Nvidia, Advanced Micro Devices (AMD), and Cerebras Systems are competing to dominate the emerging market for AI inference workloads, which is expected to surpass model training in revenue potential. Nvidia has integrated Groq’s language processing units (LPUs) into its ecosystem to enhance inference speed, utilizing a combination of LPUs and GPUs in specialized server racks. Cerebras offers a unique, large wafer-scale chip solution that provides up to 15 times the speed of standard GPUs, but at a premium price, supported by a significant $20 billion deal with OpenAI.
AMD is bolstering its position in the AI inference market through its recent acquisition of memory optimization software MEXT, which aims to mitigate high costs associated with high-bandwidth memory (HBM) shortages. By optimizing memory usage, AMD’s solutions are designed to be cost-effective while catering to the rising demand for CPUs in data centers as agentic AI workloads increase.
5 Stocks Our Experts Predict Could Double In the Next Year
By submitting your email, you'll also get a free pivot & flow membership. A free daily market overview. You can unsubscribe at any time.
