High-Performance MCU: Bridging the Last Mile to Intelligent Vehicles

This MCU breaks through the ceiling of intelligent vehicles.

Author | Hao Han

Editor | Xiao Han

Do you think cars today are smart enough?

Indeed, under the wave of development in intelligent connected vehicles, significant improvements in intelligence have been achieved in recent years. On one hand, the arms race in computing power for intelligent driving is becoming increasingly fierce; on the other hand, human-machine interaction in the cockpit is becoming richer, with face recognition, voice interaction, gesture interaction, and other configurations gradually becoming standard.

Despite this, the intelligence of vehicles is still in a fragmented state. What is hindering the implementation of advanced intelligence?

▲ A new car equipped with a high-performance computing platform

01.

Urgently seeking a “breakthrough” in intelligent vehicles

For car users, what is true intelligence? Let’s imagine a scenario.

During the rainy season before a heavy rain, the driver smokes inside the car while enjoying the coolest breeze before the storm by opening the window.

When the driver extinguishes the cigarette and prepares to drive in the heavy rain, they will need to close the windows and sunroof, turn on the air conditioning’s external circulation, and possibly turn on the fog lights or windshield wipers. In complex road conditions and bad weather, this creates more safety hazards for passengers.

However, if the vehicle could recognize that the driver has extinguished the cigarette and, based on the current weather conditions and probability of rain, automatically complete the aforementioned operations without distracting the driver, this might truly be called an intelligent vehicle.

In this case, the role of the car, in the words of Zhang Qiang, chairman of Xinchih Technology, is more like a wise, warm, safe, and reliable “car partner”.

▲ The in-car camera assesses the driver’s status

However, the above scenario is still very difficult to achieve at present. With the existing high-performance chips and perception algorithms, it is basically possible to recognize the actions of passengers in the cabin, and through the vehicle’s networking capabilities, it can also obtain weather information.

However, in the current distributed electronic and electrical architecture, each control unit is controlled by a single MCU communicating via the CAN bus. If car manufacturers want to add a new function, they need to add a new control unit, which increases the complexity and cost of the system and makes OTA upgrades difficult.

Therefore, to truly realize the above application scenarios, a centralized electronic and electrical architecture that can centrally schedule the cockpit system, intelligent driving, and MCU control units becomes crucial.

It is well-known that the centralized computing + regional control electronic and electrical architecture is an acknowledged trend. Under this architecture, not only is a central computer needed to provide a smart brain, but the originally non-communicating MCUs also need to communicate with each other, coordinating through complex communication protocols and middleware, allowing the execution side to become smarter and work together to achieve increasingly personalized travel scenarios.

Thus, if vehicles are to achieve true high-level intelligence, the regional and terminal MCUs are important pieces in realizing intelligent vehicles and must have a substantial performance improvement to help manufacturers truly achieve “turning fragments into wholes”.

Recently, an innovative automotive chip semiconductor company from China, Xinchih Technology, has bridged the last mile to achieving intelligent vehicles by launching a high-performance, high-reliability, and high-safety MCU chip, transforming cars from mere “transportation tools” to wise “car partners”.

▲ Centralized electronic and electrical architecture (top), distributed electronic and electrical architecture (bottom)

02.

The industry needs high-performance MCUs to make advanced intelligence possible

On April 12, Xinchih Technology launched its new automotive-grade MCU product – the E3 series, designed based on ARM Cortex-R5F, with a maximum CPU frequency of up to 800MHz, utilizing TSMC’s 22nm automotive-grade process.

Currently, most automotive-grade MCU chips have a CPU frequency of 100MHz, and MCUs with a CPU frequency of 200MHz or 300MHz are already considered high-performance MCU chips in the industry. Compared to these MCU chips, the performance of the MCU launched by Xinchih this time has achieved several times improvement.

▲ Xinchih Technology E3 MCU chip

Xinchih’s chief architect, Sun Mingle, stated at the launch that a single 800M CPU core can precisely monitor the status of 40 to 60 battery cells if used for BMS; for motor control, it can perform high-precision closed-loop control for 4 motors simultaneously; and if used for gateway routing, it can support real-time data exchange between 24 CANFD, 16 LIN, and 2 Gigabit Ethernet.

Moreover, the MCU is equipped with 6 such CPUs, of which 4 can be configured to run in dual-core lockstep or independently.

It can be said that Xinchih, through the E3 series MCU, has raised the performance standards of high-performance MCU chips in the industry significantly, redefining the meaning of “high performance” in MCU chips.

In addition, as a control chip, safety is of utmost importance for vehicles.

Zhang Qiang, chairman of Xinchih Technology, stated that the fault tolerance rate of automotive-grade chips is a fraction of that of consumer-grade chips. In a million-chip scale, the error probability allowed for consumer chips is 300 to 500, while the requirement for automotive-grade chips is infinitely close to 0, as it involves personal safety, where a single error can lead to fatal safety hazards.

▲ Zhang Qiang, chairman of Xinchih Technology

In fact, adhering to automotive-grade standards has been deeply ingrained in Xinchih’s brand DNA. From the moment of its establishment, Xinchih completed the highest functional safety level process certification of ISO26262 ASIL-D, and soon obtained AEC-Q100 reliability certification, ISO26262 functional safety product certification, and national secret certification, becoming the first automotive-grade chip enterprise in China with “four certifications in one”.

The newly launched Xinchih E3 series MCU chip has achieved Grade 1 level in automotive reliability standards AEC-Q100, and its functional safety level has reached ASIL-D (the highest safety level in functional safety certification), ensuring stable operation of vehicles.

Looking at the current automotive-grade MCU chip market, there are very few products globally that can meet both standards, which is also a high mountain that other competitors find it difficult to surpass.

It is understood that to achieve ASIL-B level product functional certification generally means an increase of 30% in cycle time, while achieving ASIL-D level product functional certification means nearly doubling the cycle time. The difficulty behind this is self-evident.

In addition, the Xinchih E3 also has a wide range of application scenarios, suitable for BMS, brake control, steer-by-wire chassis, ADAS, HUD, LCD instruments, and more.

For example, in the most critical battery safety of current smart electric vehicles, most MCUs in BMS (Battery Management System) generally have weak performance, unable to detect each battery cell, only able to perform sampling, resulting in overall inefficiency. However, with a high-performance MCU, each battery cell can be monitored and feedbacked quickly, predicting thermal runaway risks in advance, thereby enhancing battery and vehicle safety.

It is reported that the Xinchih E3 series MCU chip is expected to achieve mass production in the third quarter of this year. In the context of global semiconductor supply shortages, Xinchih’s MCU can be said to ride the wave, bringing more imaginative space for intelligent vehicles.

▲ Xinchih Technology E3 MCU chip

03.

Covering the Backbone Architecture of Intelligent Vehicles, Xinchih Takes a Unique Innovative Path

In the fiercely competitive chip race, Xinchih has always been a special presence. Unlike other chip companies that focus solely on cockpit chips, vehicle control chips, or autonomous driving chips, Xinchih chose a less traveled path from the beginning – a comprehensive chip solution for future vehicle electronic and electrical architecture.

With the launch of the E3 series MCU chip, Xinchih has formed a product matrix covering smart cockpit, autonomous driving, central gateway, and high-performance MCU, covering the most core application areas of intelligent vehicles.

On the day of the launch, Xinchih also introduced the future-oriented “Central Computing Architecture SCCA 1.0”, which not only supports safe and reliable task deployment but is also crucial for helping car manufacturers achieve flexible system expansion.

▲ Xinchih’s four major product series

Omni-scenario and platform-based design have always been the core logic of Xinchih’s R&D. This logic directly benefits from additive empowerment, allowing for efficient joint R&D with clients, connecting every application scenario, and ultimately achieving intelligent vehicles.

It is understood that in terms of R&D, Xinchih’s various product series adopt platform-based design, with over 60% reusability of software and hardware among the cockpit, autonomous driving, gateway, and MCU products, significantly improving R&D efficiency, achieving resource reuse, and saving costs and time.

Zhang Qiang, chairman of Xinchih Technology, stated that the integrated design of platform-based architecture not only significantly reduces R&D costs and time investment but also rapidly enhances the supply chain flexibility of car manufacturers, alleviating the risk of “chip shortage”. This can actually be evidenced by Tesla’s sales performance during the chip shortage in 2021.

According to Tesla’s second-quarter financial report for 2021, it still achieved increased production capacity under the global chip shortage environment.

The main reason behind this is that Tesla’s electrical and firmware teams were designing, developing, and verifying new MCUs and upgrading the vehicle’s electronic and electrical architecture, reducing the use of wiring harnesses and chips, and avoiding the impact of chip shortages to a certain extent through further integration. This is also a core advantage for Xinchih in terms of supply chain security in the future.

It is worth noting that although the automotive chip market is becoming increasingly crowded, there are very few with complete mass production experience of automotive-grade chips.

The R&D team led by Xinchih Technology CEO Qiu Yuqing is one of the few domestic teams with large-scale automotive-grade chip mass production experience, also gathering talents from the automotive electronic and electrical architecture field and consumer electronics field.

In terms of hardware design and software development, they have a deeper understanding of domestic users’ needs, thus creating automotive-grade chip products more suitable for the Chinese market and users.

▲ Qiu Yuqing, CEO of Xinchih Technology

Qiu Yuqing once mentioned in an interview: “From day one, Xinchih has not aimed at individual point intelligence, but at the future application scenarios of intelligent vehicles, which is an inseparable complete picture.” With the gradual rollout of the four major product series, it can be said that Xinchih has taken a unique innovative path.

Behind the rapid development of Xinchih Technology, we also see the perseverance and breakthroughs of Chinese automotive people, steadily and solidly walking every step on the slow track of automotive-grade chips, while daring to break shackles, empowering future intelligent travel with innovative business layouts, truly assisting the Chinese intelligent automotive new era to enter a new realm.

High-Performance MCU: Bridging the Last Mile to Intelligent Vehicles

Understanding the Relationship Between Micro:bit and Arduino

Getting Started with the Arduino ESP32 Core Library

Implementing Neural Networks on a 10-Cent RISC-V MCU Without a Multiplier

Getting Started with Arduino: Light Up Your First LED from Scratch

How Highly Integrated Embedded Processors Drive the Development of Industrial Robots

Leave a Comment Cancel reply

Menu