Those who are making a lot of money in this segment are the manufacturers of specialized chips and systems. We only have to look at the case of NVIDIA, which is bathing in gold thanks to this new market that is emerging.
Meanwhile, we see how AI development companies are squandering millions of euros to lead the segment. A torrent of money that they hope to recover in the future, although it is not very clear how. But well, that will be seen in the future.
India already has its first chips created for AI
Ola Electric will be the first Indian company to offer specially designed chips for artificial intelligence. The first three models will hit the market in 2026, with another model coming in 2028.
There is currently a strong demand in the country for solutions to create different systems focused on this segment. The Indian company Yotta, specialized in servers and Data Centers, has an order for 16,000 NVIDIA GPUs for next year. It should be noted that the company has recently received a quantity of 16,000 GPUs.
The company has just announced the Bodhi 1, Ojas and Sarv 1 chips, which will be launched next year. Bodhi 2 will be the successor to the first model and will not be launched until 2028.
Bohdi 1 is a chip designed for AI inference, being an optimal solution for LLM and visual models. The company highlights that this chip offers the best energy efficiency in its class.
Ojas is an Edge AI chip, which is created for more specific applications. The company highlights that it can be used for vehicles, smartphones, IoT and many other systems. Ola also highlights that the idea is to integrate these chips for its future vehicles, running systems such as charging, ADAS and more.
Finally, they have presented the Sarv 1 model, a chip based on the heavy ARM architecture for Data Centers. They have not given much information about this model, but it could be a highly efficient solution.
According to Ola Electric, these new chips offer better performance and energy efficiency than NVIDIA GPUs. Unfortunately, they have not said which NVIDIA solution they have compared them with. They simply pointed out that this GPU operates at 200 watts, so the NVIDIA solution would not be delivering its full potential, predictably.
During the presentation, the company did not reveal who will be the manufacturer of these new chips. Nor did they say what type of node they will use. Although it may seem absurd, these two factors play an important role in the final capabilities of the chip.
What seems clear is that India wants to get involved in the artificial intelligence market. It will not be easy to break into a segment led by China and the United States. The factor that India can take advantage of is that it is the most populous country in the world and has a huge amount of talent.