OpenAI’s Next Project Might Not Sit Well With Nvidia

Powerful GPUs are the backbone of AI, and Nvidia is the main company supplying these chips. Companies’ dependence on the chips is the reason why Nvidia is one of the biggest tech companies in the world. However, it seems that OpenAI wants to break away from Nvidia and start building its own chips.

At this point, we don’t really know anything about this chip or what it could offer. However, based on the company’s plans, it seems that we won’t have to wait too long. OpenAI is planning on moving quickly with its chips.

OpenAI wants to have its own custom AI chips

This news came out of left field, but it shouldn’t be too much of a surprise. OpenAI is aiming to be the definitive AI company. This includes independence, as it is looking to cut its dependence on Nvidia chips. This way, it could have more control over how the chips operate.

While information on these new chips is scarce, we know that OpenAI is not planning to wait around. The San Francisco-based firm has partnered with TSMC, the world leader in chip making, and Broadcom to make this a reality. At the time of writing this, OpenAI is working on finalizing its design before it presents it to be made. This process could take some time.

Be that as it may, OpenAI plans to have these chips hit mass production sometime in 2026. This means that these chips should be in use within the next couple of years. This is if nothing happens in the meantime.

Is this a smart idea?

There’s a reason why companies turn to Nvidia to train their models. This is because the company makes quality chips. Its chips live in millions of computers around the world, and they’re powering many of the AI models we use today. This comes from years of refinement and new iterations.

However, OpenAI will be unveiling a first-generation chip to the market. This means that the quality might not be quite as high. So, this is something that the company will need to consider going forward.