Virtualization Technology News and Information
VMblog Expert Interview: Navigating Data Store Challenges in 2023: Insights from Hao Zhong, CEO of ScaleFlux


In this exclusive VMblog Expert Q&A, we take a look at the ever-evolving landscape of data storage in 2023. In this insightful discussion, we have the privilege of sitting down with Hao Zhong, the esteemed CEO and founder of ScaleFlux, a leading innovator in storage solutions. With a wealth of knowledge and expertise, Zhong offers unique insights into the emerging trends shaping the industry, the challenges faced by data stores today, and the state of computational storage.

Please read as we embark on a journey to uncover the future of data storage and the transformative potential of computational storage solutions.

VMblog:  Can you share your high-level view of some emerging trends you're seeing this year in data storage?

Hao Zhong: 

  • As we enter the next wave in edge computing, its success largely depends on its cost-effectiveness.

As computation and storage move closer to data sources, edge computing significantly improves end users' quality of service (QoS). But, adapting to the complexities of massive data growth introduces new challenges. Heterogeneous and domain-specific computing becomes critical in maintaining the cost-effectiveness of storing, moving, and working on this wave of data generation. Computational storage is a crucial component in this re-architecture of the compute infrastructure. As edge computing and the cloud complement each other, forming a foundation for a pervasive IT infrastructure, distributed, domain-specific computing technologies, such as Computational Storage, Smart NICs, and GPUs are critical to scaling that new IT infrastructure. In the coming years, we can expect the continued growth of the cloud in the form of centralized data centers and the rapid expansion of edge computing.

  • Data center sustainability requirements drive storage innovation.

As enterprises evaluate the total costs of their infrastructure, there is an increased focus on improving sustainability along with price and performance. We see enterprises adding environmental footprint calculations to their infrastructure decision-making. Newer and more efficient data centers have the edge over aging facilities, and storage is an obvious resource to examine closely in this pursuit. In write-intensive environments, improvements in the lifespan of SSDs reduce waste and carbon footprint. Additional savings in power can be made by reducing unnecessary processing on the CPU, delivering on the performance and cost demands in the modern data center.

  • Plug-and-play SSD options drive Computational Storage adoption by mainstream customers.

Commodity flash storage is not up to the task for performance-intensive workloads in the modern data center. But finally, customers have an alternative to alleviate these issues by offloading some functions from the server's central processing unit (CPU) to improve the utility of their servers, including better endurance performance, higher capacity, and cost savings. Faced with the reality that "scale-out" is not a long-term sustainable strategy, customers struggling to optimize their data centers realize that physical space is at a premium and every additional node triggers a cascade of costs and architecture challenges. As a result, we expect to see the adoption of easy-to-use computational storage grow significantly through 2023 to help customers "scale-in."

VMblog:  What are some of the data storage challenges organizations face in 2023?

Zhong:  Cost improvements without compromise are the current attitude of enterprise customers in 2023, and who can blame them? Rethinking choices that felt immutable only six months ago is necessary as the economy contracts and competition continues to heat up. As painful as it is, this also creates an opportunity for startups like ScaleFlux to move in. One of the most interesting new trends is a wave of cloud repatriation surges through the market as customers look to optimize their costs and performance of workloads like high-performance transactional databases.

VMblog:  Where are we with computational storage? What is the state of the industry?

Zhong:  Now that there are low-complexity, NVMe-compatible options available, computational storage adoption is accelerating into the mainstream. Once only a reality for sophisticated customers with exotic use cases, computational storage is now a real option for anybody trying to modernize and optimize their data center. Significant trends like edge computing are driving some of this growth. The first killer feature baked into the technology is transparent compression. Flipping traditional thinking on its head, hardware-based compression changes compression from a performance limiter into a performance accelerator. It adds multiples of better latency, endurance, and capacity to SSD's capabilities, creating better overall system performance and TCO improvements.

VMblog:  Can you share some of the fundamental problems that mainstream enterprise customers use computational storage to solve and the benefits they achieve?

Zhong:  High-performance databases, edge workloads, and AI process more data faster than ever, but customers seeking to optimize the economics of their data centers run into the limits of scale-out architectures as costs, space, uptime, performance, and licensing all suffer. Customers want an easy way to address all these things without adding complexity, and looking at your SSDs is an excellent first step.

VMblog:  Where is this all going? What's next?

Zhong:  SSDs integrating computational storage capabilities will continue multiplying performance, endurance, and capacity. Still, there is a long roadmap of features coming to the industry as more storage processing gets pushed down into the drive for security, database enhancements, and reliability, especially as more ISVs get on board and add support to their products.


Published Monday, June 05, 2023 7:32 AM by David Marshall
Filed under: ,
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<June 2023>