Google (GOOGL) Cloud revealed its eighth generation of custom-made AI chips early this week, to give customers more power for less money. This hardware release strikes at the heart of the industry, where Nvidia (NVDA) currently holds a massive $5 trillion market lead. The search giant wants to offer its own fast alternatives to help businesses build and run advanced AI models more efficiently.
Claim 55% Off TipRanks
Forget margin or options. Here's how the pros trade AMZNGoogle Cloud Rolls Out a Two-Pronged Strategy
Google decided to split its latest technology into two specialized tools to get the best results for its users. One chip, named the TPU 8t, focuses on training large AI models, while the other, the TPU 8i, handles inference. Inference is the work a model does after a user submits a prompt. This specific design allows each chip to perform its designated task with much higher precision.
This new hardware offers a giant leap in performance compared to previous versions. Google says these chips are 3x faster at training AI models and offer 80% better performance for every dollar spent. The company can even link over 1 million of these chips together in a single group to solve massive problems. The firm is providing more computing power while using significantly less energy through this new hardware.
Google Maintains its Complex Bond with Nvidia
Google continues to work closely with Nvidia even as it builds its own custom hardware. These new chips supplement the Nvidia systems already available in the cloud. The search giant even promised that it will offer Nvidia’s newest chip, known as Vera Rubin, to its customers later this year. Both companies are also teaming up on a software project called Falcon to help different computer systems work together more smoothly.
Winning against Nvidia remains a difficult task for any tech company. The chip maker has reached a market value of nearly $5 trillion, and most of the world’s AI still runs on its technology. While giants like Google, Amazon (AMZN), and Microsoft (MSFT) are all building their own custom chips, they still rely heavily on Nvidia to meet the huge demand for AI power. Google’s growth as an AI provider often leads to more business for Nvidia as well.
Is Google a Good Stock to Buy?
Alphabet’s stock (GOOGL) continues to carry a Strong Buy consensus, based on 31 analyst ratings over the past three months. Out of those, 26 call it a Buy, while five recommend a Hold. None of the analysts currently suggest a Sell.
The average 12-month GOOGL price target sits at $387.68, which represents 14.4% upside potential.
