Despite odds and challenges amid tightening US restrictions, Huawei Technologies continues to carry on innovation efforts as the company recently announced its 2023 OlympusMons Award challenges, which seek the best solutions for "storage technologies that deliver ultimate per-bit cost efficiency" and "data enablement and resilience technologies for emerging services."
The OlympusMons Award was established in 2019, and there are two OlympusMons Award winners and five OlympusMons Pioneer Award winners each year.
Each OlympusMons Award winner will be awarded CNY1 million (about US$137,883), and each OlympusMons Pioneer Award winner will receive CNY200,000 (about US$27,479) for their work. The bonuses are allocated over two years, with 50% of the funding being disbursed to each winner every year.
According to the Huawei website, OlympusMons Awards were set up to "drive global research in foundational data storage theories, overcome crucial technical challenges, accelerate the industrialization of scientific research achievements, and foster a mutually beneficial collaboration between industry, academia, and research." The OlympusMons Awards have gained significant acclaim in the industry, with more than 180 scholars from over 60 scientific research teams participating since their establishment.
Huawei uses its OlympusMons challenges to prepare for its future solutions to meet the data storage demand in the AI era, aiming to provide emerging services to facilitate data enablement and resilience. It should be noted that it is the fourth year in a row for "Data Storage with Ultimate Per-Bit Cost Efficiency" to remain one of the challenges. High bandwidth memory (HBM), known for its rapid growth thanks to AI demand, presents challenges due to its multilayer stacking, making it 30 to 40 times more expensive than regular servers in AI applications. The cost per bit for HBM is estimated to be 20 times that of regular DRAM, maintaining a substantial gap of 5 to 7 times even in the long term. Huawei's purpose in seeking an ultimate per-bit cost-efficiency solution may be targeting an alternative solution for HBM.
The other challenges over the three years between 2020-2022 are: "Data Infrastructure with Maximum Energy Efficiency," "Foundational Technology Breakthrough in the Next-Generation Storage Industry," and "Achieving the self-driven full-lifecycle data management."
The solutions of past award winners include: "Key technologies of highly reliable and high-performance distributed storage systems," "Key technologies of embedded resistive random access memory (RRAM) on advanced nodes," "Phase-change memory (PCM)," "Persistent Memory Storage System Construction and Key Technologies", "Data-Centric & Data-Driven Storage System Design for High Performance, Efficiency and Reliability," to name just a few.
Is Huawei paving the way for its ambition to enter data storage or memory space? Time will tell. With the memory market being an oligopoly dominated by Samsung, SK Hynix, and Micron Technology, which focuses on standardized products, Huawei is perhaps aiming for memory innovations in a niche direction.