Las Vegas : The CEO of AWS, Adam Selipsky, unveiled a number of new products on Tuesday, one of which is a chip designed specifically for generative AI and machine learning training, which will increase the adoption of GenAI across businesses of all sizes.
The CEO of AWS stated that 80% of unicorns—companies valued at $1 billion or more—operate worldwide on their cloud and that, with the advent of GenAI, innovation will only spread throughout businesses of all sizes. For over 25 years, the cloud giant has been utilizing AI and ML solutions.
During the company’s annual flagship “re: Invent” conference, Selipsky gave the keynote address. He stated that innovators all over the world, from major financial service providers to automotive conglomerates, depend on AWS to power their businesses.
“With GenAI, we are constantly reinventing ourselves.” Companies of all sizes are experimenting with GenAI and developing large language models (LLMs) for early-use cases. “We are providing a generative AI stack for enterprises,” he told a crowded audience here.
Selipsky unveiled the next generation of two AWS-designed chips, AWS Graviton4 and AWS Trainium2, for a wide range of customer workloads, including machine learning (ML) training and generative artificial intelligence (AI).
Graviton4 processors outperform current generation Graviton3 processors by up to 30% in compute performance, 50% in cores, and 75% in memory bandwidth, delivering the best price performance and energy efficiency for a wide range of workloads running on Amazon EC2.
More than 50,000 customers—including the top 100 EC2 customers—use Graviton-based instances to get the best price performance for their applications. AWS has built more than 2 million Graviton processors and presently offers more than 150 different Graviton-powered Amazon EC2 instance types globally at scale.
During the event, the company unveiled three new serverless innovations that span its analytics and database portfolio and enable customers to scale their data infrastructure more quickly and easily to meet their most demanding use cases.
The new serverless innovations “build on the foundation to make it easier for customers to scale to millions of transactions per second, quickly add capacity at a moment’s notice, and dynamically adapt workload patterns to optimize for performance and cost,” according to Dr. Swami Sivasubramanian, vice president of Data and Artificial Intelligence at AWS.
Additionally, the business unveiled a brand-new identity service that uses palm scanning to verify employees’ identities before letting them into actual buildings.
During its “AWS re:Invent 2023” event here, the company announced that the service, known as “Amazon One Enterprise,” eliminates operational overhead related to the management of traditional enterprise authentication methods, like badges and PINs.
Advanced artificial intelligence and machine learning are used by palm-recognition technology to generate a palm signature linked to identification credentials such as a PIN, employee ID, or badge.
According to Dilip Kumar, vice president of AWS Applications, “Amazon One Enterprise’s palm recognition technology is designed to deliver a highly accurate identification service that increases an organization’s overall security, while offering seamless authentication management with lower operational overhead.”
Amazon has introduced new devices that cost $195 and enable corporate users to access virtual desktop environments via the internet, such as Amazon WorkSpaces. Hardware from Fire TV Cubes houses the devices.