The design priorities for consumer and automotive chips differ greatly, and so do their process nodes.

Different priorities

Mobile phone chips: speed above all

Whether for phones, tablets, set-top boxes or wearables, consumer chips are designed around performance, power consumption and cost. In the smartphone era, chip performance has become a key criterion for judging a model. Multiplayer games and battle royale titles alike benefit from a stronger CPU. Qualcomm's Snapdragon 865, for example, uses one Cortex-A77 at 2.84 GHz, three Cortex-A77 at 2.42 GHz and four Cortex-A55 at 1.8 GHz, with an NPU capable of 15 trillion operations per second and an ISP processing 2 billion pixels per second, supporting a 200 megapixel camera.

Billions of transistors switching at high frequency generate significant dynamic, short circuit and leakage power. Without control this not only produces incorrect results but can even fuse parts of the circuit together, making the chip unrecoverable. Consumer chips therefore balance performance with power consumption; otherwise devices become hot, battery life falls and the user experience suffers.

As performance rises, chip prices climb and take an increasing share of total phone cost. The Snapdragon 865 costs about 700 yuan, accounting for 14 percent of the Xiaomi Mi 10 Pro bill of materials, 23 percent of the Redmi K30 Pro, 10 percent of the OPPO Find X2 Pro and 7 percent of the Samsung S20 Ultra. The Kirin 990 costs about 500 yuan, roughly 16 percent of the Huawei nova 6 price, 10 percent of the P40, 7 percent of the P40 Pro and 5 percent of the P40 Pro Plus. MediaTek's Dimensity 1000 costs 280 yuan, about 9.8 percent of the OPPO Reno3 price. Controlling chip cost is therefore essential both for product competitiveness and for profitability.

1.1 An introduction to autonomous driving compute chips

Because they are used in vehicles, automotive chips place great emphasis on reliability, safety and longevity.

Why reliability comes first. Automotive chips face several specific conditions.

First, the operating environment is harsh. Engine bay temperatures range from -40℃ to 150℃, so automotive chips must cover this wide range, while consumer chips only need to cover 0℃ to 70℃. Vehicles also experience vibration and shock, and the humidity, dust and corrosive conditions inside a vehicle far exceed what consumer chips face.

Second, automotive products are designed for a longer life. A phone has a lifecycle of three years, at most five. Vehicles are generally designed for 15 years or about 200,000 km, far longer than consumer electronics. Automotive chips are therefore required to have a product lifecycle of more than 15 years, with supply periods that can reach 30 years.

Under these conditions, maintaining chip consistency and reliability is the first concern in automotive chip design.

Safety is also especially important. Automotive chip safety has two aspects: functional safety and cybersecurity.

A phone chip that fails can simply be restarted, but an automotive chip failure can cause a serious accident, which is unacceptable to consumers. Functional safety must therefore be built into the architecture from the start. Automotive chips adopt an independent safety island design, with ECC and CRC data checking in key modules, compute modules, buses and memory, and automotive-grade processes throughout manufacturing, to ensure functional safety.

With the spread of connected vehicle technology, cybersecurity has become increasingly important. A vehicle is an always-online device, and its communication with networks and with the in-vehicle network must be encrypted to prevent intrusion. High performance cryptographic and checking modules should therefore be embedded in the chip in advance.

For functional safety, IEC has published IEC 61508, from which a series of standards for different industries has been derived, as shown below.

Finally, automotive chip design must consider longevity.

Consumer chip development roughly follows Moore's law, with a new generation every year and a new flagship phone every year. A chip only needs to meet software performance requirements for two or three years. Vehicle development cycles are much longer, and a new model needs at least two years from development to market validation, which means automotive chip design must be forward looking enough to meet customer needs three to five years ahead. As the amount of software in vehicles grows, chips must support multiple operating systems and ongoing software iteration.

Automotive-grade chips therefore have long industrialisation cycles and high barriers to entry in the supply system. Entering the mainstream automotive electronics supply chain requires meeting several basic requirements: the AEC-Q100 (ICs), AEC-Q101 (power devices) and AEC-Q200 (passive components) reliability standards issued by the North American automotive industry; ISO 26262 for automotive electronics and software functional safety; ISO 21448 for safety of the intended functionality, covering risks arising from non-systematic failures; ISO 21434 for cybersecurity, safeguarding vehicle and system network security; and IATF 16949 for zero-defect supply chain quality management. Automotive qualification of a chip generally takes three to five years, a substantial technical, production and time cost for chip makers. Mobileye took eight full years to win its first vehicle maker order, and development of Nvidia's current main chip Xavier cost 2 billion US dollars.

Different process nodes

Reducing the distance between circuits allows more transistors on a smaller chip, giving higher computing performance and lower power consumption. Process node size has therefore mattered greatly from the early micrometre era to the nanometre era. Shrinking cannot continue indefinitely, however: below about 20 nm transistors suffer quantum effects and leakage, which cancels out the benefit of a shorter gate. To solve this, Professor Hu Chenming of the University of California, Berkeley invented the fin field effect transistor (FinFET), which greatly improves channel control and reduces leakage current.

Mobile phone chip processes have progressed from 90 nm through 65, 45, 32, 28, 16, 12, 7 and 5 nm to the current 3 nm, with 1 nm on the horizon.

Conventional automotive chips did not need to chase advanced nodes as urgently, because a vehicle has plenty of space and integration density is less critical than in a phone. Automotive chips also concentrate on low compute domains such as alternators, chassis, safety and lighting control, so makers preferred mature processes. As vehicles become more intelligent, however, higher levels of autonomous driving demand high compute, which is pushing automotive compute platforms toward 5 nm and below.

AEC-Q100: the threshold for automotive chips

1. AEC-Q certification is the entry threshold for automotive electronics. AEC stands for Automotive Electronics Council. It was founded in 1994 by Chrysler, Ford and General Motors, and its membership now spans major vehicle makers, automotive electronics companies and semiconductor manufacturers worldwide. Components that comply with AEC specifications can be adopted by all three manufacturers at the same time, which encouraged part makers to exchange product characteristic data, drove standardisation of automotive parts and laid the foundation for the rapid growth of the automotive parts market. AEC-Q is the automotive reliability test standard issued by the AEC and an important passport for part makers entering automotive electronics and tier one supply chains.

AEC-Q100 was the first AEC standard, a set of stress test specifications designed for integrated circuits in automotive applications. It is important for improving product reliability and quality assurance. AEC-Q100 confirms the quality and reliability of each chip against possible conditions and latent failure states, with standardised tests of product function and performance in particular.

2. Automotive ICs required to pass AEC-Q100. Single chip ICs: control, driver, computing, memory, sensor, communication, power, power supply, security and analogue chips.

3. AEC-Q100 operating temperature grades for integrated circuits.

▸ Grade 0: ambient operating temperature -40℃ to 150℃
▸ Grade 1: ambient operating temperature -40℃ to 125℃
▸ Grade 2: ambient operating temperature -40℃ to 105℃
▸ Grade 3: ambient operating temperature -40℃ to 85℃

4. AEC-Q100 validation flow for automotive IC products.

5. AEC-Q100 test groups.

Group A, accelerated environmental stress tests (PC, THB/HAST, AC or UHAST or TH, TC, PTC, HTSL): 6 tests
Group B, accelerated lifecycle simulation (HTOL, ELFR, EDR): 3 tests
Group C, package assembly integrity (WBS, WBP, SD, PD, SBS, LI, BST): 7 tests
Group D, die fabrication reliability (EM, TDDB, HCI, BTI, SM): 5 tests
Group E, electrical verification (TEST, HBM, CDM, LU, ED, FG, CHAR, EMC, SC, SER, LF): 11 tests
Group F, defect screening (PAT, SBA): 2 tests
Group G, cavity package integrity (MS, VFV, CA, GFL, DROP, LT, DS, IWV): 8 tests

6. NTEK chip testing capability and AEC-Q100 requirements.

About NTEK

NTEK Group was founded in 2009 and is headquartered in Shenzhen. The group provides quality research and technical services covering R&D verification, inspection and testing, failure analysis, simulation and market access for intelligent connected vehicles, electronic communications and new energy.

NTEK has extensive experience in automotive-grade electronic certification and has helped more than 100 companies pass AEC-Q series certification, confirming the quality and reliability of each chip case against AEC-Q100.

Driven by the need of vehicle makers to localise automotive-grade chips and supported by the domestic semiconductor industry, NTEK provides comprehensive testing and certification services, applying AEC-Q100 automotive standards to control the safety and quality of automotive chips, supporting the growth of domestic automotive-grade chips and adding new momentum to the intelligent vehicle safety system.