Cisco Systems introduces innovative AI networking chips to redefine the AI supercomputers’ landscape with amplified performance and efficiency.
Cisco Systems, a renowned supplier of networking equipment, is pushing its boundaries by stepping into the AI supercomputers market.
The company recently launched networking chips to strengthen its foothold in the sector and offer stiff competition to key players like Broadcom and Marvell Technology.
The SiliconOne Innovation
According to a company statement, the new chips, part of Cisco’s SiliconOne series, are under evaluation by five of the six leading cloud providers.
While the firms’ names have not been disclosed, top players such as Amazon Web Services, Microsoft Azure, and Google Cloud dominate the cloud computing market, per Bofa Global Research.
AI applications’ growing popularity, including platforms like ChatGPT powered by a network of specialized chips called graphics processing units (GPUs), has elevated the significance of individual chip communication speed.
The new Cisco chips, tested by major cloud providers, suggest the company’s strategic move towards catering to this rising demand.
The G200 and G202 Ethernet Switches
Cisco, best known for its networking equipment—including ethernet switches that connect computers, laptops, routers, servers, and printers to a local area network—has unveiled the latest ethernet switches, the G200 and G202.
The company claimed these new entrants bring double the performance compared to their predecessors and can connect up to 32,000 GPUs.
A Cisco fellow and former principal engineer, Rakesh Chopra, described the G200 and G202 as “the most powerful networking chips in the market fueling AI/ML workloads enabling the most power-efficient network.” This indicates Cisco’s ambition to transform the AI/ML workload handling landscape with its new products.
The AI and Machine Learning Revolution
The newly launched chips promise not only enhanced performance but also improved efficiency. Cisco mentioned that the chips could execute AI and machine learning tasks with 40% fewer switches and lesser lag, implying more power efficiency.
This increased efficiency could be a game-changer in a market where power efficiency and speed are highly sought-after, particularly with the surge in AI and machine learning applications.
By providing power-efficient solutions that can handle heavy AI/ML workloads, Cisco is positioning itself as a major contender in this competitive market.
The Competitive Landscape
The launch of Cisco’s AI networking chips comes on the heels of Broadcom’s announcement in April about the Jericho3-AI chip, which can also connect up to 32,000 GPU chips together.
The move signifies Cisco’s strategic attempt to compete with existing heavyweights in the AI supercomputer market.
The race to deliver the best networking chips for AI supercomputers continues as technology advancements accelerate. With Cisco’s new products, the company aims to carve a niche in this competitive landscape, banking on improved performance and power efficiency.
While it remains to be seen how these chips will fare against their competitors, Cisco’s latest venture certainly points towards an exciting future for AI and machine learning applications.