As the degree of automobile electrification and networking continues to deepen, automobiles are gradually becoming one of the important entrances to the Internet of Things. The time and energy originally used for driving will gradually be released by advanced assisted driving, and digital smart cockpits are becoming more and more important in automobiles.
As the core hardware processing chip, it needs to handle functions including video access from multiple cameras, neural network accelerator NPU, audio processing of in-car voice, image rendering and output of multiple display screens, cellular, Bluetooth, WiFi data transmission, etc. Therefore, high-performance chips are indispensable. Among these high-performance car chips, the one with the highest exposure rate should be Qualcomm Snapdragon 8155.
But the car robot ROBO-01 released by Jidu Auto the day before yesterday (June 8) will be equipped with the latest Qualcomm Snapdragon smart cockpit chip 8295, which is also the first release of Qualcomm Snapdragon 8295 chip in China. Speaking of this, I believe many people are "black question mark". 8155 has just started to be widely used in cars, and suddenly 8295 has been released? Don't worry, the following will analyze this Qualcomm Snapdragon 8295 chip in detail.
Qualcomm Snapdragon 8295 chip, the specific chip model is SA8295P, is a product of Qualcomm's fourth-generation Snapdragon automotive digital cockpit platform. The platform was released on January 27, 2021 and is divided into three levels to support corresponding models: performance level (Performance) for entry-level platforms, flagship level (Premiere) for mid-level platforms, and supreme level (Paramount) for supercomputing platforms.
In terms of parameters, the Snapdragon 8295 chip is a 5nm automotive-grade chip (the Snapdragon 8155 chip uses a 7nm process), and its CPU part uses the same sixth-generation Kyro architecture as Qualcomm Snapdragon 888, that is, a typical large and small core architecture of one X1 super core + three A78 large cores + four A55 small cores. Qualcomm first introduced the X1 super core launched by ARM last year on the sixth-generation Kyro architecture. The single-core performance is about 50% higher than the A76 used in the 8155 cockpit chip, which was modified from the previous generation Snapdragon 855.
Qualcomm's smart cockpit chip strategy has always been to focus more on GPU and NPU, while the GPU of Snapdragon 8295 uses a newly unveiled model Adreno 690. From the name, it can be seen that the architecture of Adreno 690 has not changed. It can basically be regarded as an overclocked version of Adreno685 installed on Microsoft SQ1 (a Microsoft laptop). The GPU computing power has reached an amazing level of 3100GFLOPS, which is nearly 200% higher than 8155 and 130% higher than Snapdragon 888. Such amazing performance makes it enough to crush its peers.
According to Qualcomm's official announcement, the Snapdragon 8295 chip can support the integration of multiple ECUs and domains, including instrument panels and cockpits, AR-HUD, information and audio, rear-seat displays, mirror replacements (electronic mirrors), and in-car monitoring services. In addition, the chip also provides video processing capabilities and supports integrated driving recording and monitoring functions. At the same time, through Qualcomm's Soft SKU function for cloud services, consumers can use OTA upgrades to continuously obtain the latest features and functions throughout the life cycle of the car.
In the process of domain controllers developing towards the next generation of zone controllers, the concentration of computing power is inevitable. Processing multiple screens at the same time means more pixel computing requirements and memory usage, which is also the main reason why Qualcomm has strengthened the GPU in successive generations of cockpit chips.
In terms of NPU, the AI computing power of Qualcomm Snapdragon 8155 chip is 4TOPS, while the AI computing power of Snapdragon 8295 has increased by 7.5 times to 30TOPS. As a cockpit chip, the computing power of 8295 in terms of NPU has surpassed many driving assistance chips on the market.
In Jidu ROBO-01, the Snapdragon 8295 chip mainly drives the front-row through-screen and various intelligent interactive functions, with full offline voice, millisecond-level response, 3D human-machine co-driving map, in-car and out-car scene interaction (visual perception, voiceprint recognition, lip capture), etc. At the same time, the performance redundancy of the 8295 chip can support its smart cockpit to obtain different new functions through OTA upgrades.
Huawei is Qualcomm's rival in smart cockpit chips. What about its Kirin 990A chip? What functions can the Hongmeng smart cockpit equipped with this chip achieve?
Although the Huawei Kirin 990A chip is extremely similar to the Kirin 990 chip on Huawei P40 and other mobile phones in name, they are two completely different chips. The Kirin 990A chip uses Huawei's self-developed Taishan architecture, which is usually used in server chips and is known for its stability and high performance.
In terms of specific parameters, the Kirin 990A chip uses a 28nm process, with four cores of Taishan V120 Lite + four cores based on the Arm V7A series Cortex A55, and the GPU part uses Mali G76. In addition, Huawei has added a Da Vinci architecture computing chip to this chip, which is two D110 + one D100 large and small cores, which can provide 3.5TOPs of computing power, which is comparable to the computing power level of Snapdragon 8155, but there is still a certain gap compared with the performance of Snapdragon 8295.
Currently, all cars that support Hongmeng OS smart cockpit use Kirin 990A chip, which supports more than 30 application software, gesture control, full-scene voice, smart split screen, in-car selfie and other functions. In addition, Hongmeng OS smart cockpit can also realize interconnection with Hongmeng OS-equipped smartphones, smart watches and Huawei smart home devices.
Taking the Hongmeng car machine of the Wenjie M5 as an example, its smoothness of use is no different from that of ordinary Huawei mobile phones and tablets, and the operating logic is the same as the Hongmeng system on mobile devices. You can also see the entrance to the home interconnection function. In the future, the Hongmeng car machine will strengthen the seamless flow function between mobile phones and improve the user experience.
From the introduction of the strongest cockpit chips of the above two manufacturers, it can be seen that even though the Snapdragon 8295 chip is obviously better than Huawei's Kirin 990A in performance, the functions that can be achieved in the smart cockpit are relatively close. In that case, isn't it enough to use a chip with the performance level of Kirin 990A? The answer is correct, but not entirely correct.
Indeed, to realize the smart cockpit function, Kirin 990A and Snapdragon 8155 chips of the same performance level are sufficient, and even have certain performance redundancy. At present, the smart cockpit of automobiles is developing in the direction of high integration and high intelligence. In the short term, we can see that high-resolution and high-refresh-rate screens will be installed in the car, more screens will process information flow, and more complex in-vehicle interconnection functions will be taken into account, such as in-vehicle AI learning and subsequent OTA updates.
In other words, the cockpit chip must have the performance to ensure smooth use of current functions, and also have the performance redundancy for subsequent upgrades. This is why all the newly launched smart cars are equipped with the Qualcomm Snapdragon 8155 chip with performance redundancy, rather than the Snapdragon 820A chip. The ultra-high performance of the Qualcomm Snapdragon 8295 is prepared to cope with the development of smart cockpits in the future.
With the rapid development of automobile electrification, intelligence, networking, and sharing, the requirements for chips are getting higher and higher. Qualcomm Snapdragon 8295 chip, which is a smart cockpit chip, has caught up with mobile chips in terms of process technology. In addition to outstanding foreign players such as Qualcomm, Samsung, and Nvidia, Chinese companies such as Huawei and Horizon have also achieved considerable results in automotive smart chips. There are also Tesla and Apple, which is rumored to be building cars. The evolution of automobiles seems to have fallen into the circle of Moore's Law. I believe that in the next one or two years, the development speed of smart cars will accelerate, and the competition in smart cockpits will intensify. Higher-performance car chips will also bring new imagination space for the intelligent capabilities of automobiles.
Previous article:Changan Automobile is equipped with Chinese "chips", and domestically developed LTE Cat.1 automotive-grade chip modules are commercially available
Next article:NIO launches self-developed chassis domain controller ICC and takes the lead in mass production
- Popular Resources
- Popular amplifiers
- Car key in the left hand, liveness detection radar in the right hand, UWB is imperative for cars!
- After a decade of rapid development, domestic CIS has entered the market
- Aegis Dagger Battery + Thor EM-i Super Hybrid, Geely New Energy has thrown out two "king bombs"
- A brief discussion on functional safety - fault, error, and failure
- In the smart car 2.0 cycle, these core industry chains are facing major opportunities!
- The United States and Japan are developing new batteries. CATL faces challenges? How should China's new energy battery industry respond?
- Murata launches high-precision 6-axis inertial sensor for automobiles
- Ford patents pre-charge alarm to help save costs and respond to emergencies
- New real-time microcontroller system from Texas Instruments enables smarter processing in automotive and industrial applications
- Innolux's intelligent steer-by-wire solution makes cars smarter and safer
- 8051 MCU - Parity Check
- How to efficiently balance the sensitivity of tactile sensing interfaces
- What should I do if the servo motor shakes? What causes the servo motor to shake quickly?
- 【Brushless Motor】Analysis of three-phase BLDC motor and sharing of two popular development boards
- Midea Industrial Technology's subsidiaries Clou Electronics and Hekang New Energy jointly appeared at the Munich Battery Energy Storage Exhibition and Solar Energy Exhibition
- Guoxin Sichen | Application of ferroelectric memory PB85RS2MC in power battery management, with a capacity of 2M
- Analysis of common faults of frequency converter
- In a head-on competition with Qualcomm, what kind of cockpit products has Intel come up with?
- Dalian Rongke's all-vanadium liquid flow battery energy storage equipment industrialization project has entered the sprint stage before production
- Allegro MicroSystems Introduces Advanced Magnetic and Inductive Position Sensing Solutions at Electronica 2024
- Car key in the left hand, liveness detection radar in the right hand, UWB is imperative for cars!
- After a decade of rapid development, domestic CIS has entered the market
- Aegis Dagger Battery + Thor EM-i Super Hybrid, Geely New Energy has thrown out two "king bombs"
- A brief discussion on functional safety - fault, error, and failure
- In the smart car 2.0 cycle, these core industry chains are facing major opportunities!
- The United States and Japan are developing new batteries. CATL faces challenges? How should China's new energy battery industry respond?
- Murata launches high-precision 6-axis inertial sensor for automobiles
- Ford patents pre-charge alarm to help save costs and respond to emergencies
- New real-time microcontroller system from Texas Instruments enables smarter processing in automotive and industrial applications
- Design an electronic meter that can measure 0-1.5A current and automatically switch ranges
- Analysis of embedded C language pointers
- What does 0402_SHORT mean in the schematic diagram and how to draw the PCB
- CB140 core board voltage value difference solution
- RJ45 selection for POE powered device PD
- How are the results of CST PCB simulation displayed?
- Arrow Live: Focus on "TI FPD-Link III Automotive Chipset", the ideal solution for automotive video transmission
- Application of “C2000+TMS570” dual-chip solution in automotive electric drive functional safety
- A complete list of filter circuits, save it for future use!
- Automatic License Plate Recognition System Based on ARM Cortex