Google vs NVIDIA, Google has announced the launch of its new artificial intelligence supercomputer, TPU v4, which the company claims is faster and more efficient than Nvidia’s AI systems. The announcement was made on Wednesday, highlighting Google’s dominance in the field of AI, as it continues to lead the industry with its Tensor Processing Units (TPUs) that have been in use since 2016.
Google Showcasing its Power
Google has long been recognized as an AI pioneer, with its employees contributing some of the most significant advancements in the field over the last decade. However, the company has been lagging behind in commercializing its inventions, causing concerns that it may have squandered its lead. To address this Google vs NVIDIA, the company has been racing to release products to prove that it still holds the top spot in the field, and the TPU v4 is one of the products designed to achieve this.
AI models such as Google’s Bard or OpenAI’s ChatGPT, which are powered by Nvidia’s A100 chips, require a vast number of computers and chips to train models, with the computers running continuously for weeks or months. In contrast, Google’s TPU v4 can train AI models faster and more efficiently, using fewer power resources.
The TPU v4 supercomputer comprises between Google vs NVIDIA over 4,000 TPUs connected with custom components that enable the machine to run and train AI models. The system has been in use since 2020 and was used to train Google’s PaLM model, which competes with OpenAI’s GPT model. The TPU v4 was found to be 1.2x to 1.7x faster and 1.3x to 1.9x more power-efficient than the Nvidia A100.
There’s a Catch
Despite the success of the TPU v4, Google researchers did not compare its performance with the latest Nvidia AI chip, the H100, which is more recent and was made with more advanced manufacturing technology. Nonetheless, the TPU v4 supercomputer is regarded as the “workhorse” of large language models, thanks to its impressive performance, scalability, and availability.
Summing Up Google vs NVIDIA
The need for large computer power in AI is expensive, making it crucial for industry players to develop new chips, components, and software techniques that reduce the amount of power required. This presents an opportunity for cloud providers such as Google, Microsoft, and Amazon, which can rent out computer processing by the hour and provide credits or computing time to startups to build relationships.
In summary, Google’s TPU v4 supercomputer promises to revolutionize the AI industry, making it more efficient and faster. The product marks another milestone in the company’s bid to retain its position as a leader in AI research and development.