The artificial intelligence (AI) race is getting tighter and tighter as the days go by. You have the success of the likes of OpenAI’s ChatGPT and Google’s Bard, AI tools that an increasing number of people are continuously using. Then, you’ve got companies like NVIDIA creating chips intended for AI and AMD creating GPUs.
But some technology companies, even those prominent ones, seem to be struggling to keep up in the intensifying AI race – such as Intel.
Though Intel produces some graphics processing units (GPUs), experts say these GPUs are still rudimentary. Furthermore, you may also be familiar with its Ponte Vecchio chip suffering years of delays, a reason the company doubled down on enhancing its CPU technology. Intel’s Ponte Vecchio GPU utilizes Generation 12.5 architecture and is created using a 10 nm production process in the company.
However, Intel is determined to keep up in this tightening AI race. The pressure is on for this leading technology company, and it is willing to make an effort also to win the race, such as what it unveiled recently at the ISC High Performance Conference in Hamburg, Germany.
During the conference, Intel announced its supercomputing chip roadmap for the next three years, outlined some of its products bound to compete against AMD and NVIDIA, and provided some details about a chip optimized for AI computing that it plans to introduce this 2025. Read on for more information on this news.
The competition
Today, two of the market leaders and prominent leaders when it comes to products intended for AI are NVIDIA and AMD. For one, NVIDIA’s technology for GPUs has helped it swiftly deliver relevant solutions and scale these up to keep up with the growing demand.
AMD is a prominent player in the GPU market as well. Despite its focus on CPUs and accelerated processing units (APUs), it is still challenging Intel in the market for laptops and desktops. Also, in fact, AMD is slated to challenge NVIDIA with the MI300 chip. MI300 is a sizable chip with 14080 shading units, 880 texture mapping units, and 0 ROPs.
Intel’s Falcon Shores: Will this shift the AI race?
Intel’s shot at the AI race is probably the development of its own products dedicated to AI applications, namely Falcon Shores and the Aurora supercomputer. The company said it has nearly finished the Argonne National Lab’s Aurora supercomputer shipments based on its Ponte Vecchio.
Then, Intel also announced Falcon Shores, its next-generation Max Series GPU, which will provide customers the flexibility to implement system-level CPU and discrete GPU combinations for the demands of the future.
“Intel is committed to serving the HPC and AI community with products that help customers and end-users make breakthrough discoveries faster,” Jeff McVeigh, Intel’s Super Compute Group corporate vice president and general manager, said.
Intel is very willing to correct its past mistakes and beat others in the AI race. What do you think? Will Intel succeed?
