10,000 USD chips from NVIDIA aim to power the AI revolution

Advanced language models powered by Artificial Intelligence have taken the technology industry by storm of late. Platforms like Google and Microsoft are racing to integrate AI capabilities into their search engines after OpenAI's revolutionary model ChatGPT led the way in integrating AI to produce outstanding results.

AI models are getting highly complex as they take on next-level challenges, such as conversational AI. Such platforms need immense computing and processing power for AI, ML, and data analytics workloads. Most of these language models are making use of NVIDIA's state-of-the-art A100 chip, which is powered by its Ampere Architecture, with the chip acting as the engine of the NVIDIA data center platform. 

"AI is at an inflection point, setting up for broad adoption reaching into every industry. From startups to major enterprises, we are seeing accelerated interest in the versatility and capabilities of generative AI," said Jensen Huang, founder and CEO of NVIDIA, in a press release. 

On the back of renewed demand for its AI chips, the firm recently announced revenue for the fourth quarter ended January 29, 2023, of $6.05 billion, up 2 percent from the previous quarter.

A powerful solution for AI-tasks

NVIDIA claims that its $10,000 A100 provides up to 20X higher performance than the prior generation and can be partitioned into seven GPU instances to dynamically adjust to shifting demands. 

The advanced variants are available in 40GB and 80GB memory versions. The firm claims that A100 80GB "debuts the world’s fastest memory bandwidth at over 2 terabytes per second (TB/s) to run the largest models and datasets."

According to the firm, its range of computing solutions ensures "rapid deployment, management, and scaling of AI workloads in the modern hybrid cloud."

NVIDIA's partnerships with cloud service providers 

The firm is partnering with leading cloud service providers to offer AI-as-a-service that provides enterprises access to NVIDIA’s world-leading AI platform. "Customers will be able to engage each layer of NVIDIA AI – the AI supercomputer, acceleration libraries software, or pre-trained generative AI models – as a cloud service," said the firm in a blog post.

The firms utilizing their services will able to engage an NVIDIA DGX™ AI supercomputer through the NVIDIA DGX Cloud, which is "already offered on Oracle Cloud Infrastructure, with Microsoft Azure, Google Cloud Platform and others expected soon." 

The service will also enable customers to access NVIDIA AI Enterprise for training and deploying large language models or other AI workloads. Furthermore, NVIDIA will offer its customizable AI models NeMo and BioNeMo to clients "build proprietary generative AI models and services for their businesses."

The company believes that its range of solutions is ideal for applications like language modeling, recommender systems, image segmentation, translations, object detection, automatic speech recognition, text-to-speech, and high-performance computing.