Most Read: Google’s AI Search Updates, Nvidia CEO on Generative AI Redefining the Future

Also inside, a PGIM report delves into future data center power consumption, plus Microsoft’s latest generative AI partnership

June 7, 2024

4 Min Read

Here are the most-read stories on AI Business this week:

Google Refines AI Overviews After Bizarre Responses, Limits Content

Google has enhanced its AI Overviews feature with "technical improvements" following a series of viral bizarre responses.

AI Overviews (formerly Search Generative Experience) uses Google’s Gemini model to generate natural-sounding answers to search queries.

Following its recent expansion following Google’s I/O event, users began sharing screenshots of outlandish Overviews responses.

The weird responses included suggesting non-toxic glue be added to pizza, that jumping off the Golden Gate Bridge was a cure for feeling depressed and eating rocks was a good source of minerals.

To address the issue, Google said it has made more than a dozen “technical improvements” including an improved detection mechanism for “nonsensical queries.”

Google has also limited the use of user-generated content in responses.

Read more about how Google has updated AI Overviews

Nvidia CEO: Generative AI, Accelerated Computing to Redefine the Future

Nvidia CEO Jensen Huang said he believes that generative AI and accelerated computing will “redefine the future” during a keynote at the annual Computex event in Taiwan.

“Today, we’re at the cusp of a major shift in computing,” Huang said. “Generative AI is reshaping industries and opening new opportunities for innovation and growth.”

Related:Generative AI in the Spotlight at AI Summit London

Nvidia has positioned itself as a leader in the AI space. Its accelerated computing hardware including its GPUs is sought after by businesses looking to power and scale new generative AI applications and services.

Huang said AI is redefining accelerated computing across consumer-facing AI PCs and enterprise-level computing platforms in data centers.

“The future of computing is accelerated,” Huang said. “With our innovations in AI and accelerated computing, we’re pushing the boundaries of what’s possible and driving the next wave of technological advancement.”

Learn more about Nvidia’s announcements from Computex

AI Workloads to Double Data Center Power Demand by 2026

A PGIM report has warned that the share of global electricity consumption used by data centers could soar from 2% today to more than 20% by 2030.

Data centers use massive amounts of power to run and cool their servers. The Fueling the Future report, suggests global data center power consumption will more than double by 2026, consuming the same amount of electricity as Japan.

According to the report, factors increasing data center energy demands include intensive workloads for training large language models.

Related:Nvidia Hits $3 Trillion Market Cap, Overtakes Apple as 2nd Most Valuable Company

“The exponentially increasing demand for computational capacity to power the training of large language models… will have profound implications for data centers and may be one of the most overlooked aspects of the energy transition,” the report said.

PGIM's report found that demand for data center operations is growing in developed countries that are behind in expanding their power infrastructure.

For example, electricity demand from data centers in Ireland is set to double by 2026, accounting for one third of the country’s electricity demand.

PGIM suggests data center operators should look to balance increasing their compute workloads with their ability to source power.

Explore the report’s findings in detail

AMD New Chips Take on Nvidia, Power AI PC Workloads

AMD has launched new AI-focused hardware offerings at Computex 2024 to continue its rivalry with Nvidia.

The company unveiled new AI accelerators including the Instinct MI325X which will directly compete with Nvidia’s H100 units for powering AI workloads in data centers.

During the company’s Computex keynote, Lisa Su, AMD’s chair and CEO, offered a preview of AMD’s new EPYC server processors, codenamed Turin.

Turin and the new MI325X GPUs will be released in the second half of 2024 — half a year before Nvidia launches its new Blackwell hardware.

AMD also plans to release the MI350 GPU in 2025 and the Instinct MI400 GPUs in 2026, shortening its release cycle to copy Nvidia’s new one-year product road map, which was also unveiled at Computex.

“This is an incredibly exciting time for AMD as the rapid and accelerating adoption of AI is driving increased demand for our high-performance computing platforms,” said Su.

Find out more about AMD’s Computex announcements

Microsoft, Hitachi Sign $18.9B Generative AI Partnership

Hitachi will leverage Microsoft’s AI and cloud services through a multiyear, multibillion-dollar partnership.

Under the three-year deal, Hitachi will integrate Microsoft cloud services, including the Azure OpenAI Service, into its digital solutions business, Lumada.

Lumada will gain access to OpenAI models through the Azure service, which will aid it in building customers' digital solutions.

Lumada will also have access to GitHub Copilot to assist developers with coding tasks and productivity tools through Copilot for Microsoft 365.

The companies estimate the joint generative AI efforts will generate $18.9 billion in revenues for Lumada in 2024 alone.

Read more about Microsoft’s latest generative AI partnership

Keep up with the ever-evolving AI landscape
Unlock exclusive AI content by subscribing to our newsletter!!

You May Also Like