The article focuses on the comparison between the AI capabilities of the Kirin 970 and the Snapdragon 845. The author believes that the Kirin 970 has a slight edge over the 845 in terms of AI performance. EETOP does not make a judgment on which is stronger or weaker, so please don’t focus only on the soft text about the 970.
In the second half of 2017, the mobile phone market began to emphasize the term “artificial intelligence.†Apple and Huawei, both leaders in self-developed chip technology, launched the A11 Bionic and Kirin 970 chips, integrating AI features into their flagship devices.
While the industry was closely watching other high-end chip manufacturers like Qualcomm and Samsung, Qualcomm finally released the Snapdragon 845, adding AI capabilities to its lineup. However, its release came later than expected, raising questions about how it compares to the Kirin 970 in AI performance.
This article delves into the AI architecture of both chips, comparing their strengths and weaknesses.
The implementation of AI requires collaboration between cloud and device sides to deliver an optimal user experience. Over the past year, AI has gradually shifted from the cloud to the device, with smartphones being the most advanced end-side AI devices. However, they still lack sufficient computing power.
Improving AI computing power is a key challenge for smartphone chips, and each manufacturer has taken different approaches in their AI architecture.
In addition to traditional components like CPU, GPU, and ISP, the Kirin 970 includes a dedicated NPU (Neural Processing Unit) designed specifically for AI tasks. This HiAI architecture significantly enhances AI performance, offering about 25 times the performance and 50 times the energy efficiency compared to four Cortex-A73 cores. This allows for better image recognition, voice interaction, and smart photography, making the phone "understand you better."
In contrast, the Snapdragon 845 uses a more flexible machine learning architecture, with AI calculations handled through the Kryo 385 CPU, Adreno 630 GPU, and Hexagon 685 DSP. While it leverages integrated AI algorithms to improve overall performance, Qualcomm claims the 845 offers three times the AI computing power of the 835. Both companies are also building open AI ecosystems for developers.
Which approach is more effective? The Kirin 970’s dedicated NPU avoids resource conflicts and improves AI efficiency, while the 845 relies on existing hardware units, which may lead to higher energy consumption and slower performance.
Additionally, since the Kirin 970 is used in Huawei devices, it can be fully integrated with the system to provide a seamless AI experience. The 845, on the other hand, offers an AI platform but doesn't require manufacturers to build applications around it.
Qualcomm highlights improved camera, voice, and gaming experiences with the 845, but there's no specific mention of AI applications. Given the cost, these enhancements are likely limited to high-end models, and the market share of Snapdragon chips is shrinking.
Huawei Mate 10 series, equipped with the Kirin 970, delivers rich AI applications:
- **AI Scene Recognition**: It can identify 13 scenes and objects in real time, adjusting camera settings automatically for better photos.
- **Intelligent Voice Assistant**: Beyond basic commands, it understands user needs, answers questions, and even predicts actions based on voice patterns.
- **App Integrations**: Apps like Vibrato, Daily P, and Today’s Headlines have joined the Huawei Developer Alliance, with AI versions planned for release later this year.
Beyond AI, 5G is another key trend in the mobile industry. It complements AI by enhancing cloud-device communication, improving user experiences. In baseband performance, the Kirin 970 outperforms the 845, supporting faster download speeds and optimized high-speed rail communication.
At the 2017 GTI Symposium, Qualcomm showcased dual-SIM dual-4G, but it was the Kirin 970 that first brought this feature to market, offering high-definition VoLTE calls.
Security is also crucial when combining cloud and device AI. The Kirin 970 uses an inSE solution (Integrated Secure Element), providing hardware-level security. It supports multiple encryption algorithms and has achieved financial-grade security certifications.
In summary, the competition between Kirin and Snapdragon has evolved from a dislocation race to a true battle at the leading edge. Both chips bring enhanced experiences to consumers, with AI, communication, and security becoming essential in daily life. Whoever masters these technologies will have a stronger position in the future market. As users, we’re excited to see such advancements shaping the next era of mobile technology.
Radiators in this case are manufactured with Stainless steel (SS304, SS316 and SS316L).
These radiators are manufactured with both 1mm CRCA sheet and 1.2 mm CRCA sheet as required and centre distance varying from 600 mm to 4000 mm. Stainless steel radiators can be offered with and without paint.
Through the method, the stainless steel plate type radiator for the transformer is simple in structure, free of complex treatment process on the surface of the radiator, not prone to oxidizing and corroding, long in service life, and high in welding strength.
Stainless Steel Radiator,Stainless Steel Transformer Radiator,Stainless Steel Cooling Radiator,Stainless Steel Weather Proof Radiator
Shenyang Tiantong Electricity Co., Ltd. , https://www.ttradiator.com