6 minutes and 37 seconds, the brand new Xpeng P7 received over 10,000 pre-orders. Of course, placing an order only requires 99 yuan, and it’s refundable if not desired.
Reviewing the past year’s sales data of pure electric sedans in the 200,000 yuan range, Xiaomi SU7 and Tesla Model 3 topped the sales charts, far outpacing other competitors. This proves that to successfully sell a car, outstanding product capabilities are just the beginning.
He Xiaopeng stated that when the Xpeng P7 was launched in 2020, it immediately became one of the top three in the market over 200,000 yuan. By 2025, he still expects the new P7 to remain a “top three”.
What product strengths is the new P7 using to captivate people?
- 0-100 km/h acceleration in as fast as 3.7 seconds, braking distance from 100 km/h to zero is 33.16 meters, top speed is 230 km/h.
- Standard dual-chamber air suspension, supporting 50mm height adjustment.
- All models offer long range, with long range, extra-long range, and high performance being 702 km, 820 km, and 750 km respectively.
- Equipped with 800 V SiC platform and 5C batteries across the board, charging up to 525 km in just 10 minutes.
- Features Ultra Intelligent configuration, including three Turing chips, delivering 2,250 TOPS of computing power.
More importantly, the new P7’s design that prompts a “wow” reaction provides this car with a unique competitive edge.
Last night, after the first public appearance of the new P7, media outlets including the No. 42 Garage jointly interviewed Xpeng Motors Chairman and CEO He Xiaopeng, Xpeng P7 model product head Nick, and Xpeng Motors Design Center General Manager Wang Tan. This interview offered deeper insights into the strategies of the new P7 and design considerations. We have compiled the entire interview into a text version for reference.
Q: Why is the P7 considered to have entered the second generation platform?
He Xiaopeng: I think the P7 will remain our icon for a considerable time and our best-selling car, especially in 2021 and 2022. Currently, automotive cycles are much faster than before; previously, many companies used one car per platform for seven years, now I think it’s more like two and a half years for a generation. I believe five years is a major cycle.
For example, the original P7 had only 30 TOPS computing power, and there were very few cars with over 600 km range back then. Now, models with over 800 km range are stronger, without 800 V or 5C at that time.I look forward to the P7 once again becoming the totem of Xpeng; and re-entering the top three above 200,000. I believe it used to be one strong player, then Xpeng’s P7 and BYD’s Han entered to form the top three. With the launch of the new P7, I believe it has re-entered the top-three system, which I am very excited about. Whether it’s design, performance, assisted driving, or intelligent cabin, it surpasses the previous generation by far.
Q: Where is the greatest differentiation of the brand-new Xpeng P7?
Henry Xia: We aim to be both aesthetically pleasing and differentiated, which is very difficult to achieve simultaneously. Often, looking good doesn’t offer differentiation, and standing out isn’t aesthetically appealing, which is really challenging.
At the same time, when our highest computing power becomes the strongest differentiation is extremely important. The internal goal I set is for Xpeng’s smart assisted driving to be several times better than the current version, achievable in the near future, and I’ll share more then.
At the beginning of 2020, our executives first tested the P7’s highway intelligent assisted driving, and many quietly decided to purchase this configuration afterward, whereas they previously didn’t buy the version with assisted driving; they didn’t believe in it. In just five years, today’s new energy vehicles’ intelligence, not just in China but also in the global market, has seen a revolutionary change. I believe both China and the U.S. lead in smart assisted driving, and even in autonomous and driverless driving, we will be able to fully overcome the technical challenges in the next five years, addressing cost, safety, and regulatory issues afterward. This indicates that higher-level assisted driving will emerge.
Q: Has there been any advance stocking for the new P7, and what about production capacity?
Henry Xia: Our modularization is improving, such as in cell modularization and the Ultra smart driving, which is already on another vehicle. The P7 will see some cars produced in advance to sell as inventory when delivering.
The previous order model made it difficult to accurately predict user preferences. Now we first broadly understand user preferences in advance, and second, sell inventory cars that can’t have their configurations chosen. New car forces and traditional carmakers differ significantly. In a dealership model, dealers hold stock, but we don’t. Some (brands) have inventory periods of 2, 3, or even 5 months; ours is about 0.35 months. This cycle aims to minimize the risk to dealers, the supply chain, and ourselves, though it makes the first batch delivery very challenging. We will progressively grow and learn.
Q: How do you differentiate between the new P7, P7i, and P7+ products?
Henry Xia: I consider both the P7 and P7i as the first-generation P7. Today, we’ve launched a new generation of P7, which can practically be regarded as the next generation of P7i. Both P7+ and P7 belong to the P7 family, one focusing on family, and the other on aesthetics, sportiness, and individuality, but both vehicles are very tech-savvy. The P7+ places more emphasis on space and comfort, while the P7 focuses more on aesthetics and sportiness.
Q: How would you persuade users to choose P7 over SU7?
Henry Xia: I believe both Xpeng’s new P7 and SU7 are excellent vehicles, worthy of everyone’s attention.“`markdown
Q: Why does Xpeng use both VLM and VLA?
He Xiaopeng: I think we’ll see more similar architectures emerge next year, primarily due to two reasons. First, Xpeng’s in-house development of computing power considers multiple scenarios and product lines, which is a significant differentiation as most companies develop in-house computing mainly for AD and smart driving without addressing other issues.
Second, as I mentioned earlier, internally I demand VLA to be a safer driver and VLM to be a smarter housekeeper. Today’s computing power seems ten times higher than before, and in 5-10 years, it is expected to be over ten times more powerful than now. Therefore, not only Xpeng but also our American counterparts will use this enhanced computing power.
The transformation brought by this computing is a substantial leap, not merely a percentage increase. It happens so rapidly that you can see the change within months. The value of computing power will become evident to all car manufacturers by next year, at the latest the year after. We are determined to be pioneers.
Q: What is the progress of Xpeng’s Robotaxi?
He Xiaopeng: We’re getting there soon.
Q: Have the issues of self-developed chip yield and automotive-grade certification been resolved?
He Xiaopeng: We completed tape-out last August and the chips were installed in actual cars by October. By this month, in August, we expect a limited user experience for G7 customers. The hardware issues were resolved months ago, and now we’re tackling software.
I firmly believe that the value of our self-developed computing power is immense. Some American competitors achieve impressive outcomes with less computing power, thanks to factors beyond the models alone. Thus, the self-developed computing power magnifies the value of third-party computing multiple times. The deeper you delve into it, the more challenging yet rewarding it becomes. This is why I am confident that Xpeng’s smart driving will lead significantly in the near future.
Q: Has there been a market shift in recent years regarding consumer demands, and what’s the order of importance?
He Xiaopeng: In mid-to-high-end car brands, the first priority used to be different. Previously, for cars sold at lower prices, cost performance was the top priority.
Initially, we prioritized technology; however, from this year onwards, aesthetics take precedence. Before, beauty took a backseat to technology. Now, if aesthetics and technology “quarrel,” aesthetics takes precedence. They are equally important, but aesthetics lead by a slight margin.
I believe safety is paramount. Xpeng is committed to not only hard and soft integration research but also cross-domain integration, with proactive safety becoming increasingly crucial.
Moreover, why pursue ultra-high computing power? For instance, achieving safe highway assisted driving at 200 km/h is unsafe with many current solutions. To provide a safe highway driving aid at such speeds requires significant computing power. All Xpeng products, including our Ultra, consider global operations. Thus, in my opinion, safety quality is foundational.
“`Design, technology, brand, and indeed business acumen are crucial. You may excel in all areas, but if everything is overpriced, it becomes challenging.
Designing cars truly isn’t a job for the faint-hearted; it’s a never-ending marathon, likened to an enormous barrel that requires hundreds of staves, where each must be impeccable, supplemented by proficiency in specific areas. I consider it a high-tech research endeavor; styling is not merely art but scientific exploration.
Q: Why didn’t you design the P7 to mimic a “Baoshishi”?
Wang Tan: Throughout our design process, we ponder aesthetics deeply. In a recent video, I mentioned that Chinese people and Chinese cars can indeed be original. I believe aesthetics transcend beauty; it involves design philosophy and profound thought. A design is an integral part of a product and must align with the pioneering spirit of the brand capable of producing such a product. We strive to innovate and lead with technology and creativity for our consumers’ benefit.
With these reflections, every action embodies our design philosophy. You’ll see that the latest Xpeng P7 harmonizes left-hand technology with right-hand art—melding emotion with rationality.
During design, we focus heavily on subtraction. Adding elements is easy but lacks longevity; the challenge lies in effective subtraction. The sleek surfaces highlight dynamic tension, conveying immense emotional value.
The true fusion of emotion and rationality constitutes our design philosophy for the P7. Originality and other considerations emerge from these logic layers.
Q: Does the P7’s interior resemble the SU7?
Wang Tan: In design, imagination runs wild. However, when narrowing down to a solution, it’s key to consider the real value to users. For instance, the exterior should communicate our technological prowess and design philosophy, forming the first impression.
He Xiaopeng advised me that the exterior must evoke an “Wow” upon the first glance, not merely look good.
As for the interior, you will spend years with this car. How the interior conveys its purpose—be it domestic or sporty—is paramount. It should also stand the test of time, enduring as a companion for 10, 20 years, or longer.And speaking of the steering wheel, I’ve noticed quite a bit of online chatter comparing this steering wheel to that of certain other models. It’s a very classic T-type three-spoke steering wheel, complete with satellite buttons. Essentially, in the dynamic context of racing, this is designed for the most convenient switching of driving modes. It’s a function-oriented feature that we’ve successfully blended with our design.
Q: NIO L90 was recently launched and has been quite successful in deliveries. In what aspects does the new Xpeng P7 reflect the changes in Xpeng Motors over the past few years?
He Xiaopeng: Congratulations to NIO for their excellent work on the L90, but I believe this market competition is just beginning.
We are swimming in a sea of blood, not even the red sea yet. Whether it is Xpeng or others, I think many automakers are striving hard. We aim to elevate the P7 further, enhancing its overall design, performance, technology, details, quality, and safety to an entirely new level. I believe Xpeng’s comprehensive capability has significantly improved compared to five years ago.
Often, a company may easily focus only on its strengths, but I think the other two areas are equally crucial. One is the shortcoming, as per the barrel theory, the shortest plank is the most alarming. The second is the average level where some companies might define it as 70 points, others 75 or 80, but I believe it’s essential to raise this average level significantly. For the new Xpeng P7, my goal is to have no short planks, a high average level, and several areas of exceptional strength.
Q: What do you consider Xpeng’s core advantages to be at this moment?
He Xiaopeng: Internally, I hope we do four things well over the next two years, and that pace is termed “steady progress,” meaning moving slowly to lay a strong foundation and enhance quality.
I believe the four directions are, firstly, technology, secondly, achieving organizational excellence, including team efficiency and group efficiency. Thirdly, mastering commercialization, which involves how to save money, how to spend it, and how to earn it. Lastly, developing a comprehensive global capability, which is quite challenging. Many people say China’s new energy vehicles are outstanding and ready to take on the global stage. In reality, if you assess the globe, Chinese cars are merely much stronger than before and still need to earnestly learn for another 10 or even 20 years.
These are the four capabilities we are striving towards, not necessarily our strengths yet but the directions we hope to excel in. Thus, we are on this journey, continuing our efforts.
Q: How do you balance innovation speed with safety responsibilities?
He Xiaopeng: I think the new car manufacturers are tackling this. Traditional car manufacturers focused predominantly on hardware quality, but now it’s about software quality, hardware quality, and global quality.
Balancing these three will challenge us all. Why are we pursuing steady progress? We spent five years building a foundation before truly entering Europe. The first four years were purely groundwork, so a steadier approach translates to greater safety.
Q: In your view, should design pursue mass appeal or niche appeal?## Wang Tan
Design isn’t simply an art piece; it’s meant to be something appealing and cool for the majority. However, designers must be forward-thinking. We need to consider how people will perceive the product two years from now. In other words, we innovate one step ahead. Going two steps might make it an artwork rather than a well-designed product.
Q: How was the new P7’s design conceived?
Wang Tan:
When we contemplated designing the new P7, it posed a significant challenge. The previous P7 stood out because it followed conventional design language and philosophy. We asked ourselves which design elements of the P7 should continue. But that’s not the answer for the new P7’s design. The real answer, I believe, is to elevate it—to embody the pioneering spirit the P7 had when it first hit the market.
As explorers of future mobility, exploration involves trying new things and diversifying design. If today’s design blends cautious, conservative elements, it’s not innovative, creative, or original. When designing the new P7, we indeed inherited the P7’s pioneering spirit.
Q: What are your plans for technological intelligence?
He Xiaopeng:
Today, our road rage emoji is an emotional expression fueled by technology. First, it needs an AR-HUD, then a high computing power. To further expand its capabilities, it requires superior software integration.
Traditional car companies rarely developed full stacks, and future car companies won’t suffice with just full stacks—they need cross-domain integration. The primary valuable aspect is called functional, performance, or safety value. However, emotional value is also significant. This cool feature might seem trivial but can significantly enhance your mood, creating a distinctly different experience.
Q: What is your view on homogenized design?
Wang Tan:
Indeed, styling designers are like dancing with shackles. But I believe there are always opportunities in automotive design for innovation. Finding innovative points in design combinations is essential and should be a skill every designer possesses.
Whether from different companies or design styles, good design can genuinely resonate and be perpetuated. Still, design also evolves and stays contemporary. Some good, new designs will iterate in the next generation of models. Regarding the clone phenomenon, it simply reflects the diverse strategies car companies use to succeed in selling their products.
Q: Earlier, Xpeng focused more on technology; has it shifted towards design now?
He Xiaopeng:
We invest nearly 2 billion on autonomous driving annually, but our styling is excellent too. Why wouldn’t you spend 2 billion on design? I haven’t seen any Chinese car company spending 2 billion a year just on design.
The user buys a car, and the top three or at worst top four designs are certainly among them. Why is there no equivalent investment in research and development? Thus, we allocated greater weight and resources to it.
In Shanghai, over the past 12 months, we rented two buildings for the styling team, providing them with whatever they needed. The best resources and treatment must first be allocated to styling here in China.
Q: What changes will there be in the design of upcoming Xpeng vehicles?
Wang Tan: Firstly, the future models will undoubtedly become more refined. Secondly, we will unify the design language with both emotional and rational logic. Thirdly, since we are a tech company, a future mobility exploration company, the upcoming designs will increasingly highlight our technology. Lastly, I believe that the design might also convey more emotional value.
Q: Recently, Labubu has become popular. Is it possible for Chinese cars to lead globally in this way?
He Xiaopeng: I don’t really know why Labubu became popular. As a counterexample, in 2021, we insisted on taking the P7 overseas, where few users found it beautiful. I was quite surprised, as I thought the car was beautiful, but apparently not.
There are at least two points, European users think differently. First, European users believe that no matter how beautiful it is, it must also be practical for family use, and the trunk must not be small. Moreover, I think all beautiful coupes in the future may need air suspension, which in Europe might not be achievable.
From another angle, the emotional value of China is different from global emotional value.
Q: Is there still a debate between LiDAR and the visual approach?
He Xiaopeng: Regarding LiDAR or the visual approach, we saw changes in our products last year, but the decision was made the year before. For current intelligent assisted driving, the next generation of autonomous driving, and even driverless driving, we are fully committed to the visual approach.
Previously, vision wasn’t effective due to inadequate computational power. The vision system didn’t have enough pixel density, frame rate, or spatiotemporal logic, so it didn’t work well.
For instance, in the future, can we visually detect a nail in the road? Future vision systems will be able to achieve this, but LiDAR cannot. The upper limit of vision far exceeds that of radar. By 2027, this won’t be an issue, though today it remains a topic of debate.
This article is a translation by AI of a Chinese report from 42HOW. If you have any questions about it, please email bd@42how.com.