These eight domestic automotive-grade chips are hyped

Automotive chips have been a hot topic in recent years. With the continuous advancement of the wave of automotive electrification and intelligence , many automotive original equipment manufacturers (OEMs) have put forward new requirements for the localization rate. Under such industry changes, domestic chips have undoubtedly ushered in more opportunities.

However, in recent years, the competition in the domestic market has been extremely fierce, and various parties have launched automotive-grade chips with excellent performance. In this dazzling ocean of products, it is not easy to find a satisfactory chip. The following chips have gone through many tests in the industry and are very popular in industry applications.

01. Horizon-Journey 5

Among the automotive-grade chips , autonomous driving chips are crucial. Autonomous driving chips are a type of high -computing power chip that has emerged with the development of smart cars . Mainly SoC chips , automotive SoCs mainly integrate system-level chip control logic modules, CPU cores, graphics processors , DSP modules, memory modules, external communication interface modules, analog front-end modules with ADC /DAC, power supply and power management modules, etc. Since 2015, with the rapid installation of intelligent driving , intelligent and multi-screen car systems, and intelligent modules such as HUD and voice recognition , the trend of automotive intelligence has accelerated. SoC chips have gradually replaced MCUs as automotive main control chips due to their advantages in functional integration, performance and hardware scalability . Relevant data show that for each level of autonomous driving, the computing power requirement will increase by more than ten times.

According to the matching data between autonomous driving capability and chip computing power requirements, the AI ​​computing power required for L3 reaches 30TOPS, the AI ​​computing power required for L4 is close to 400TOPS, and the AI ​​computing power required for L5 is more stringent, reaching 4000+TOPS. From the perspective of market application, autonomous driving chips in the commercial stage are mainly concentrated in the field of advanced driver assistance systems , which can achieve L1-L2 level assisted driving functions. Some chip companies also claim that their products can achieve L3 level functions. However, L4-L5 level autonomous driving is still some distance away from large-scale commercialization.

At present, NVIDIA , Mobileye , and Qualcomm are leading in the field of autonomous driving SoC, and each of them has different advantages. Among them, NVIDIA is the king of high-computing chips; Mobileye is the leader in the field of assisted driving; Qualcomm is the leader in cockpit chips. In recent years, emerging chip technology companies represented by Huawei , Horizon Robotics, Black Sesame, and CoreDrive Technology have entered this blue ocean market with the advantages of AI algorithms. Many manufacturers have launched a series of products with excellent performance and high reliability. Here we want to mention a popular autonomous driving chip, Horizon Robotics' Journey 5.

According to statistics from Gasgoo Automotive Research Institute, in 2023, the number of pre-installed standard intelligent driving domain controllers for passenger cars in the Chinese market (excluding imports and exports) will reach 1.839 million, a year-on-year increase of about 70%, and the pre-installation rate will be about 8.7%. In addition, in the ranking of installed capacity of intelligent driving domain control chips in the Chinese market in 2023:

The first place is Tesla 's FSD chip, with shipments of about 1.208 million, accounting for 37%; the second place is Nvidia's Orin-X chip, with shipments of 1.095 million, accounting for 33.5%; the third place is Horizon's Journey 5 chip, with shipments of 200,000, accounting for 6.1%; the fourth place is Mobileye's EyeQ4H chip, which is equivalent to the shipment volume of J5 (Journey 5), also about 200,000, accounting for 6.1%; the fifth place is Mobileye's EyeQ5H chip, with shipments of 174,000, accounting for 5.4%.

Horizon Journey 5 is a popular high-performance, high-computing vehicle-mounted smart chip. Journey 5 is Horizon's third-generation automotive-grade product, and is also the first domestic vehicle-mounted smart chip developed in accordance with the ISO 26262 functional safety certification process and certified by ASIL-B. Based on the latest Horizon BPU Bayesian architecture design, it can provide up to 128TOPS computing power; it has rich external interfaces and can access more than 16 high-definition video inputs; relying on powerful heterogeneous computing resources, it is not only suitable for the most advanced image perception algorithm acceleration, but also supports laser radar , millimeter- wave radar, and other multi- sensor fusion; it supports predictive planning and H.265/JPEG real-time encoding and decoding, and is an ideal choice for mass production of high-level autonomous driving and smart cockpits .

What cars are available in Horizon Journey 5?

According to data released by Horizon Robotics in September last year, the shipment volume of the Journey 5 chip has exceeded 200,000 pieces since its global mass production in September 2022 to the release of data in September 2023, with an average monthly shipment of more than 20,000 pieces, helping a number of mid-to-high-end new energy models win the sales champion in their market segments.

In terms of specific vehicle installation, many models of Ideal Auto , including Ideal L9 and Ideal L8, are already equipped with the Ideal AD Pro intelligent driving system equipped with the Horizon Journey 5 chip. With the support of the Journey 5 chip, these models have achieved the industry-leading high-speed NOA function, providing users with the ultimate intelligent driving experience. In addition to Ideal, Horizon Journey 5 has also obtained mass production cooperation with many mainstream car companies, including NIO , SAIC Group, Changan Automobile, GAC Aion, FAW Hongqi, Nezha Automobile, Chery Automobile, etc.

In April this year, Horizon launched the new generation of in-vehicle intelligent computing solutions, the Journey 6 series, and the Horizon SuperDrive scenario intelligent driving solution. At the press conference, Horizon officially announced the first batch of 10 mass production cooperation car companies and brands for the Journey 6 series, including SAIC Group, Volkswagen Group, BYD , Ideal Auto, GAC Group, Deep Blue Auto, BAIC Group, Chery Automobile, Xingtu Auto, Lantu Auto, etc., as well as many Tier1, software and hardware partners. The Journey 6 series will start the delivery of the first pre-installed mass production model in 2024, and it is expected to achieve mass production and delivery of more than 10 models in 2025. At the same time, SuperDrive will reach cooperation with many top Tier1 and automobile brands in the second quarter of 2024, will launch a standard version of the mass production solution in the fourth quarter, and will achieve the delivery of the first mass production cooperation model in the third quarter of 2025.

02. Core Chi - Cabin Core X9

Smart cockpit chips are an important part of smart cars. They are responsible for processing various information inside the vehicle, including vehicle status, navigation, entertainment, etc. However, the development and manufacturing of smart cockpit chips is a very complex task with many technical difficulties. According to statistics, a high-level self-driving car can generate up to 10TB of data per second, including vehicle status, sensor data, map data, etc.

These data need to be processed quickly by the chip in order to achieve real-time decision-making and control systems . Therefore, the smart cockpit chip needs to use high-performance processors and algorithms to ensure that the data can be processed quickly. Secondly, the performance and reliability requirements that the smart cockpit chip needs to meet are also very high. In a high-speed vehicle, the chip needs to be able to respond quickly to various instructions while ensuring the accuracy and security of the data.

In addition, the smart cockpit chip needs to work closely with other systems of the vehicle. This requires the chip to have a high degree of integration and scalability, and to be able to seamlessly connect with other systems. Let's look at the development cycle and cost of the smart cockpit chip. Developing a smart cockpit chip requires multiple links, including design, verification, manufacturing, testing, etc., and the development cycle is often long. According to statistics, it takes at least 24 months to develop a high-end smart cockpit chip, and the manufacturing cost is as high as millions of dollars.

As the smart cockpit enters the 3.0 era, smart cockpit chips will continue to evolve. The Core X9 series processors are automotive-grade chips designed for the new generation of automotive electronics cockpits. They integrate high-performance CPUs, GPUs, AI accelerators, video processors, and a variety of in-vehicle scene communication interfaces , audio and video input/output interfaces, and storage interfaces, which can meet the needs of the new generation of automotive electronics cockpit applications for powerful computing and communication capabilities.

At the same time, the entire series of products have built-in high-performance HSM modules and independent safety islands, meeting the ASIL B functional safety standard and can be used in scenarios with more stringent safety performance requirements. The X9 series of products covers 3D instruments, IVI, cockpit domain control, cabin integration and other cockpit application scenarios from entry-level to flagship level. The entire series of products maintains hardware Pin-to-Pin and software compatibility.