In a recent press release, IBM announced the development of an artificial intelligence chip that can be applied to machine-learning tasks such as self-driving vehicles and virtual assistants.

The system is designed to provide an easier way for developers to train deep neural networks, decreasing the time it takes to train them from days or weeks down to minutes.

The company called the AIU an application-specific integrated circuit that can do anything deep learning-related, like rapidly processing images or translating spoken languages.

Research and development of the AIU took five years and is designed to handle AI calculations that require matrix and vector multiplication. With AIU's plug-and-play design, deep learning models can be trained and run faster than on a CPU.

Regarding the new chip, IBM said that CPUs and GPUs couldn't keep up with the demands of the developed general-purpose deep learning models.

Though AI models have been more efficient and accurate recently, they're limited compared to modern computer systems.

IBM stated that while some AI models would require devices with the processing power of edge devices and cloud servers, the more complex models will require investments in new hardware AI platforms.

In 2019, IBM established its AI Hardware Centre to optimize and increase AI hardware performance 2.5 times a year. By 2029, the company hopes to have reduced the time required to train and run an AI model by one thousand.

"Deploying AI to classify cats and dogs in photos is a fun academic exercise. But it won't solve the pressing problems we face today. For AI to tackle the complexities of the real world — things like predicting the next Hurricane Ian, or whether we're heading into a recession — we need enterprise-quality, industrial-scale hardware," Said IBM.

Also, read Oracle Is Joining Forces with NVIDIA to Speed up the Adoption of Enterprise AI