新湖畔网 （随信APP) | 一体化驾驶平台中心处理器，骁龙至尊版，引领汽车未来。
文章目录[隐藏]
新湖畔网 （随信APP) | 一体化驾驶平台中心处理器，骁龙至尊版，引领汽车未来。
【微信/公众号/视频号/抖音/小红书/快手/bilibili/微博/知乎/今日头条同步报道】
在《黑神话：悟空》这款游戏中，玩家想要打出高额伤害，就需要积攒「棍势」，然后打出重击，但「棍势」不是凭空获得，需要玩家在游戏里命中攻击，蓄力和闪身才能获得。
老话就是「不积跬步，无以至千里」，俗话说就是「开大得先积攒能量」。
今年的高通骁龙峰会 2024 就有一种高通攒出了四段「棍势」然后打出的感觉，首日的骁龙 8 至尊版移动平台发布足够炸场，性能飙升；汽车平台则首次成为骁龙峰会的重磅主题之一，第一天刚说完「Snapdragon Everywhere」，次日就推出了全新骁龙座舱至尊版平台和 Snapdragon Ride 至尊版平台，让骁龙越跑越快，跑向无处不在。
一言蔽之，高通对汽车市场的重视程度前所未有，同时当下汽车市场的变局，刚好让高通再遇良机，就像当年高通随着功能手机转向智能手机的历史转变而成为科技巨头那样。
一芯双平台，高通的野心不止座舱
自骁龙汽车 602A 平台和骁龙汽车 820A 平台（也就是第一代和第二代的骁龙座舱平台）开始，高通开始进入并站稳汽车车机领域，当时智能座舱概念还不清晰，大多采用 Android 系统的车机还可以被认作是一个对安全性和稳定性要求更高的车载平板。
稍微关注这一年来汽车发布会的人就会明显感知到，骁龙 8295 出现在发布会上的频率越来越高，而这是骁龙第四代的座舱产品，它就像骁龙 8 系移动芯片出现在手机发布会的一样。
汽车不仅仅是一种交通工具，也是一种智能联网设备，不管是座舱的影音娱乐设备，还是诸多舒适性配置，都有赖于强大的座舱芯片来驱动座舱内的数个屏幕，数十个扬声器。
在安全稳定可靠之外，座舱系统的流畅与否亦是汽车产品力的重要指标，而座舱芯片是否强大，直接关系到座舱系统的流畅度。
如此背景下，在座舱芯片上已然处于领先位置的高通，推出了骁龙座舱至尊版平台来支持先进的数字化体验。
骁龙 8 至尊版移动芯片依靠高通自研的 Oryon CPU 架构实现了相比于上代产品巨大的性能飞跃，超大核的主频更是达到了移动芯片未曾触及的 4.32GHz 。同样的，骁龙座舱至尊版平台和 Snapdragon Ride 至尊版平台也采用了一样的高通 Oryon CPU 架构，相较于上一代产品，实现了 3 倍的 CPU 性能提升，同样的，GPU 性能也实现了 3 倍的性能提升，而 Hexagon NPU 实现的 AI 性能提升则高达 12 倍。
之所以把骁龙座舱至尊版平台和 Snapdragon Ride 至尊版平台放在一起说，是因为这两个平台可以分开，汽车厂商可以用骁龙座舱至尊版平台打造顶级的座舱体验，也可以用 Snapdragon Ride 至尊版平台实现高阶智能驾驶功能，也可以在同一块 SoC 上同时运行数字座舱和智能驾驶功能。
所以作为骁龙数字底盘解决方案的两个组成部分（另外两个部分是骁龙汽车智联平台和骁龙车对云服务），骁龙座舱至尊版平台和 Snapdragon Ride 至尊版平台具有相当大的灵活性。
当然，强大依然是这两个平台的第一标签，以骁龙至尊版座舱平台为例，Oryon CPU 架构能够支持多个虚拟环境和多样化的跨域应用，因为座舱系统往往会同时运行多个应用，比如导航、娱乐和通信，并且都处于台前常驻，而 Oryon CPU 架构能够保证多个应用程序同时流畅且无延迟的运行。
骁龙座舱至尊版平台中集成的 Hexagon NPU 是各模块中性能提升最多的，AI 算力是前代骁龙 8295 的 12 倍，这意味着它能够在本地处理数十亿参数的大语言模型，而不必等云端大模型的反馈。
在小红书等社交平台上经常有新手司机问汽车屏幕上的故障灯或者故障警报是什么意思，实际上这种事儿问大语言模型就行了，一些大语言模型会根据车辆维修手册进行训练，AI 助手会检索和解读这些故障灯和故障警报的意思，给出处理和维修建议。
相信不少人也注意到了，以往汽车里面也就两个屏幕，仪表屏和中控屏，但是现在汽车里面的屏幕数量正在飞速增加，除了基础的仪表屏和中控屏，还有副驾屏，后排娱乐屏；如果是 6 座车，很可能后排娱乐屏就得配四个，说不定还有投影大屏和扶手处的小控制屏，数量多之外，分辨率都从 1080P 奔着 4K、8K 的趋势去了。
不光是分辨率高了，现在汽车座舱系统的动效也越来越炫酷，能上 3D 动效就不用 2D，能高帧率动画就不搞 PPT 动画，不光一个屏幕动，所有屏幕都在动，甚至在今年骁龙峰会上，做游戏和游戏引擎的 EPIC 也来登台了，以后的座舱系统，还会用到虚幻引擎来做渲染，好看是好看了，芯片压力大啊。
比屏幕更多的是扬声器，现在国产新能源卷配置，20 多万的车，也配 20 多个扬声器，加上座舱内语音交互频率高，但是乘客很可能有几个，而且座舱内放音乐，座舱外还有路噪风噪，也对声音的处理提出了挑战。
综上要求，一般车机芯片看了都要罢工，不过既然骁龙座舱至尊版平台号称强大，不光是要搞定上面的挑战，也要给未来留出冗余。
骁龙座舱至尊版平台和 Snapdragon Ride 至尊版平台配备面向汽车应用设计的 Adreno GPU，拥有独立于 NPU 的专用图形处理器，前面说的 3D 动效和界面，还有高帧率高分辨率实时光追游戏等等都没问题。
而新平台中负责将内容传输到车辆屏幕的显示处理单元（DPU）能够支持 16 个 4K 像素的屏幕，也能混合处理不同层级和类型的数据，包括视频、图形渲染以及摄像头内容，让各种信息有序连贯呈现。
全新座舱平台支持分区音频体验，通过定位乘客的车内位置，并根据座位了解乘客需求，让车内的每个人都拥有自己专属的工作或娱乐区域。另外，新平台专用的 AI 音频处理器能够让座舱系统在各种语言、方言和背景噪音中进行高级唤醒词检测。
如果说骁龙座舱至尊版平台的进化，是进一步巩固自己在座舱平台领域的领先地位，那么 Snapdragon Ride 至尊版平台则是高通的野心所在：智能座舱在手，再图智能驾驶。
和骁龙座舱至尊版平台支持如此多的屏幕类似，Snapdragon Ride 至尊版平台也支持超过 40 个传感器，包括车外多个 1600 万像素的摄像头，以及面向乘客的 360 度全景红外摄像头，并利用神经网络实现这些传感器数据的低级别融合，然后对物体和轨迹进行稳定性检测、分类和预测。
智能驾驶的难点之一在于车辆行驶的情况极其复杂，黑夜环境、逆光环境和雨雪雾天气会影响毫米波雷达和摄像头的感知，所以 Snapdragon Ride 至尊版平台特别支持多模态传感器数据流，提升不同外部条件之下的图像效果，另外，手机相机有 HDR 功能，Snapdragon Ride 至尊版平台也有高动态范围的曝光设置来应对复杂光照环境，清晰识别物体、标识和车道标记。
激光雷达抗干扰能力强，但是获得的点云数据分辨率不够高，当然，激光雷达也朝着高分辨传感器的方向快速发展，Snapdragon Ride 至尊版平台采用了异构架构，也配有专用的处理器，能够处理激光雷达的点云数据或更低级别的数据，并实现与同一 SoC 上其他传感器的低级别融合，而不会将数据集中在单个内核中进行高负载处理。
Snapdragon Ride 至尊版平台有希望让这个功能变得又省心又省电，因为它的设计可以将智能电源管理硬件和软件相结合，平衡功耗，实现始终在线的低功耗安全监测。
另外，自特斯拉 FSD V12 版本把端到端推到行业台前，目前业界主流智驾方案都已经或者即将迁移到这个高效但神秘的技术路线当中。对于 Snapdragon Ride 至尊版平台来说，支持端到端架构是必然的，更重要的是，给端到端以及其他智驾路线留出安全冗余：内置的安全岛控制器和硬件架构可以实现隔离和无干扰运行；主控制域的 TrustZone 则被用于专门管理一些关键变量，比如转向指令或者刹车请求。
过往已经有不少智驾厂商基于 Snapdragon Ride SoC 或者 Flex SoC 等芯片做出了智驾方案，而在这次骁龙峰会上，历史最悠久的汽车厂商之一的梅赛德斯奔驰，和史上最快突破 100 万辆销量的新能源品牌理想，都宣布了将会在其未来的量产车型中采用至尊版骁龙汽车平台，这也算是这款新平台的初战告捷。
一芯不止二用，至尊版骁龙汽车平台为何能分能合？
如开头所言，骁龙攒出「棍势」并打出爆炸效果之前，已经做了非常充分的准备了，而行业变革的节奏，与之也暗暗契合。
自骁龙 602A 平台诞生，骁龙的座舱平台已历十年四代；Snapdragon Ride 平台诞生也有四年，Snapdragon Ride Flex 则是汽车行业首款同时支持数字座舱和 ADAS 的可扩展系列 SoC，提供从入门级到超级计算级别的可扩展性能。
这段时间恰好是汽车行业变革最为猛烈的十年。
汽车的电动化、智能化和网联化固然是趋势，但变化并非只有好处，比如在这个过程中，汽车的电子控制单元（ECU）越来越多，从传统的引擎控制系统、安全气囊、防抱死系统、电动助力转向、车身电子稳定系统；再到智能仪表、娱乐影音系统、辅助驾驶系统；还有电动汽车上的电驱控制、电池管理系统、车载充电系统，以及蓬勃发展的车载网关、T-BOX 和自动驾驶系统等等。
传统汽车的电子电气架构大多是分布式的，多一个功能或者模块，就可能多一个电子控制单元（ECU），最终电子控制单元（ECU）数量可能达到上百个之多，然
英文版:
In the game "Black Myth: Wukong", players need to accumulate "stick power" in order to deal high damage and then deliver heavy blows, but "stick power" is not obtained out of thin air, it requires players to hit attacks, charge up, and dodge in the game to obtain it.
As the saying goes, "a journey of a thousand miles begins with a single step", and another saying goes, "charge up before going big".
At this year's Qualcomm Snapdragon Summit 2024, Qualcomm has accumulated four stages of "stick power" and then delivered a powerful performance. The first day of the Snapdragon 8 flagship mobile platform release was explosive, with a significant performance boost; the automotive platform also became one of the highlights of the Snapdragon Summit for the first time. Just after mentioning "Snapdragon Everywhere" on the first day, a new Snapdragon Cockpit Extreme platform and Snapdragon Ride Extreme platform were launched the next day, making Snapdragon run faster and run everywhere.
In summary, Qualcomm's increased focus on the automotive market is unprecedented, and the current changes in the automotive market provide a perfect opportunity for Qualcomm, much like Qualcomm's transformation into a tech giant with the shift from feature phones to smartphones.
Dual-Platform, Qualcomm's Ambition Goes Beyond Cockpit
Since the introduction of the Snapdragon Automotive 602A platform and the Snapdragon Automotive 820A platform (the first and second generations of Snapdragon Cockpit platforms), Qualcomm has entered and established a strong position in the automotive infotainment sector. At that time, the concept of smart cockpits was not well defined, and most car infotainment systems using the Android system could be considered as a car-mounted tablet with higher security and stability requirements.
Anyone who has paid attention to automotive events in recent years can clearly see that the Snapdragon 8295 has been appearing more frequently at events, which is the fourth generation cockpit product of Snapdragon, similar to how the Snapdragon 8 series mobile chips appear at smartphone events.
Cars are not just a means of transportation; they are also intelligent networked devices. Whether it's the entertainment system in the cockpit or various comfort configurations, they all depend on powerful cockpit chips to drive multiple screens and dozens of speakers inside the cabin.
In addition to safety and reliability, the smooth operation of the cockpit system is also an important indicator of a car's product strength, and the power of the cockpit chip directly affects the smoothness of the cockpit system.
Against this background, Qualcomm, which has already established a leading position in cockpit chips, has introduced the Snapdragon Cockpit Extreme platform to support advanced digital experiences.
The Snapdragon 8 Extreme mobile chip achieves a significant performance leap compared to its predecessors through Qualcomm's self-developed Oryon CPU architecture, with the main frequency of the super-large core reaching a record-breaking 4.32GHz for mobile chips. Similarly, the Snapdragon Cockpit Extreme platform and Snapdragon Ride Extreme platform also use the same Qualcomm Oryon CPU architecture. Compared to the previous generation products, they achieve a 3x CPU performance improvement; GPU performance also sees a 3x improvement, and the AI performance achieved by the Hexagon NPU is up to 12 times higher.
The reason for mentioning the Snapdragon Cockpit Extreme platform and Snapdragon Ride Extreme platform together is that these two platforms can be used separately. Automakers can use the Snapdragon Cockpit Extreme platform to create top-notch cockpit experiences, or they can use the Snapdragon Ride Extreme platform to achieve advanced intelligent driving functions. It is also possible to run digital cockpit and intelligent driving functions simultaneously on the same SoC.
Therefore, as two components of the Snapdragon digital chassis solution (the other two being the Snapdragon Automotive Smart Connected Platform and the Snapdragon Vehicle Cloud Service), the Snapdragon Cockpit Extreme platform and Snapdragon Ride Extreme platform offer considerable flexibility.
While power remains the primary label for these two platforms, the Snapdragon Extreme Cockpit platform, for example, can support multiple virtual environments and diverse cross-domain applications through the Oryon CPU architecture. Since the cockpit system often runs multiple applications simultaneously, such as navigation, entertainment, and communication, and all of them are in the forefront, the Oryon CPU architecture ensures that multiple applications run smoothly and without delay.
The Hexagon NPU integrated into the Snapdragon Cockpit Extreme platform is the module with the highest performance increase, with AI power 12 times higher than the previous Snapdragon 8295. This means it can handle billions of parameters of large language models locally, without waiting for feedback from cloud-based large models.
On social platforms like Xiaohongshu, new drivers often ask what the fault lights or alerts mean on the car screen. In fact, you can just ask a large language model about these issues. Some large language models are trained based on vehicle maintenance manuals, and AI assistants can search for and interpret the meanings of these fault lights and alerts, providing handling and repair suggestions.
It's also worth noting that in the past, cars typically had only two screens, the instrument panel and the central control screen. Now, the number of screens inside cars is rapidly increasing, with additional screens like the co-driver's screen, rear entertainment screens, and in a 6-seater vehicle, there might be four rear entertainment screens, possibly even a projection screen and a small control screen at the armrest. Not only has the number of screens increased, but the resolution has also jumped from 1080p toward 4K and 8K trends.
It's not just about higher resolution; the dynamic effects of car cockpit systems are becoming increasingly cool. If it can be in 3D motion, there's no need for a 2D interface, and if high frame rate animation is possible, avoid using PowerPoint-style animations. Every screen is moving, and at this year's Snapdragon Summit, EPIC Games, a game and game engine developer, made an appearance. In the future, cockpit systems may also use Unreal Engine for rendering, which not only looks good but also puts a lot of pressure on chips.
More important than screens are the speakers. In a domestic new energy vehicle configuration, even a car costing over 200,000 yuan will come equipped with over 20 speakers, and with high-frequency voice interactions inside the cockpit, but passengers may have several, and with music playing inside the cabin and road noise outside, it also poses a challenge for sound processing.
With all these requirements, most car chips are likely to go on strike. However, since the Snapdragon Cockpit Extreme platform is said to be powerful, it not only needs to handle the aforementioned challenges but also allow for future redundancy.
The Snapdragon Cockpit Extreme platform and Snapdragon Ride Extreme platform are equipped with an Adreno GPU designed for automotive applications, with a dedicated graphics processor separate from the NPU. The 3D effects and interfaces, as well as high-frame-rate real-time ray tracing games are no problem.
The display processing unit (DPU) in the new platform can support 16 4K pixel screens, as well as mix and process data of different hierarchies and types, including video, graphics rendering, and camera content, ensuring a coherent presentation of various information.
The new cockpit platform supports partitioned audio experiences by locating passengers inside the car and understanding their needs based on their seating positions, allowing each person in the car to have their own exclusive workspace or entertainment area. In addition, the dedicated AI audio processor in the new platform allows the cockpit system to perform advanced wake-word detection in various languages, dialects, and background noises.
If the evolution of the Snapdragon Cockpit Extreme platform aims to further consolidate its leading position in the cockpit platform field, then the Snapdragon Ride Extreme platform embodies Qualcomm's ambition: with intelligent cockpit in hand, aim for intelligent driving.
Similar to the support for multiple screens from the Snapdragon Cockpit Extreme platform, the Snapdragon Ride Extreme platform also supports over 40 sensors, including multiple 16-million-pixel cameras outside the car and a 360-degree panoramic infrared camera facing passengers. It utilizes neural networks to achieve low-level fusion of data from these sensors, and then performs stability detection, classification, and prediction of objects and trajectories.
One of the challenges of intelligent driving is the extremely complex driving conditions, where dark nights, backlighting, rain, and snow can affect the perception of millimeter-wave radar and cameras. Therefore, the Snapdragon Ride Extreme platform specifically supports multimodal sensor data streams to improve imaging results under different external conditions. Additionally, just as mobile cameras have HDR functions, the Snapdragon Ride Extreme platform also features high dynamic range exposure settings to address complex lighting environments and accurately identify objects, signs, and lane markings.
While LIDAR has strong anti-interference capabilities, the resolution of the point cloud data obtained is not high enough. However, LIDAR is also rapidly evolving toward high-resolution sensors. The Snapdragon Ride Extreme platform uses a heterogeneous architecture and a dedicated processor to process LIDAR point cloud data or lower-level data and achieve low-level fusion with other sensors on the same SoC, without concentrating on high-load processing in a single core.
Many bloggers have previously shared the benefits and troubles of the "sentinel mode", such as being able to continuously record vehicle environment information and provide evidence for incidents like malicious scratches or accidental collisions. However, this mode consumes a considerable amount of power; if used all day, it's equivalent to consuming five to six kWh of electricity. When used for long periods in public parking lots, this would cost nearly 100 yuan per month in electricity fees and also affect the vehicle's range.
The Snapdragon Ride Extreme platform has the potential to make this function both worry-free and energy-saving. Its design combines intelligent power management hardware and software to balance power consumption, achieve low-power security monitoring constantly, and stably operate.
Moreover, since Tesla's FSD V12 version brought end-to-end to the forefront of the industry, mainstream intelligent driving solutions are already moving or about to migrate to this efficient but mysterious technological path. For the Snapdragon Ride Extreme platform, support for end-to-end architecture is essential, and more importantly, providing safety redundancy for end-to-end and other intelligent driving paths: the built-in safety island controller and hardware architecture can achieve isolation and interference-free operation, while the TrustZone of the main control domain is used to manage key variables such as steering commands or braking requests.
Numerous intelligent driving companies have already developed intelligent driving solutions based on Qualcomm's Snapdragon Ride SoC or Flex SoC chips. At this year's Snapdragon Summit, Mercedes-Benz, one of the oldest car manufacturers, and NIO, the new energy brand that has surpassed 1 million units in sales faster than any other, both announced that they will adopt the Snapdragon Automotive Extreme platform in their future mass-produced models, marking a successful start for this new platform.
Single-Chip, Dual-Purpose, Why Can the Snapdragon Automotive Extreme Platform Divide and Merge?
As mentioned at the beginning, Qualcomm has made very thorough preparations before accumulating "stick power" and delivering explosive effects, and the pace of industry transformation also quietly aligns with it.
Since the birth of the Snapdragon 602A platform, Snapdragon's cockpit platform has undergone ten years and four generations. The Snapdragon Ride platform has also been around for four years, and the Snapdragon Ride Flex is the first series of SoCs in the automotive industry that simultaneously supports a digital cockpit and ADAS, offering scalable performance from entry-level to supercomputing level.
This period coincides with the most intense transformation of the automotive industry.
While the electrification, intelligence, and networking of cars are the trends, the changes are not without drawbacks. For example, in this process, there are more and more electronic control units (ECUs) in cars, ranging from traditional engine control systems, safety airbags, anti-lock braking systems, electric power steering, vehicle electronic stability control systems; then to smart dashboards, entertainment audio systems, advanced driving assistance systems (ADAS); and even the electric drive control, battery management systems, vehicle charging systems on electric vehicles, and the burgeoning vehicle gateway, T-BOX, and autonomous driving systems, and so on.
The electrical and electronic architecture of traditional cars is mostly distributed, with each additional function or module possibly leading to the addition of another electronic control unit (ECU), with the eventual number of ECUs potentially reaching well over a hundred, connected through Controller Area Network (CAN) and Local Interconnect Network (LIN) buses.
To address the various problems of distributed electrical and electronic architectures, many car companies have started to improve their architectures and have moved towards making functional domain controllers the norm.
The so-called architecture based on functional domain controllers consolidates ECU functions that are similar but distributed onto a hardware platform that is more robust than a single ECU, known as the "Domain Control Unit" (DCU).
Platformization, high integration, high performance, good compatibility, rich hardware interface resources, and powerful software features allow DCUs to handle multiple aspects, simplify the electrical and electronic architecture, and reduce development and manufacturing costs.
Generally, DCUs fall into five main categories: Powertrain Domain, Chassis Domain, Body/Comfort Domain, Cockpit Domain, and ADAS Domain.
Some car companies go a step further and consolidate domain controllers into "Vehicle Domain Controller (VDC), ADAS Domain Controller (ADC)," and Cockpit Domain Controller (CDC)," forming a three-domain architecture.
The evolution of the automotive electronic and electrical architecture can be summarized as transitioning from "Working Alone," to "Feudal Seclusion," to "Central Supercomputer + Domain Control." This is why the new Snapdragon Cockpit Extreme platform and Snapdragon Ride Extreme platform can be separated and used together, one of the reasons being that different carmakers have different electronic and electrical architectures in their cars, different technology roadmap choices, and different definitions of domain controllers for cars, requiring flexible technical solutions to meet complex customer demands.
When independent, the Snapdragon Cockpit Extreme platform and Snapdragon Ride Extreme platform handle the intelligent cockpit domain controller (CDC, Cockpit Domain Controller) and intelligent driving domain controller (ADC, ADAS Domain Controller) independently, which is the current mainstream solution in the industry.
Additionally, as mentioned earlier, platformization and high integration are key trends. While the division of three domains may be simple enough, moving boldly towards an integrated cockpit and driving system is also possible. This is where the significance of Qualcomm embedding a dual-platform in the Snapdragon automotive platform lies, embracing the concept of a car central supercomputer, similar to how phone chips are consolidated into SoCs, concentrating CPU, GPU, NPU, modem, and other components into a unified architecture, but more complex.
The Central Supercomputer contains two platforms, not only for dual-use but also likely to run several operating systems. This requires an architecture with heterogeneous processing to address different systems or applications while maximizing performance.
The Snapdragon Automotive Extreme platform also incorporates Qualcomm Technologies' comprehensive software stack, supporting hardware virtualization through the Type-1 Secure Management Monitor program. This management program supports multiple client VMs with unique operating systems, enabling them to run independently and undisturbed.
In an era where everything can be OTA, the Snapdragon Automotive Extreme platform naturally aims to facilitate the industry's transition to software-defined cars, ensuring a unified software framework for upgradeability. It also aims to allow automakers to accelerate feature development through cloud-based workstations to simplify software development processes for continuous improvement and faster market launch of new features and services.
In short, the Snapdragon Automotive Extreme platform brings fast upgrades, quick error correction, and rapid feature additions to the cockpit and intelligent driving systems.
When the Snapdragon Automotive Extreme platform appeared at the Snapdragon Summit, it feels more like a perfect fit than a newcomer—a journey of Qualcomm beyond just smartphones, tablets, and PCs, also extending to VR and AR, and now, entering the automotive sector that faces a century-old transformation. Qualcomm is a driver and participant in this transformation, also aspiring to be one of the guiding lights in this change.
Author: 刘学文
In the greenhouse of destiny, where every cabbage sprayed with too much pesticide once dreamed of becoming organic and pollution-free organic vegetables.
Contact: Email, Sina Weibo
Follow us: 爱范儿 |
Original Article |
[Comments](https://www.ifanr.com/1603656
舱驾一体，中央超算，至尊版骁龙汽车平台想把汽车的未来「集中」起来
#舱驾一体中央超算至尊版骁龙汽车平台想把汽车的未来集中起来
关注流程：打开随信App→搜索新湖畔网随信号：973641 →订阅即可！
公众号：新湖畔网 抖音：新湖畔网
视频号：新湖畔网 快手：新湖畔网
小红书：新湖畔网 随信：新湖畔网
百家号：新湖畔网 B站：新湖畔网
知乎：新湖畔网 微博：新湖畔网
UC头条:新湖畔网 搜狐号：新湖畔网
趣头条：新湖畔网 虎嗅：新湖畔网
腾讯新闻：新湖畔网 网易号：新湖畔网
36氪：新湖畔网 钛媒体：新湖畔网
今日头条：新湖畔网 西瓜视频：新湖畔网
