On April 29, 2025, Huawei presented the AI Data Lake Solution on the 4th Huawei Innovative Data Infrastructure (Idi) Forum in Munich, which is supposed to accelerate the introduction of AI in all industries. Peter Zhou, Vice President of Huawei and President of Huawei Data Storage Product Line, presented the solution in his keynote: “Data Awakening, Accelerating Intelligence with Ai-Ready Data Infrastructure.”
Even if the digital transformation has developed over decades and has brought profound changes, one thing remains constant: the crucial importance of data. This was emphasized in the speech by Zhou: “To be ready for AI, you have to be ready for data. The continuous deepening of digitization of industry is a process of conversion of data into information and knowledge.”
Through the integration of data storage, data management, resource management and Ki toolchain, the AI Data Lake Solution delivers high-quality AI-Corpus and accelerates the model training and the inference, which enables companies to use AI.
In his speech, Zhou dealt with the products and technologies that are part of the data Lake solution:
- Storage of data: continuous innovation in terms of performance, capacity and resistance
- Accelerated AI model training and inference: The powerful AI memory of the Oceanstor A-series from Huawei offers extraordinary performance. Among other things, the developer of AI technologies Iflytek was able to significantly increase the efficiency of cluster training. The advanced technology for inference acceleration improves inference performance, reduces the latency and improves the user -friendliness of the application – and thus accelerates the use of inference applications with large models in production environments.
- Efficient storage of AI mass data: The Oceanstor Pacific Scale-out memory with all-flash technology offers a high capacity density of 4 PB/2 HE and an extremely low power consumption of 0.25 W/TB. It is designed for the simple management of data in the Exabyte area and is ideal for data-intensive workloads in the areas of education, scientific research, medical imaging and media.
- Securing AI-Corpus and Vectord Database: The OceanProtect-Backup memory from Huawei offers ten times higher backup performance than other common solutions and is characterized by a recognition accuracy of 99.99 % in ransomware attacks. He thus protects important data from training corpora and vectorship banks in areas such as oil and gas as well as MSPs.
- Data management: data transparency, manageability and mobility across regions
Huawei DME is a data management platform that integrates the Omni Dataverse to support customers in the elimination of data silos in geographically distributed data centers. In addition, the ability of DME helps to access data from over 100 billion files in a matter of seconds, to efficiently process data to process data and to exploit their full potential.
- Resource management Merging of various XPUS and intelligent planning of AI resources
The DCS platform is based on virtualization and container technologies and offers efficient XPU resource pooling and intelligent planning, which increases resource use. In addition, the datamaster in DME enables a AI-based O&M with AI copilot that offers a number of AI applications such as intelligent questions and answers, an O&M assistant and an inspection expert and thus creates an extraordinary O&M experience.
Photo – https://mma.prnewswire.com/media/2676754/image_5011010_26316166.jpg
View original content:https://www.prnewswire.com/news-releases/huawei-veroffentlicht-ai-data-lake-solution-zur-beschleunigung-der-branchenintelligenz-302443966.html
OTS original text press release with the exclusive in terms of content of the sender – www.ots.at | Prn