SoftBank and Intel Partner on New AI Memory Chip, Halving Power Consumption

SoftBank and Intel are partnering to develop an AI-specific memory chip, Saimemory, aiming to reduce AI computing power consumption by up to 50%. This stacked DRAM chip, utilizing an innovative wiring architecture, promises improved energy efficiency and supports sustainable AI data centers. With an estimated $64 million investment, a prototype is expected within two years, targeting commercial deployment in the late 2020s, starting with SoftBank’s AI data centers. Other research institutions and the government are expected to offer support as the project progresses.

“`html

In a move that could reshape the landscape of artificial intelligence, SoftBank Group, the Japanese tech behemoth, and Intel, the chipmaking titan, have announced a strategic alliance to develop a revolutionary AI-specific memory chip. This collaboration aims to tackle the energy-guzzling nature of current AI computing, with the potential to slash chip power consumption by a staggering 50% – a breakthrough with significant implications for the future of global AI infrastructure.

The core of this partnership lies in the development of a novel stacked DRAM chip. This chip will leverage an innovative wiring architecture, diverging significantly from the existing High Bandwidth Memory (HBM) solutions. This cutting-edge design promises not only a dramatic improvement in energy efficiency, but also provides crucial technological support for the greening of AI data centers. Industry analysts are already hailing the potential impact, noting that the successful commercialization of this technology could drastically reduce the operational costs associated with AI computing, thereby contributing significantly to sustainable development goals.

To spearhead this ambitious project, the two giants are forming a dedicated company, Saimemory. This entity will bring together Intel’s core chip technology and patents from Japan’s leading research institutions such as the University of Tokyo, creating a unique intellectual property portfolio and technological advantage.

In terms of operations, Saimemory will concentrate on chip design and patent management, while outsourcing manufacturing to specialized foundries. This collaborative approach is designed to maximize research and development efficiency, allowing each partner to focus on their core strengths.

The project roadmap projects the completion of a chip prototype design within two years, followed by assessment for mass production, with a target for commercial deployment by the late 2020s. The total investment is estimated at 10 billion Japanese yen (approximately $64 million USD), with SoftBank as the lead investor committing to a 3 billion yen (about $19.2 million USD) injection.

Adding further intrigue, research institutions, also show interest in potentially providing financial or technical support. The project participants are also actively seeking government support to create a favorable environment for both technological advancements and industrialization.

SoftBank has made it clear that this groundbreaking memory chip will initially be deployed in its own AI training data centers. As AI technology continues its march into enterprise management, smart manufacturing, and other high-end sectors, the demand for high-performance, low-power computing hardware is poised for explosive growth.

耗电量减半!软银与英特尔合作研发新型AI内存芯片

“`

Original article, Author: Tobias. If you wish to reprint this article, please indicate the source:https://aicnbc.com/1585.html

Like (0)
Previous 4 days ago
Next 4 days ago

Related News