Vast Data sets sights on analytics, AI
Vast Data Grows: Database Management, Analytics, And AI Focus
Date of Publication: August 8, 2023
Vast Data is bringing storage and data management closer together with a fresh product, bridging the divide between them as artificial intelligence (AI) rapidly increases its presence in the market.
Vast Data's Universal Storage, a file system that utilizes flash technology and allows for extensive sharing and scalability, has undergone a transformation and is now known as Vast Data Platform. This upgrade includes the addition of features related to database management and analytics, expanding its existing storage capabilities. It's worth mentioning that the introduction of these new features does not affect the pricing of Vast's pure storage offering. Furthermore, customers who are satisfied with Vast solely as a storage provider are not obligated to utilize the newly added features. This development is occurring during a time of immense interest in the field of AI, prompting a reevaluation of storage and its future role.
According to IDC analyst Dave Pearson, the areas of advanced analytics that are currently gaining the most attention and experiencing rapid growth include AI, gen AI, machine learning, and HPC (high-performance computing).
According to Pearson, although they might be the most captivating, they do not pose the greatest difficulties in terms of capacity or constitute the majority of storage usage scenarios.
AI Takes Center Stage In Expanding Storage
Max Tegmark, an AI researcher at MIT and professor of physics, highlighted at the recent Build Beyond event by Vast that society has always progressed by utilizing human intelligence. He further emphasized that artificial intelligence has the capability to enhance this progress significantly, potentially leading to faster solutions. To fully harness AI's potential, it is vital to discover innovative methods of data processing.
Tegmark emphasized that the crucial aspect is the way information is processed, disregarding everything else.
At first, Vast offered a marketplace for storing files and objects and a storage area that improved the effectiveness of quad-level cell flash memory by including additional functions like safeguarding data and encoding it.
During the Build Beyond event, Renen Hallak, the co-founder and CEO at Vast, announced that the company has integrated a table and the standard query protocol into their Vast Data Platform. These recent enhancements now enable the execution of SQL queries on file and object metadata.
According to Jeff Denworth, who is the co-founder and chief marketing officer at Vast, this capability allows for the swift analysis and handling of unorganized information. This is extremely important for artificial intelligence tasks since unorganized data serves as the foundation for various advanced deep learning methods.
According to the speaker, the DataStore is a top-notch storage solution provided by the vendor. It is equipped with NAS protocols and a network file system protocol that enables efficient parallel processing of files. Additionally, it offers support for High-Performance Computing (HPC) and is compatible with emerging Artificial Intelligence (AI) workloads.
According to Denworth, AI has a significant impact, and storage remains an essential part of the infrastructure. Although Vast Data is labeled as a storage provider, their primary concern is finding the most efficient way to store data for effective application usage. The data stored on popular cloud platforms like AWS, Azure, Google Cloud Platform, and Oracle holds great importance to companies but none of these providers are primarily known for their storage offerings.
"Storage is the core component of the system, however, the system encompasses much more than just storage," as stated by Denworth.
Expanding Market, Steady Storage Growth
According to Denworth, the shift from Universal Storage to the Vast Data Platform was a growth of the existing technology provided by Vast. This technology allows for the integration of different layers rather than a complete change in direction.
"Scaling is the most straightforward approach to enhance a business, and the same principle applies to the product [like this] -- it's a broadening of functionalities and the target audience," commented Denworth.
Steve McDowell, an analyst and founding partner at NAND Research, states that the new data platform seamlessly aligns with Vast's existing business model. Vast is transitioning from a focus on significant, rapid storage capabilities to a close integration of analytics and artificial intelligence that harness the power of fast storage.
According to McDowell, the integration of analytics systems is intricate, and the relocation of data in proximity to the analytics process might result in enhanced performance.
"He mentioned that this streamlines data analysis by condensing the system."
According to McDowell, although the storage industry as a whole will not completely transition to offering both data management and storage services, there will be other companies like Vast that will offer products focused on making analytics easier. McDowell predicts that the companies to follow Vast's lead will likely be file system vendors such as Weka, Panasas, or IBM.
According to Pearson from IDC, the launch of Vast Data Platform signifies the vendor's effort in aligning itself with the direction it believes the market is heading towards. IT excels in storing vast quantities of disorganized data and is proficient in working with relational databases. Merging these two capabilities would be the natural progression.
According to Pearson, our current focus in advanced analytics is combining those two elements in a way that surpasses the usual constraints of each one.
New & Old: Perfect Harmony
According to Denworth, the fusion of analytics and storage has the potential to revolutionize data storage. With the growing prevalence of AI, businesses will discover fresh requirements for data that would typically be stored away. As an illustration, he noted that deep learning thrives on exposure to greater amounts of data, leading to improvements in learning models.
According to Denworth, the traditional methods of storage may not be suitable for the evolving trends in applications.
According to Pearson, the way a company stores its data is still determined by how it is used. There are still numerous instances where the primary factors for storage decisions are the size, capacity, and long-term preservation of data. Although managing data is a crucial aspect of global data storage, it does not overshadow the importance of other storage tasks. Throughout the past two decades, the significant catalyst for storage has been the continuous growth of data. Therefore, data management is not the ultimate objective in the field of storage.
He stated that the expansion of data is the increasing wave that is boosting all vendors in the storage industry. However, identifｙｉng the specialized areas in storage that offer tangible benefits to businesses will set vendors apart and enhance their earnings.
According to McDowell, a particular area of interest is generative AI, and storage providers are exploring ways to align themselves with it. Vast has already made progress in this regard, particularly with its exceptional performance and file system. On the other hand, other vendors will concentrate on providing data to AI systems. The Vast Data Platform efficiently organizes metadata to enable fast retrieval and transfer to analytics platforms. Even though this product is intriguing, it caters to a specific market segment. Currently, the demand for integrated storage and analytics solutions is limited in size.
Adam Armstrong is a writer for TechTarget Editorial who reports on hardware related to file and block storage, as well as private clouds. He has prior experience working at StorageReview.com.
