AMD's Announces A New AI GPU Chip To Take On Nvidia's Market Dominance!

Move over, potato chips; there is a new chip in town!

As defined by CSET, "AI chips" encompass specialized hardware for AI computations, such as graphics processing units (GPUs), field-programmable gate arrays (FPGAs) and some forms of application-specific integrated circuits (ASICs).

While PC sales (a typical driver of semiconductor processor sales) has taken a dive, AMD aims to get in on the AI game with one of the semiconductor industry's bright spots – AI GPU chips!
Recently, AMD released a new AI GPU processor known as the MI300X that is optimised for Large Language Models and other state-of-the-art AI models, thanks to its CDNA (Converged Data Network Architecture).

The MI300X supports up to 192GB of memory, making it capable of running AI models that would otherwise require significantly larger processors. The memory limit of Nvidia's competitor, H100, is 120GB. #AMDTakesTheLead

There is a correlation between the size of the language model and the amount of memory it consumes in Generative AI applications. The AMD MI300x was shown off while running the Falcon model, which has 40 billion parameters. For reference, there are 175 billion parameters in OpenAI's GPT-3 model!

AMD CEO Lisa Su said, “GPUs are central to this, as they are enabling generative AI. Model sizes are getting much larger and you actually need multiple GPUs to run the latest large language models, noting that with the added memory on AMD chips, developers wouldn’t need as many GPUs using the new chip.”


AMD's AI GPU chip may think it's the best in the business but Nvidia is a beast that’s been wearing the AI crown and sipping on silicon smoothies for a long time now.

According to Statista, Nvidia is widely regarded as a pioneer and leader in the field of accelerated computing and generative AI applications. Nvidia's revenue in the fiscal year 2023 was over $27 billion of which Nvidia spent a record $7.3 billion on research and development to expand its product portfolio. This would surely ensure that Nvidia continues to dominate in the AI industry.

Kevin Krewell, principal analyst at TIRIAS Research, said AMD must catch up to Nvidia on software and research trends if they want a chance to shine. "They have to connect with the research community to understand what the market is doing and what they need to do, well ahead of when something becomes important."


Is AMD's new AI GPU a fierce contender, or will Nvidia steamroll it into a minor silicon speed bump? Who will emerge victorious in this epic clash of the titans? Let us know your thoughts in the comments section below!