How AI Technology Is Already in Our World Today

Written by Adrian Rubin

Artificial intelligence (AI) is one of the most important emerging technologies in development today. Advancements in AI and machine learning (ML) are progressing at incredible speeds. As they drive computer chip innovations and transform hardware development, we are continually reminded of AI’s potential to transform global society with each new application.

Several companies are embracing AI technology right now. From obvious companies like Amazon, Microsoft, and Google to ventures in cryptocurrencies, emerging AI applications represent the new Wild West in the technology sector. AI is not just for digital personal assistants. Innovative leaders, like Israel, rely upon AI for its national defense. This has bolstered hardware business for Intel, Cisco, Qualcomm, and even Apple. The current and emerging developments specific to AI hardware development are outlined below.

AI Hardware Investments

AI is developing so rapidly that it is difficult to stay on top of all the technologies currently available, let alone predict what will be on the market within the next year. There are several computing applications that demand increasingly more power than even today’s processors are capable of supplying. Deep machine learning is one of these. Three significant areas of AI hardware investment are: 

  • neuromorphic/neurosynaptic architectures
  • graphics processing units (GPUs)
  • quantum computers
Ai Technology2 Adrian Rubin

Emerging AI Technology

Besides Intel, AMD, NVIDIA, and the big tech companies like Microsoft, Google, IBM, and Amazon, several other companies are making waves in the AI Chipset Market. Allied Market Research expects the CAGR of 45% to propel the market to over $91M by 2025. The report also lists major players and important factors that impact innovation. Digital Journal and Nanalyze list other key players. Here are some recent developments:

Graphcore provides efficient acceleration of AI applications. According to HPCwire (October 18), its first AI architecture is due out soon. Penguin Computing is one of the first in line to try out samples.

Cerebras Systems is a rapidly growing startup that recently hired former Intel executive Dhiraj Mallick. He will provide data center experience for Cerebras’ next generation machine learning chips.

Xilinx is making waves in field-programmable gate arrays (FPGA). It boasts the industry's first Adaptive Compute Acceleration Platform (ACAP) for data center applications. In Beijing on October 16th, the company announced the major platforms currently employing their FPGA-as-a-service model. Xlinx’s FPGA developments intensifies competition for companies like Nvidia and AMD.

With more competition, Nvidia has intensified R&D efforts focused on projects like its Hyperscale Inference Platform. It also designs a GPU for Tesla.

Several Chinese companies are developing new AI chips amidst the Ministry of Science and Technology’s call for innovation to challenge Nvidia. DeePhi raised $40 million. The entrance of Alibaba and Baidu into a Cambricon-dominated Chinese AI chip market has been compared to Google’s anomalous market activity.

Inbenta is innovating machine learning for natural language processing. It also has applications for branding.

Hardware for deep machine learning architectures finds inspiration from brain infrastructure. GPUs are an optimal modern technology for running AI algorithms. Quantum computers are transitioning AI development from hardware based on binary logic to a system based on the principles of quantum mechanics.

The AI budgets of early adopters is expected to rise by at least 50% annually over the next few years. This is due directly to research and development. The motivation for financial expenditures and allocating development resources is sound. It is driven by the historical success of AI to deliver exponential returns on investment. 

Chips are at the Center of AI Hardware Development

AI stimulated computer chip development is unlike anything before. Machine learning demonstrated so much potential that new processor development began emerging from unconventional sources. Of course, Intel had its hand in creating the current market. While companies like Microsoft, Apple, and IBM are not surprising developers of AI processors, they do illustrate the increasing demand. Then there are companies like Google and Amazon. Yes, these two are branching out into seemingly everything tech, but developing an AI chip themselves is quite a commitment. If this is not enough to indicate a radical shift in the global technology sector, consider Tesla. The electric automobile maker is exploring engineering its own chips for use in its cars.

Beyond the general promise of AI to produce rewards, developing a company-specific chip has clear advantages. It allows for the kind of highly specialized tasks that broad chip architectures are either incapable or inefficient at providing. AI requires processes that differ dramatically from conventional computer applications. Being able to optimize the use of past experiences becomes increasingly complex in high-level organizations. Organizationally, specific chip development is crucial for many of today’s big businesses. It will likely become increasingly important with each passing year, as well as, when more companies look into AI. 

AI Hardware Summit

The AI Hardware Summit is one of the most important events in a market that exceeds $7 billion. Within the coming years, AI chip development is expected to grow to more than $50 billion. This might be a conservative number, considering that AI research and development greatly outpace suitable platforms. Developing hardware to support AI innovations is a priority. The current infrastructure for hardware development is proving inefficient at meeting the highly specific demands of independent organizations.  

In 2018, more than 600 senior technology leaders signed up to attend the summit. They represent a diverse background of semiconductor companies, chip developers, system vendors/OEMs, financiers, investors, data centers, start-ups, and end users. Some of the topics discussed include developments within the industry, the role of software, and alternative innovation.

Conclusion

The potential for AI technology to transform our world is profound. Once a novelty, AI inspired the kind of deep machine learning that has made scientists rethink what is possible. Modern AI technology has become more organizationally specific. This has brought some surprising players into AI hardware development. The recent shift in the industry indicates a more unpredictable outcome than many foresaw. 

About The Author

Adrian Rubin

Adrian Rubin

Writer, Editor

Adrian Rubin works as an editor for the New York Daily Herald and the Pittsburgh Gazette. He writes for newspapers across America. https://www.adrianrubin.com/

Share Article:

Sign up to the Hardware Massive newsletter

Related Guest Articles