Want more intelligent visions of your inbox? Subscribe to our weekly newsletters to get what is concerned only for institutions AI, data and security leaders. Subscribe now
Since AI applications are increasingly permeating the operations of institutions, from enhancing patient care through advanced medical imaging to the operation of complex fraud detection forms and even to help wildlife preserve, the critical bottleneck often appears: storing data.
During Venturebeat Transfer 2025Greg Matson, President of Products and Marketing, Solidigm and Roger Cummings, CEO of Peak: AIO talks with Michael Stewart, M12’s administrative partner on how to enable innovations in storage technology to use AI Enterprise in health care.
The Monai frame is a boom in medical photography, its construction faster, safer and safer. Progress in storage technology is what enables researchers to build on this framework, repeat and innovate quickly. PEAK: AIO has partnered with Solidgm to integrate energy, performance and high capacity storage that enabled Monai to store more than two million full bodies of CT scans on one knot within their information technology environment.
“As the Foundation’s infrastructure develops quickly, the storage devices are increasingly needing specific cases of use, depending on where they are in the artificial intelligence data pipeline,” Matson said. “The type of use that we talked about with MONAI, the state of the edge use, as well as feeding the training set, is well provided through the storage solutions of the very high capacity solid condition, but actual inference and training the form programming model.”
Improving the conclusion of artificial intelligence on the edge
For peak performance on the edge, it is important to expand the scope of storage to one knot, in order to make the semantic closer to data. What the key is to remove memory bottlenecks. This can be done by making memory part of Amnesty International’s infrastructure, in order to expand its scope with data and descriptive data. The data is close to an account significantly increases the time to insight.
“All huge publishing operations see AI’s large green field data centers, using very specific devices designs to make data as close as possible to graphics processing units,” said Matson. “They have built their data centers by storage in a very high -capacity storage, to bring in abdominal storage, and can be accessed at very high speeds, to graphics processing units. Now, the same technology occurs in a mini image on the edge and in the institution.”
It has become very important for buyers artificial intelligence systems to ensure that you get the most performance of your system by operating it on all solid condition. This allows you to bring huge amounts of data, enable incredible processing force in a small system on the edge.
The future of artificial intelligence devices
“It is necessary to provide open, developed and memory solutions, using some of the latest and largest technologies to do this,” said Kamings. “This is our goal as a company, to provide this openness, that speed, and the scale the organizations need. I think you will see economies that match it too.”
For the general training and inference pipeline, and within the same reasoning, the needs of the devices will increase, whether it is a high -speed SSD or an effective high -capacity solution in energy.
“I would like to say that it will move further about it, about a very high capacity, whether it is the educational SSD one from now from now passing a very low force that can replace four times the hard drives, or a very high -performance product near almost memory speeds,” Matson said. “You will see that adult GPU sellers are looking at how to determine the following storage structure, so that it can help in a close, hbm in the system. The SSD for general purposes in cloud computing is now turning into power and performance.
https://venturebeat.com/wp-content/uploads/2025/07/VBTRANSFORM25-0204-X2.jpg?w=1024?w=1200&strip=all
Source link