to a "puzzle". IP cores are used by other chip design companies and are called IP multiplexing in the industry. Specially design relatively independent circuit function modules, the purpose is to promote to other chip design companies for reuse, this kind of design work is called IP development. Companies that specialize in IP development are called IP vendors, or IP providers. It is an IP transaction behavior for IP manufacturers to sell IP to chip design companies.

2. The origin and function of IP

The origin of IP starts from the early chip design process. The integrated scale of the early chips was limited and the design complexity was not high. All the circuits on the chip were completed by the chip designer independently. Chip companies with low design levels and limited capabilities can only design small and simple chips. Only chip companies with high design levels and strong capabilities can design large-scale chips with complex functions. In this period, no matter the chip scale is large or small, the chip from the "head" to the "foot" is designed by the chip company itself. The early high-end chips were basically controlled by a few large international chip companies.

As the modern information society has higher requirements for chips, the scale of chips has increased exponentially, and the complexity has increased dramatically. It is almost impossible for small and medium chip companies to independently complete a complex chip design. Especially in the late 1980s, the foundry business model emerged in the chip industry, and a large number of small and medium-sized microchip design companies (Fabless) emerged. During this period, the chip design industry urgently needs to solve the problem that small chip companies cannot design large chips.

There are many enlightening ideas to solve this problem. For example: building blocks and puzzles and drawing toys; designing large machines from standard parts; calling and designing large software by software subprograms (or middleware); building large electronic systems with chips, etc. The idea is to reuse pre-designed mature components to build a more complex system, eliminating the need to consider the internal problems of the components, and simplifying the complexity; reusing components, reducing repetitive labor, and saving time; reusing components to improve the whole The success rate of complex system construction.

IP core development and IP reuse in the chip design industry are inspired by these ideas. The IP core is similar to the above-mentioned components. The IP core is a pre-designed circuit module design with independent functions. With the IP core, the design of large and complex chips becomes easier, shorter in cycle, and easier to succeed.

There are four main functions of IP. One is to simplify the chip design, shorten the chip design cycle, and improve the success rate of complex chip design. The second is IP development and IP multiplexing technology that make it possible for small companies to design large chips; the third is to enable system companies to design their own chips, and to improve independent innovation capabilities and the independent intellectual property content of the entire system; fourth is to enable chip design The industry got rid of the traditional IDM model and became an independent industry in the industry chain, which promoted the rapid development of the chip design industry.

At present, although many small and medium-sized microchip design companies have limited design capabilities and levels, in order to seize the market and shorten the chip design cycle, they will purchase many IP cores to complete their own chip design projects. The number of IP developers and IP providers in the industry is increasing and becoming more and more professional. Various functions and various types of IP cores continue to emerge. IP trading activities are becoming more and more common, and the transaction amount is getting larger and larger.

3. Types and examples of IP

IP cores are classified into three levels of behavior (Behavior), structure (Structure) and physical (Physical), corresponding to three types of IP cores. They are soft cores (Soft IP Core) designed by hardware description language. ), the solid core (Firm IP Core) that completes the structure description and the hard core (Hard IP Core) based on physical description and process verification.

1. IP soft core: it is an independent functional circuit module designed with hardware description language (HDL). From the perspective of chip design, it has only undergone RTL-level design optimization and functional verification, and is usually submitted to users in the form of HDL text. So it does not contain any physical realization information, therefore, the IP soft core has nothing to do with the manufacturing process.

After the user purchases the IP soft core, the correct gate-level design netlist can be synthesized, and the subsequent structural design can be carried out, which has great flexibility. With the help of EDA synthesis tools, users can easily integrate with other IP soft cores and self-designed circuit parts, and design chips with different performances according to various semiconductor processes.

2. IP solid core: Its design level is between IP soft core and IP hard core. In addition to completing all the design of IP soft core, it also completes design links such as gate-level circuit synthesis and timing simulation. Generally, it is provided to users in the form of a gate-level circuit netlist.

3. IP hard core: It provides a mask-level circuit module at the final stage of circuit design. It is provided to users in the form of the final layout and routing netlist. IP hard core not only has the predictability of results, but also can optimize power consumption and size for a specific process or a specific IP provider. Therefore, the three types of IP cores are products of circuit functional module designs at different design stages, as shown in Figure 5.

Figure 5. Different types of IP cores are available at different stages of circuit function module design. After careful evaluation and selection, users purchase IP cores from IP manufacturers and begin to design their own chips. As mentioned earlier, a complex chip is generally composed of a purchased IP core and a circuit part designed by the user. The chip design process includes three stages: behavioral level, structural level, and physical level. The work in the behavioral and structural design phases is generally referred to as front-end design, and the work in the physical-level design phase is generally referred to as back-end design. The schematic diagram in Figure 6 illustrates that different types of IP cores are added to the entire chip design at different design stages.

Figure 6. The three types of IP cores are merged into the design at different stages of the chip design, forming a whole. The three types of IP cores have their own advantages and disadvantages, and users will choose according to their actual needs. The following is a brief summary of the advantages and disadvantages of the three IP cores.

IP soft core: It is delivered to users in the form of integrated source code. Its advantage is that the source code is flexible, it can be reconfigured at the function level, and the target manufacturing process can be flexibly selected. High flexibility, strong portability, allowing users to configure themselves. The disadvantage is that the predictability of circuit function modules is poor, and there is a possibility of error in the subsequent design, and there is a certain design risk. Moreover, the protection of intellectual property rights of IP soft cores is relatively difficult.

IP solid core: Its flexibility and success rate are between the IP soft core and the IP hard core, which is a compromise type. Compared with the IP soft core, the design flexibility of the IP solid core is slightly inferior, but the reliability is greatly improved. At present, IP solid core is one of the mainstream forms of IP core.

IP hard core: Its biggest advantage is to ensure that the performance, such as speed, power consumption, etc., achieve the expected results. However, the IP hard core is related to the manufacturing process, it is difficult to transfer to a new process or integrate into a new structure, and cannot be reconfigured. The characteristics of IP hard cores that cannot be modified make it difficult to reuse, so they can only be used in certain specific applications and have a narrow range of use. But IP hardcore intellectual property protection is the most convenient.

Examples of IP cores include various types of CPU IP cores from ARM. Many IP vendors provide DSP IP cores, USB IP cores, PCI-X IP cores, WiFi IP cores, Ethernet IP cores, embedded memory IP cores, etc., with a wide variety of varieties.

If divided by major categories, it can be roughly divided into processor and microcontroller IP, memory IP, peripheral and interface IP, analog and mixed circuit IP, communication IP, image and media IP, etc.

Some of the world's major EDA suppliers are also IP suppliers. For example, the United States Synopsys (Synopsys) can provide thousands of various types of IP. Covers Logic Libraries, Embedded Memories, Analog Libraries, Wired and Wireless Interface, Security, Embedded Processors, and subsystems (Subsystems) and other aspects of the IP.

Fourth, the development and application of IP

It can be seen from Figure 4 and Figure 5 that the development of IP cores is similar to the process of chip design. Both need to keep up with the development trend of chip technology, accumulate chip design technology and understand the development pace of chip manufacturing process. The difference is that the business model of a chip design company is to design a complete chip and push it to the market to realize chip sales profits. However, IP developers do not design complete chips. They only design certain functional modules of the chip. The design of these functional modules (IP cores) is optimized, and the adaptability of IP cores in multiple multiplexing situations must be realized. ,Facilitation. IP developers seek that these IPs can be purchased and adopted by many chip design companies and realize IP sales profits.

However, the technical difficulty and requirements of IP development are very high, and it is not something that any chip design company wants to do. This is mainly reflected in, first, the completeness requirements of IP. As a component of chip design, IP core can be compared with component (middleware) of software design. Small flaws can be hidden in the software components, and the integrated system software can be continuously improved by patching and upgrading when it is put into actual application. But the IP core must be very perfect, and there must be no flaws. Because any small flaw in the IP core may affect the success or failure of the entire complex chip design, and affect the investment and output of the entire chip research and development that has cost millions or even hundreds of millions of yuan.

Therefore, the price of complex IP cores is extremely expensive, ranging from hundreds of thousands to millions of dollars. In order to facilitate promotion, IP manufacturers generally adopt a business model of charging an IP license fee (IP License) in the early stage + a royalty fee (Royalty) calculated based on chip shipments in the later stage to reduce users' R&D investment. The second is the multi-parameter requirements of IP, including process nodes, power supply, power consumption, performance, etc. All of these require IP developers to be very familiar with chip design, manufacturing processes, and industry applications. The third is the user’s credibility of the IP provider

Requirements, including the company’s technical strength, influence, and stability. It is difficult for users to accept IP provided by a small and unstable IP manufacturer, even if these IPs are cheap or even free. Users would rather buy more expensive IPs provided by well-known and large IP vendors, because such IPs have the guarantee of quality and follow-up services.

Since 2000, under the promotion of the National Development Document No. 18 and the eight national IC design industrialization bases, domestic IC design companies have sprung up and developed rapidly. Most of these companies are small, medium and micro IC design companies. Thanks to IP development and IP reuse technology, these small and medium IC design companies can develop various functions of system-on-chip (SOC) and artificial intelligence (SOC) in a timely manner according to market needs. AI) chips, Internet of Things (IoT) chips, etc. These have promoted the rapid development of the domestic IC design industry and promoted the development of my country's new generation of information technology industry. It can be said that foundry technology, IP development and IP multiplexing technology have achieved the foundation for the rapid development of my country's current chip industry.

Postscript: IP referred to in the chip industry is also called IP core. IP core refers to the mature design of circuit modules with independent functions in the chip. In the three different stages of circuit module design, the different types of IP cores that can be obtained are called IP soft cores, IP solid cores, and IP hard cores. From softening to curing, then to hardening, the configurability and flexibility of the IP core become smaller, but the integrity is improved, and the risk of reuse is reduced. Chip design companies buy IP cores from IP manufacturers and can quickly design chips with complex functions through IP multiplexing, which can save design time and improve the success rate of chip design. IP development and IP multiplexing technology have greatly promoted the rapid development of the chip design industry.