Technology peripherals
AI
The 'emergence' moment of artificial intelligence: How do data centers solve problems?
The 'emergence' moment of artificial intelligence: How do data centers solve problems?
When hundreds of large AI models for industries appear, the data center hosting the large models is quietly changing. Large models require large computing power. On the one hand, data centers will provide diverse computing comprehensive capabilities as the computing power base for digital transformation to meet the intelligent needs of different industries; on the other hand, data centers continue to improve energy use efficiency, which requires A better computing architecture and lower energy consumption generate greater computing power, which not only achieves green and low carbon itself, but also empowers the intelligent transformation of other industries and promotes the carbon reduction of the whole society.
Data from the Ministry of Industry and Information Technology shows that in 2022, the total size of data center racks in use in my country will exceed 5.9 million standard racks, the number of servers will be approximately 20 million, and the average annual growth rate of data center storage capacity will exceed 50%. At the same time, against the background of the explosive demand for data and computing power in the whole society, the power consumption of data centers has inevitably increased rapidly. So, facing the dual test of "computing power" and "consumption reduction", how will future data centers continue to optimize, grow steadily in scale, and gradually improve energy efficiency? What will the data center of the future look like?
Recently, Huawei released the "Data Center 2030" report, which provides a reference direction for the transformation and upgrading of future data centers.
Innovation emerges and the new direction of development around “five-effect improvement”
Today, data centers are almost closely related to daily life. Familiar terms such as AI, cloud computing, smart cities, and the metaverse are all inseparable from the support of data centers. At the same time, trends such as energy conservation, sustainability, and safety are also gradually strengthening. Major enterprises are increasing investment and construction of data centers to meet the growing demand for computing power.
It is foreseeable that with the rapid growth of the global economy, the development and construction of data centers will continue to be in a period of rapid development. Coupled with the strong support from emerging industries in various places, it has brought great opportunities to the development of the data center industry. Advantage.
Without changing the original meaning, the content needs to be rewritten into Chinese. What needs to be rewritten is: But with rapid development, choosing the direction and path has become a kind of ability and wisdom. Computing power requirements and resource constraints are the main challenges for future data center development, and innovation needs to be centered around improving efficiency
"Data Center 2030" starts from the five major future scenarios affecting data centers and points out that the future development of data centers will show a number of important trends. The report predicts that computing power will show rapid development, and computing power and scale efficiency will become national and The core competitiveness of enterprises; at the same time, driven by AI, a panoramic revolution will occur in data centers. The industrial label of data centers will transform from high energy consumption to green development enablers. Multi-stream collaborative data centers will also become popular. This builds a green, low-carbon, and leading data center with computing efficiency.
It can be seen that the development of future data centers is a fast-changing and challenging process. If we want to figure out the future data center, unlock the digital economy, a market with huge imagination, and promote the high growth of enterprises in the data economy era, we need some new ideas to solve the "multiple higher-order equation" of the current data center.
In order to cope with breakthroughs in scientific research and resource demands for computing power, data centers under the new situation are in urgent need of technological innovation and model changes. The "Data Center 2030" report points out that future data centers need to improve their performance in five aspects: "energy efficiency, computing efficiency, operation efficiency, digital efficiency, and human efficiency" to cope with the computing power supply and energy consumption caused by large artificial intelligence models. challenge
Optimize energy efficiency to build a green zero-carbon data center; improve computing efficiency to achieve efficient supply of data center computing power; enhance operational efficiency to create a hyper-converged data center network; improve digital efficiency to maximize data value; ultimately improve Human efficiency, realizing automated data centers. It can be said that the improvement of "five effects" points out an important direction for the technological innovation of future data centers
It is the holistic perspective that defines the technical characteristics of future data centers for the first time
With the development of cloud computing and artificial intelligence in China, the scale of ultra-large-scale data center applications continues to increase, and the construction concept of data centers has also changed. The most typical change is that the construction of data centers and IT equipment are more integrated. close. As an infrastructure, the data center will also undergo top-down changes accordingly. To realize the linkage between applications and technology from a business perspective, the data center needs to consider cooling, power supply, monitoring and operation and maintenance as a whole to achieve vertical integration
The mainstream of future data center technology development will be achieved through system-level innovation, which means that the next generation of data centers needs to break the tradition and greatly improve the efficiency of the data center through software and hardware collaboration
Specifically, what are the key features of future data centers? What are the evaluation indicators for new data centers? The "Data Center 2030" report believes that new data centers in the future will have six major technical characteristics: diverse ubiquity, security and intelligence, zero-carbon energy saving, flexible resources, peer-to-peer interconnection, and system Moore.
Diverse and ubiquitous: In the future, data centers will move towards the polarization of ultra-large clusters and ultra-light edges. At the same time, facing new scenarios, a variety of innovative data centers will also appear, such as space data centers, undersea data centers, etc.
Security Intelligence: Future new data centers need to have the characteristics of high security, high reliability and high intelligence. Al and data enable the entire life cycle of data center planning, construction, and operation, promoting the development of data centers in the direction of high efficiency, energy saving, and intelligence.
Zero-carbon energy saving:The proportion of clean energy such as wind energy and solar energy in the energy structure of the data center increases. At the same time, energy storage technology has become an important way to reduce the power cost of data centers through "peak shaving and valley filling"; liquid cooling technology helps data centers save energy and reduce noise.
Flexible resources: Cloud architecture will become one of the "standard configurations" of future data center infrastructure. The next generation cloud data center architecture will follow the lines of "full pool" and "soft computing" , the direction of "pan-collaboration" continues to evolve.
Peer-to-peer interconnection: In order to improve data processing efficiency, future computing will break through the bottleneck of the von Neumann architecture. Computing, storage and communication modules are peer-to-peer interconnected through a unified bus. In the future, new data centers will have Features of hyper-convergence, high performance and endogenous light.
System Moore: Moore's Law, which dominates the development of semiconductor integrated circuits, is encountering dual limitations of physics and economics, and there is an urgent need to adopt new technologies to promote the continued vigorous development of the future information industry.
Six technical characteristics of new data centers
Not only that, in order to match the multi-form data center, the report also proposed the "6 new" reference architecture of the new data center in the industry for the first time, including the new infrastructure layer, the new computing power base layer, the new resource scheduling layer, the new There are six levels of data management layer, new collaborative service layer, and new intelligent management layer to indicate as clearly as possible the key points and elements of new data center construction.
New Data Center “6 New” Reference Architecture
We can see some new ideas for future data centers from the "Data Center 2030" report, which can solve this "multiple higher-order equation". However, to get the "correct answer" in the complex calculation process requires not only the correct problem-solving ideas, but also the cooperation of powerful "problem solvers" to jointly promote industry innovation and progress
Cooperate to solve problems and lead the upgrading of the data center industry
Today, data centers gather technological innovation, and are also responsible for protecting the national economy and people's livelihood, and shouldering practical social responsibilities. Therefore, the future data center market needs more experienced “problem solvers” to participate.
As a leader in the global data center industry, Huawei has begun to make preparations to "solve problems" with all sectors of the industry to meet the challenges of data centers. The "Data Center 2030" report not only provides reference indicators for the construction of new data centers, but also details the practical direction of future data center technology innovation.
Huawei is willing to work together with the entire industry chain to be a problem solver for new data centers and lead the innovation and development of the future data center industry.
The above is the detailed content of The 'emergence' moment of artificial intelligence: How do data centers solve problems?. For more information, please follow other related articles on the PHP Chinese website!
Hot AI Tools
Undresser.AI Undress
AI-powered app for creating realistic nude photos
AI Clothes Remover
Online AI tool for removing clothes from photos.
Undress AI Tool
Undress images for free
Clothoff.io
AI clothes remover
AI Hentai Generator
Generate AI Hentai for free.
Hot Article
Hot Tools
Notepad++7.3.1
Easy-to-use and free code editor
SublimeText3 Chinese version
Chinese version, very easy to use
Zend Studio 13.0.1
Powerful PHP integrated development environment
Dreamweaver CS6
Visual web development tools
SublimeText3 Mac version
God-level code editing software (SublimeText3)
Hot Topics
1377
52
Bytedance Cutting launches SVIP super membership: 499 yuan for continuous annual subscription, providing a variety of AI functions
Jun 28, 2024 am 03:51 AM
This site reported on June 27 that Jianying is a video editing software developed by FaceMeng Technology, a subsidiary of ByteDance. It relies on the Douyin platform and basically produces short video content for users of the platform. It is compatible with iOS, Android, and Windows. , MacOS and other operating systems. Jianying officially announced the upgrade of its membership system and launched a new SVIP, which includes a variety of AI black technologies, such as intelligent translation, intelligent highlighting, intelligent packaging, digital human synthesis, etc. In terms of price, the monthly fee for clipping SVIP is 79 yuan, the annual fee is 599 yuan (note on this site: equivalent to 49.9 yuan per month), the continuous monthly subscription is 59 yuan per month, and the continuous annual subscription is 499 yuan per year (equivalent to 41.6 yuan per month) . In addition, the cut official also stated that in order to improve the user experience, those who have subscribed to the original VIP
Context-augmented AI coding assistant using Rag and Sem-Rag
Jun 10, 2024 am 11:08 AM
Improve developer productivity, efficiency, and accuracy by incorporating retrieval-enhanced generation and semantic memory into AI coding assistants. Translated from EnhancingAICodingAssistantswithContextUsingRAGandSEM-RAG, author JanakiramMSV. While basic AI programming assistants are naturally helpful, they often fail to provide the most relevant and correct code suggestions because they rely on a general understanding of the software language and the most common patterns of writing software. The code generated by these coding assistants is suitable for solving the problems they are responsible for solving, but often does not conform to the coding standards, conventions and styles of the individual teams. This often results in suggestions that need to be modified or refined in order for the code to be accepted into the application
Can fine-tuning really allow LLM to learn new things: introducing new knowledge may make the model produce more hallucinations
Jun 11, 2024 pm 03:57 PM
Large Language Models (LLMs) are trained on huge text databases, where they acquire large amounts of real-world knowledge. This knowledge is embedded into their parameters and can then be used when needed. The knowledge of these models is "reified" at the end of training. At the end of pre-training, the model actually stops learning. Align or fine-tune the model to learn how to leverage this knowledge and respond more naturally to user questions. But sometimes model knowledge is not enough, and although the model can access external content through RAG, it is considered beneficial to adapt the model to new domains through fine-tuning. This fine-tuning is performed using input from human annotators or other LLM creations, where the model encounters additional real-world knowledge and integrates it
Seven Cool GenAI & LLM Technical Interview Questions
Jun 07, 2024 am 10:06 AM
To learn more about AIGC, please visit: 51CTOAI.x Community https://www.51cto.com/aigc/Translator|Jingyan Reviewer|Chonglou is different from the traditional question bank that can be seen everywhere on the Internet. These questions It requires thinking outside the box. Large Language Models (LLMs) are increasingly important in the fields of data science, generative artificial intelligence (GenAI), and artificial intelligence. These complex algorithms enhance human skills and drive efficiency and innovation in many industries, becoming the key for companies to remain competitive. LLM has a wide range of applications. It can be used in fields such as natural language processing, text generation, speech recognition and recommendation systems. By learning from large amounts of data, LLM is able to generate text
Samsung introduces BM1743 data center-grade SSD: equipped with v7 QLC V-NAND and supports PCIe 5.0
Jun 18, 2024 pm 04:15 PM
According to news from this website on June 18, Samsung Semiconductor recently introduced its next-generation data center-grade solid-state drive BM1743 equipped with its latest QLC flash memory (v7) on its technology blog. ▲Samsung QLC data center-grade solid-state drive BM1743 According to TrendForce in April, in the field of QLC data center-grade solid-state drives, only Samsung and Solidigm, a subsidiary of SK Hynix, had passed the enterprise customer verification at that time. Compared with the previous generation v5QLCV-NAND (note on this site: Samsung v6V-NAND does not have QLC products), Samsung v7QLCV-NAND flash memory has almost doubled the number of stacking layers, and the storage density has also been greatly improved. At the same time, the smoothness of v7QLCV-NAND
To provide a new scientific and complex question answering benchmark and evaluation system for large models, UNSW, Argonne, University of Chicago and other institutions jointly launched the SciQAG framework
Jul 25, 2024 am 06:42 AM
Editor |ScienceAI Question Answering (QA) data set plays a vital role in promoting natural language processing (NLP) research. High-quality QA data sets can not only be used to fine-tune models, but also effectively evaluate the capabilities of large language models (LLM), especially the ability to understand and reason about scientific knowledge. Although there are currently many scientific QA data sets covering medicine, chemistry, biology and other fields, these data sets still have some shortcomings. First, the data form is relatively simple, most of which are multiple-choice questions. They are easy to evaluate, but limit the model's answer selection range and cannot fully test the model's ability to answer scientific questions. In contrast, open-ended Q&A
Five schools of machine learning you don't know about
Jun 05, 2024 pm 08:51 PM
Machine learning is an important branch of artificial intelligence that gives computers the ability to learn from data and improve their capabilities without being explicitly programmed. Machine learning has a wide range of applications in various fields, from image recognition and natural language processing to recommendation systems and fraud detection, and it is changing the way we live. There are many different methods and theories in the field of machine learning, among which the five most influential methods are called the "Five Schools of Machine Learning". The five major schools are the symbolic school, the connectionist school, the evolutionary school, the Bayesian school and the analogy school. 1. Symbolism, also known as symbolism, emphasizes the use of symbols for logical reasoning and expression of knowledge. This school of thought believes that learning is a process of reverse deduction, through existing
SOTA performance, Xiamen multi-modal protein-ligand affinity prediction AI method, combines molecular surface information for the first time
Jul 17, 2024 pm 06:37 PM
Editor | KX In the field of drug research and development, accurately and effectively predicting the binding affinity of proteins and ligands is crucial for drug screening and optimization. However, current studies do not take into account the important role of molecular surface information in protein-ligand interactions. Based on this, researchers from Xiamen University proposed a novel multi-modal feature extraction (MFE) framework, which for the first time combines information on protein surface, 3D structure and sequence, and uses a cross-attention mechanism to compare different modalities. feature alignment. Experimental results demonstrate that this method achieves state-of-the-art performance in predicting protein-ligand binding affinities. Furthermore, ablation studies demonstrate the effectiveness and necessity of protein surface information and multimodal feature alignment within this framework. Related research begins with "S
