Analyzing data generated within the enterprise, such as sales and purchasing data, can lead to insights that improve business operations. But some organizations struggle to efficiently process, store and use their massive amounts of data. According to an IDC questionnaire On behalf of Seagate, organizations collect only 56% of the available data in their business operations, and of that 56% they use only 57%.
Part of the problem is that data-intensive workloads require significant resources and adding the necessary compute and storage infrastructure is often expensive. For companies moving specifically to the cloud, IDG reports that they plan to spend $78 million on infrastructure this year. Thirty-six percent named controlling costs as their biggest challenge.
That’s why Uri Beitler launched pliops, a startup that develops what he calls “data processors” for enterprise and cloud data centers. Pliop’s processors are designed to improve the performance of databases and other apps that run on flash memory, saving money in the long run, he says.
“It became clear that today’s data needs are not compatible with yesterday’s data center architecture. Massive data growth has collided with legacy compute and storage shortcomings, causing computer slowdowns, storage bottlenecks and declining network efficiency,” Beitler told londonbusinessblog.com in an email interview. “As CPU performance increases, they don’t keep up, especially where accelerated performance is critical. Adding more infrastructure often proves prohibitive and difficult to manage. As a result, organizations are looking for solutions that free CPUs from compute-intensive storage tasks.”
Pliops is not the first to bring a processor for data analysis to the market. Nvidia sells the BlueField-3 data processing unit (DPU). Marvell has its Octeon technology. Oracle’s SPARC M7 chip has a data analytics accelerator coprocessor with a specialized set of instructions for data transformation. And on the startup front, Blueshift Memory and Speedata are creating hardware that they say can perform analytics tasks significantly faster than standard processors.
But Pliops claims to be further ahead than most, with implementations and pilots with customers (albeit not named) including fintechs, “medium-sized” communications service providers, data center operators and government labs. The startup’s early traction, it seems, won investors, who deposited $100 million in the Series D round that closed today.
Koch Disruptive Technologies led the tranche, with participation from SK Hynix and Lip-Bu Tan of Walden International, bringing Pliops’ total capital to over $200 million to date. Beitler says it will be leveraged to build out the company’s hardware and software roadmap, strengthen Pliops’ presence with partners and expand its international workforce.
“Many of our customers saw tremendous growth during the COVID-19 pandemic, thanks in part to their ability to respond quickly to the new work environment and uncertain conditions. Pliops certainly did. While some customers were affected by supply chain issues, we weren’t,” said Beitler. “We see no slowdown in data growth — or the need to leverage it. Pliops was strong for this latest round of financing and now even stronger.”
Speed up data processing
Beitler, former director of advanced memory solutions at Samsung’s Israel Research Center, co-founded Pliops in 2017 with Moshe Twitto and Aryeh Mergi. Twitto was a research scientist at Samsung who developed flash memory signal processing technologies, while Mergi co-launched a number of startups — including two acquired by EMC and SanDisk — before joining Pliops.
Pliop’s processor provides disk error protection for solid-state drives (SSD) and in-line compression, a technology that reduces the size of data by finding identical data sets and then saving only the first set. Beitler claims the company’s technology can reduce disk space while expanding capacity, mapping “variable size” compressed objects into storage to reduce wasted space.
A core component of the Pliops processor is the hardware-accelerated key-value storage engine. In key value databases – databases in which data is stored in a “key value” format and optimized for read and write – key value engines manage all persistent data directly. Beitler argues that CPUs are typically overused when running these engines, so apps don’t take full advantage of SSD’s capabilities.
“Organizations are looking for solutions that free CPUs from compute-intensive storage tasks. Our hardware helps create a modern data center architecture by leveraging a next-generation hardware-accelerated data processing and storage management technology — a technology that delivers an order of magnitude improvement in performance, reliability and scalability,” said Beitler. with Pliops you get more out of existing infrastructure investments.”
The Pliops processor came on the market last July. The development team’s current focus is to accelerate data ingestion for machine learning use cases, Beitler says — use cases that have grown among Pliops’ current and potential customers.
The road ahead of us
Certainly, Pliops has his work ahead of him. Nvidia is a formidable competitor in the data processing accelerator field, having spent years developing its BlueField range. And AMD acquired DPU supplier Pensando for $1.9 billion, signaling its broader ambitions.
One move that could pay off for Pliops is joining the Open Programmable Infrastructure Project (OPI), a relatively new venture under the Linux Foundation that aims to create standards around data accelerator hardware. While Pliops isn’t a member yet — current members include Intel, Nvidia, Marvell, F5, Red Hat, Dell, and Keysight Technologies — it goes without saying that becoming a member could expose its technology to a wider customer base.
Beitler was hesitant when asked about OPI, but pointed out that the data acceleration market is still in its infancy and growing.
“We continue to see both infrastructure and application teams inundated with underperforming storage and overwhelmed applications that don’t meet the company’s data requirements,” said Beitler. “The general feedback is that our processor is a groundbreaking product and without it, companies will have to invest years in software and hardware engineering to solve the same problem.”