Qualcomm Says AI Needs More Power Than Cloud – Taipei Times

BIG MARKET:
As the growth in the number of devices and data traffic accelerates, it will not be possible to send everything to the cloud, said a Qualcomm executive

Qualcomm Inc is betting that the future of artificial intelligence (AI) will require more computing power than the cloud alone can provide.

The world’s largest maker of smartphone processors is transitioning from a communications company to an intelligent edge computing company, said Qualcomm senior vice president Alex Katouzian.

The benefit in question is the mobile device a user uses to access a network or service, and Katouzian used his time headlining a major main event at the Computex Show in Taipei to demonstrate just how big the market would be.

Photo: Sam Yeh, AFP

The US company’s chips help smartphones harness artificial intelligence for everything from photo processing to malware detection. By emphasizing the AI-supporting aspects of their products, Katouzian and his colleagues have joined a barrage of companies positioning themselves as beneficiaries of a spike in demand for AI.

The company has shipped 2 billion AI-enabled products to date, he said.

As growth in the number of connected devices and data traffic continues to accelerate and data center costs rise, it simply won’t be possible to send everything to the cloud, Katouzian said.

Nor will people want to do that when personal information is involved, he added.

Soaring demand for the chips behind AI tools like Open AI’s ChatGPT is driving shares of chipmaker peer Nvidia Corp to record highs.

Unlike Nvidia, Qualcomm’s outlook fell well short of estimates on weak global demand for mobile devices.

Katouzian sees the depletion of excess inventory in the third or fourth quarter of this year.

Some customers have started increasing orders for smaller components, a leading indicator of increased demand, he said.

The broader enthusiasm for AI has prompted many companies to focus attention on their AI-related services, with more than one executive declaring that a new era of computing had begun.

Arm Ltd, for example, said its technology is enabling many AI applications already taken for granted and would be critical to building the next wave of AI innovations, Arm Chief Executive Rene Haa said in a keynote address on Monday. Computex.

Citing examples from Amazon.com Incs’ Alexa voice assistant and Alphabet Incs’ Google Pixel phones to intelligent traffic light management and robotic beehive maintenance, Haas argued that on-device and small-scale AI processors would all run Arm technology. .

He also pointed to Nvidia’s Grace Hopper next-generation architecture for accelerating AI as another example of Arm technology in the AI ​​supply chain.

The company still intends to go public later this year, a spokesman confirmed on Monday. Arm, looking for a high valuation to please parent company Softbank Group Corp, is trying to present itself as another avenue for investors to tap into this vein of tech optimism.

Given the hunt for AI titles, let’s think about SBG [Softbank Group] may want to pitch Arm as an AI game, although most of its current business is based on semiconductor IP licensing and design [intellectual property]particularly for mobile devices, said Jefferies analyst Atul Goyal.