In the evolving landscape of technology, the mobile phone market has reached saturation, with major players now shifting focus to the next big frontier: autonomous vehicles. Drawing from past trends in the internet era, it's clear that chips sit at the top of the industrial chain and are often the driving force behind innovation. Core chips shape the infrastructure of new computing eras.
As we step into this new age of computing, it presents a golden opportunity for emerging companies to overtake traditional leaders. While GPUs and FPGAs have been widely used, they are not specifically designed for artificial intelligence. This means they come with inherent limitations. AI chips, on the other hand, offer a vast, untapped market for both tech giants and startups alike.
Currently, deep learning and other AI applications rely heavily on general-purpose chips like GPUs, FPGAs, and ASICs, which excel in parallel processing. These components play a crucial role in accelerating AI workloads.
[Image: Behind autopilot: Intel and other chip giants' three-way battle]
In this critical domain, Intel, NVIDIA, and Qualcomm stand as the leading chip manufacturers, each vying for dominance in the AI-driven automotive industry.
**NVIDIA: Leading the Charge**
When it comes to AI chips, NVIDIA is undeniably a global leader, a position reflected in its soaring stock price—ten times higher than just two years ago. The company’s GPGPU (General-Purpose GPU) is widely recognized in the field of deep learning, with most training processes currently powered by NVIDIA’s GPUs.
This success stems from the superior computational power of NVIDIA’s GPUs compared to CPUs, as well as the ease of use in their development tools, which significantly lowers the barrier for developers.
Before the rise of self-driving cars, NVIDIA entered the automotive market through its Tegra processors, primarily for in-vehicle entertainment. However, recent models like the new Audi A8, featuring L3 autonomy, now utilize the Tegra K1 chip for surround vision. This chip delivers 350 GFLOPS of single-precision floating-point performance.
NVIDIA has since introduced more advanced platforms, such as Drive PX 2, which uses two Tegra Parker SoCs to achieve up to 8 TFLOPS of performance. However, this high performance comes at a cost—consuming up to 250W and requiring water cooling, making it expensive and impractical for mass production.
To address these issues, NVIDIA launched the Xavier chip, based on its Volta architecture. It offers 20 TFLOPS of single-precision performance while consuming only 30W, making it ideal for L4-L5 autonomous driving. Xavier is expected to start production early next year, with mass shipments planned for the fourth quarter.
**Intel: A Strategic Move**
For years, Intel lagged behind in the AI space, but it made a bold move by acquiring Altera in 2016 for $16.7 billion. This allowed Intel to integrate FPGA technology into its Autopilot platform, used by Audi’s new A8.
In March 2017, Intel further solidified its position by acquiring Mobileye for $15.3 billion. Mobileye’s EyeQ chips dominate the ADAS market, giving Intel a strong foothold in the autonomous driving sector.
Intel’s strategy combines its Atom processors, Mobileye’s EyeQ chips, and Altera’s FPGAs to create a comprehensive vehicle computing solution. With its strong presence in the server market, Intel can also provide cloud-based computing power, making it a viable player in smart city initiatives.
However, the full potential of this model depends on infrastructure development, which will take time and investment.
**Qualcomm: A Bold Acquisition**
Qualcomm has long focused on in-vehicle entertainment with its Snapdragon 602A and later the 820A, which included hardware acceleration for neural networks. Despite its energy efficiency, the chip lacks dedicated ADAS support, limiting its appeal.
To address this, Qualcomm announced its acquisition of NXP in 2016, aiming to strengthen its position in the automotive semiconductor market. NXP’s strong relationships with car manufacturers make it a valuable asset.
However, the deal remains under review by EU regulators, who raised concerns about insufficient disclosure. For Qualcomm, completing the acquisition is crucial to securing its future in the autonomous driving space.
As AI continues to evolve, the race for control over core chip architectures intensifies. Companies like NVIDIA, Intel, Qualcomm, and Google are investing heavily to secure their positions in the next computing era.
Over the next 3–5 years, breakthroughs in AI-specific chips will drive widespread adoption across industries. The intelligent market is expected to grow exponentially, far surpassing today’s mobile internet market.
Laptop DDR
Boluo Xurong Electronics Co., Ltd. , https://www.greenleaf-pc.com