Home >Technology peripherals >It Industry >AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean it's better to use?

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean it's better to use?

PHPz
PHPzforward
2023-09-19 12:37:021287browse

My biggest worry has come.

Just kidding, AnTuTu recently announced the performance rankings for automotive and computer versions. Frankly speaking, this is the first "bottom test" of the performance of mainstream automotive and computer chips. This is the ranking list based on a large number of benchmark scores of dozens of cars and devices collected in the backend nearly a month after the public beta of the AnTuTu car and machine version was launched.

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Picture from AnTuTu)

The top ten automotive chips in the list are very interesting, and they are currently ranked The first place is Qualcomm Snapdragon 8295, with a score of nearly 700,000 points. It uses a 16+128GB storage combination and is currently the car chip with the highest score. The first launch of this car chip should be Geely Galaxy’s new pure electric flagship model E8, which is expected to be launched later this year.

There are many manufacturers rushing to launch the Qualcomm Snapdragon 8295. They use the quad-core Kyro 695 Gold and the quad-core Kyro 695 Sliver. The GPU model is Adreno 690, which is the same as the third one currently used for the Windows On ARM platform. The 3rd generation Snapdragon 8cx is the same and I believe it will soon replace the 8155 and become the mainstream.

Qualcomm Snapdragon 8155, the most common flagship car chip at this stage, ranks fifth. This is still surprising. After all, it is the most commonly used flagship chip by car companies, and there are still some rankings. Awkward.

Qualcomm Snapdragon 8195, which ranks second, may be relatively rare. Its running score is second only to Qualcomm Snapdragon 8295. The models it is equipped with mainly include electric vehicles from traditional car companies such as Cadillac LYRIQ and Buick E5. According to Xiaotong’s understanding, the Qualcomm Snapdragon 8195 is based on the Qualcomm Snapdragon 865, which is much stronger than the 8155 that was born out of the Snapdragon 855, and the running score results also reflect this.

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Picture from Cadillac)

We can also see many unique code names from the list The automotive chipsets are RK3588 from Rockchip. This is a "domestic core" with a relatively pure pedigree. Although it uses the public version of the core, its running score strength is also in the top three on the list.

Looking carefully, Tesla’s AMD chipset is not among them. If it were, Xiaotong’s visual inspection would rank first.

Professional player Tesla (referring to performance comparable to PS5) is prohibited from participating, right? The car owner will probably complain to the official, but Tesla's operating system is different from other electric vehicles, and it has not been adapted yet. Tesla owners can also rest assured that AnTuTu will most likely not ignore such a big issue for the electric car giant.

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Picture from Tesla)

Of course, the interior space of the car is very large, and chips are often present" If one chip is not enough to make up for two, this approach can also significantly improve the computing power of the car system. This is a part that is not disclosed in the current AnTuTu car version performance list.

Xiaotong also noticed that some new car-making forces have chosen the route of "sharing mobile phone computing power" and borrowing the computing power of mobile phones to supplement the car-machine system. The question is, if the running score at this time is calculated for cars or mobile phones?

Antutu announced the car and computer benchmarking list, which probably means that the era of car benchmarking is coming.

Car and machine performance benchmarks,gives manufacturers a new direction to “brag”

Before the benchmarks and games are released, we might as well first Take a look at how new energy vehicle companies promote the performance of their vehicles.

Many official detailed introduction pages describe the car chip in an understatement. Apart from "using XXXX chips to provide powerful computing power for AI, software and entertainment", it is almost difficult to find more detailed descriptions. When car chip performance is no longer the focus, manufacturers will even delete the description of chip performance, leaving only the sentence "very strong performance."

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Picture from Li Auto)

Judging from the rules, it can be found that car companies are really interested in promoting the performance of car chips. Without any more intuitive means, we can only tell the consumer group from a relatively metaphysical perspective that "the performance of this car is very strong and can meet the needs of entertainment and AI." That's it. The relevant introduction is not as long as the chassis. , battery, and car body come in abundance.

Cars have entered the era of electrification. The traditional fuel system (except for refueling) has been cleverly replaced by a three-electric system, which greatly improves energy efficiency. In just a few years, new energy vehicles have caught up with fuel vehicle technology and are comparable to fuel vehicles in terms of chassis, integration, handling, comfort, etc., and have spawned a wealth of intelligence based on electrification. Potential, the huge battery can also drive more vehicle electrical appliances.

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Picture from Li Auto)

Obviously, fuel vehicles and new energy vehicles are no longer on the same track, and they cannot be used purely for driving. Texture summarizes the experience of a car. This is what electric vehicles should do, and if they compete with fuel stations on the same stage, in my opinion, they are no different from "electric fuel vehicles."

New energy vehicles should be more "confident". In addition to the driving quality, they should also put high-end assisted intelligent driving, AI, and entertainment richness on the table. Smart cars are also used to judge whether they are good or bad. one of the standards.

The latter two have certain requirements for the car chip. The stronger the performance, the smoother the experience will be in theory. However, smoothness itself is a subjective feeling, which is difficult to quantify through data. This also leads to consumers not understanding the greatness of the car machines promoted by automobile manufacturers. They can only promote them from a functional level. The car machine chips are only Used a "flagship" to bring it.

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Picture from Li Auto)

How to let consumers understand more intuitively what a "flagship chip" is , there has been no solution before, because the use scenarios of cars and machines are just that, no matter how powerful the performance is, they can only watch videos, listen to music and use navigation. The big deal is that the AI ​​processing speed is faster.

In order to highlight the performance of the car and machine, some manufacturers have created a set of car-machine interface based on the Unreal game engine to make it more load-bearing. The main focus is a bells and whistles.

When games and benchmarks gradually invade the automotive field, high-load scenarios belonging to the car and machine are captured by manufacturers. From now on, when looking at the performance of a car or computer, we will look at the AnTuTu benchmark of the car or computer, or conduct a long-term high-load game test like a mobile phone.

The game is relatively simple. The mainstream car and computer systems are all based on the Android system. In theory, it can come with its own game library or allow third-party applications to be installed. For example, the game suite of Denza N8 or pre-installed in Denza Something like Real Racing 3 on N7. But the problem is that many vehicles require root permissions to install apps on their own, and there are very few open installation systems like BYD.

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Pictures taken by Tram Express)

In other words, at this stage, the car and machine can even freely install software. One problem is that car companies need to do it step by step to open up the installation of third-party software and promote the improvement of the in-car entertainment ecology. It is only in the "initial stage" now.

I believe this does not need to wait too long, because car manufacturers and car chip companies have finally found a way to promote the performance of cars.

The era of full-vehicle benchmarking has not yet come

Why do we need to do vehicle-machine benchmarking? AnTuTu believes that the smart cockpit of new energy vehicles is very similar to the situation of domestic mobile phones in the early years. The large central control screen, high-precision touch and high-resolution, narrow bezel, and high-performance chips are indeed very similar to mobile phones when viewed individually. .

Therefore, the smart cockpit system must be distinguished from the car itself for separate performance evaluation.

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Picture from AnTuTu)

New energy vehicles are gradually entering the field of digital products. To put it simply Say, this commodity is like a smartphone with a mobile chassis.

In fact, most car-grade chips are derived from the mobile platform of smartphones, and some are directly used. This also leads to the fact that the replacement of car chips is generally not as fast as that of smartphones, and most models use mainstream solutions that are "unified and distributed" by chip manufacturers. The hardware gap is currently not large.

However, automotive chips are different from smartphones. The mainstream platforms are not only MediaTek, Samsung, Qualcomm, but also more competitors such as Rockchip. There will be more chip choices than mobile phones. .

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Picture from NIO)

"Using 8155 engine does not mean smoothness" Xiaotong still agrees that new energy vehicles are defined by software, and hardware can only determine the lower limit, while software maintenance and optimization are the key to improving the upper limit of experience. Cars will always reach a point where, like smartphones, host manufacturers and chip suppliers will cooperate in depth to bridge the gap between software and hardware.

But software adaptation and optimization is an extremely painful process. Car companies did not have much relevant software development experience when they transformed. They either cooperated with mobile phone manufacturers, or directly "taken over" the software departments of mobile phone manufacturers to accelerate Intelligent development. We cannot see the results in the short term, and we still need patience and support.

At least until the industry has completed educating users, car and machine benchmarks and games will still tend to be “meaningless shows off”.

AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean its better to use?
(Picture from AnTuTu)

If there are no more new features specific to the car scene, then Xiaotong believes that games and benchmarks are the "vent" for car and machine performance. If consumers don't understand, then use more realistic data and performance. No matter what, consumers ultimately just want to know that the car and machine they are looking for are smooth enough, but under the concept of "shopping around", they also want to know whose car and machine are stronger.

In short, when car companies release new cars in the future, it will definitely be inevitable to "run a score". I wonder if everyone likes to compare benchmarks?

This article comes from the WeChat public account: Dianchetong (ID: dianchetong233), author: Tea Governor Shopkeeper

Advertising statement: The external jump links contained in the article (including Not limited to hyperlinks, QR codes, passwords, etc.), used to convey more information and save selection time. The results are for reference only. All articles on this site include this statement.

The above is the detailed content of AnTuTu announces its first car-machine benchmark rankings. Does a high score really mean it's better to use?. For more information, please follow other related articles on the PHP Chinese website!

Statement:
This article is reproduced at:ithome.com. If there is any infringement, please contact admin@php.cn delete