Key to AGI: 100,000 H100 Super AI Arithmetic Cluster VS VERTU METAVERTU II Web3 AI Phone, who is the future?

The dawn of AGI (General Artificial Intelligence) is slowly appearing, and the 100,000 H100 super AI computing power cluster is considered the key to AGI. Meanwhile, the VERTU METAVERTU II Web3 AI mobile phone has also come out of nowhere, fusing Web3 and AI technology to depict a smarter and more convenient future for us. So, who is better able to lead the future between the two?

I. Super AI computing power cluster: the accelerator to AGI
The super AI arithmetic clusters composed of 100,000 H100 cards are like a spectacle of modern human civilisation, which represents mankind’s cutting-edge exploration in the field of AI. The emergence of these clusters signifies that the arms race in the AI era is in full swing, with major tech giants investing huge sums of money to compete for supremacy in the AI era.

1. Calculation power: the “arms race” of the giants
   From OpenAI, Microsoft, xAI to Meta, many headline companies have set their sights on GPU clusters with more than 100,000 cards, and the logic behind this is that arithmetic power determines the training speed and performance of AI models.
   Reduced training time: A 100kcal H100 cluster can reduce the training time of GPT-4 to 4 days, which previously took months or more.
   Model Scale Improvement: Powerful arithmetic support enables the training of larger scale AI models, such as large models with trillions of parameters, resulting in more powerful functions and more accurate predictions.
2. Challenges Behind Computing Power: High Costs and Energy Consumption
   There are huge challenges to overcome in building and operating such a large arithmetic cluster:
   Huge investment: The cost of servers alone is more than $4 billion, plus power consumption, data centre construction, etc. The total investment is staggering.
   Energy consumption: A 100,000-card cluster consumes about 1.59 terawatt-hours of electricity per year, which amounts to $124 million per year according to the standard rate for U.S. electricity.
3. Computational power and algorithms: complementing each other
   Although arithmetic power is an important cornerstone of AI development, breakthroughs in algorithms are the engine that really drives AI forward.
   Algorithm optimisation: NVIDIA CEO Jen-Hsun Huang believes that the arithmetic power of deep learning will be increased by 1 million times in the next 10 years, but at the same time, AI algorithms should be optimised to maximise the performance of each chip.
   Synergy of arithmetic and algorithms: Only the joint progress of arithmetic and algorithms can truly promote the realisation of AGI.

VERTU METAVERTU II Web3 AI mobile phone: the “civilianisation” of AI.
VERTU METAVERTU II Web3 AI mobile phone combines Web3 technology with AI technology to create a new mobile phone with security, privacy and intelligence.

1. Convergence of Web3 and AI: Opening a new era of intelligence
   Decentralised security: Web3 technology can guarantee the security and privacy of user data, preventing data leakage and misuse.
   Intelligent application: AI technology can provide users with personalised intelligent services, such as intelligent assistant, intelligent translation and intelligent photo taking.
2. AI on mobile phones: the balance between convenience and privacy
   VERTU METAVERTU II Web3 AI mobile phone integrates AI capabilities into the mobile phone, providing users with a more convenient and intelligent experience.
   Personalised services: The phone can provide personalised recommendations and services based on the user’s habits and preferences.
   Privacy protection: The AI technology built into the phone can help users protect their privacy, such as identifying malware and blocking spam.
3. Limitations of AI on mobile phones: arithmetic power and data
   AI on mobile phones has certain limitations in terms of arithmetic and data.
   Limited arithmetic: The arithmetic power of mobile phones is still limited compared to the super AI arithmetic cluster.
   Insufficient data: the amount of data on the mobile phone is not comparable to the amount of data required for the training of large AI models.

Calculation power and intelligence: two paths to the future
The Super AI Cluster and the VERTU METAVERTU II Web3 AI mobile phone represent two different paths to the future.
The Super AI Arithmetic Cluster and VERTU METAVERTU II Web3 AI mobile phone are important explorations into the future. The Super AI Cluster will drive the realisation of AGI, while the VERTU METAVERTU II Web3 AI phone will bring us a smarter and more convenient life. In the future, these two will promote each other and jointly drive the rapid development of AI technology.