TDK Ventures adds AI accelerator Groq to its investment portfolio
TDK Corporation announced that subsidiary TDK Ventures Inc. adds Groq to its growing investment portfolio, a world leader in the development of artificial intelligence and machine learning platforms that offer twice the inference performance and drastically reduced infrastructure costs. TDK Ventures Inc. was launched with a mission to invest in innovative startups that focus on digital and energy transformation and contribute to sustainability. As TDK Ventures’ sixth investment since its July 2019 launch, Groq is its first investment in AI chipset solutions, with a solution that drastically drives energy requirements down at the highest compute power possible.
Data centers use an estimated 200 terawatt hours (TWh) each year, or about 1% of global electricity demand, and contribute 0.3% to overall carbon emissions. The carbon footprint for data center construction, upkeep and use, is set to reach double-digit percentage of the world’s electricity by 2025.
“In line with TDK’s environmental transformation mission, it is critical we support and invest in solutions that drastically reduce energy needed for very high levels of compute power,” said Nicolas Sauvage, Managing Director, TDK Ventures. “Groq brings the highest AI compute for a given amount of energy, typically measured as Tera Operations Per Second (TOPS) per watt, demonstrated by its first product, the TSP100, with further projected improvement in their next gen platform.”
Groq’s TSP100 solution is ideal for customers who need predictable, reliable, low-latency. automotive ADAS solutions that can react quickly to changing road conditions, pan across connectivity with 5G, support low-latency sensor fusion, and require heavy in-car local computing. Groq’s runtime latency and throughput, known at compile time and without the need to profile code nor worry about power spikes, is a clear benefit to novel use-cases such as ADAS, that require incredibly reliable and predictable performance within a given latency budget, sometimes mandated in SLAs.
“Groq’s architecture enables a whole new class of innovations in power and energy,” said Jonathan Ross, CEO of Groq. “Through its investment and support, TDK is helping Groq bring these capabilities to customers concerned about ML compute power demands and energy consumption. We’re excited to work with TDK to deliver uniquely-scalable, high-performance ML, while driving sustainability and lowering total cost of ownership. We’re delighted that TDK has recognized the power and potential of Groq, making them an ideal partner to continue deepening integration of our technology across automotive, data center and other key industry verticals.”
TDK Ventures specifically sought out AI inference chipset solutions with the highest TOPS/Watt performance metrics, as inference use-cases exponentially increase. “The world cannot wait for 20% of our energy to be used for inference use-cases,” Sauvage added. “Groq offers twice the inference performance while drastically reducing infrastructure costs. At TDK, we believe that we need to invest right now for an attractive future tomorrow and Groq technology supports this strategy.”
The article is originally written by @SemiMediaEdit posted on SemiMedia. All rights reserved.