Presented by SQream
The challenges of AI get worse as it advances: data preparation requirements, large data sets and data quality, time wasted by long-running queries, batch processes, and more. Moreover. In this VB Spotlight, William Benton, Principal Product Architect at NVIDIA, and others explain how your organization can simplify today’s complicated things.
The growing transformative power of AI is hamstrung by a very real challenge: not only the complexity of analysis processes, but also the interminable time it takes to go from executing a query to accessing data. information you are looking for.
“Everyone has worked with dashboards that have a little bit of latency built in,” says Deborah Leff, chief revenue officer at SQream. “But you get to very complex processes where you’re waiting hours, sometimes days or weeks, for something to be finished and get a specific insight.”
At this recent VB Spotlight event, Leff was joined by William Benton, principal product architect at NVIDIA, and scientist and journalist Tianhui “Michael” Li, to talk about how organizations of any size can overcome common obstacles to leverage the power of data analytics at the enterprise level – and why an investment in today’s powerful GPUs is crucial to improving the speed, efficiency and capabilities of analytics processes, and will lead to change paradigm in how businesses approach data-driven decision-making.
Accelerating business analytics
Although there is a lot of excitement around generative AI and it is already having a powerful impact on organizations, enterprise-level analytics has not evolved as much over the same period.
“A lot of people still have analytics issues with the same architectures,” Benton says. “Databases have seen many incremental improvements, but we have not seen this revolutionary improvement that impacts everyday practitioners, analysts, and data scientists to the same extent that we see with some of these perception problems in AI, or at least they did. It did not capture the popular imagination in the same way.
Part of the challenge is this incredible waste of time, Leff says, and solutions to these problems have been prohibitive until now.
Adding more hardware and computing resources to the cloud is expensive and adds complexity, she says. A combination of brains (the CPU) and brawn (the GPUs) is required.
“The GPU you can buy today would have been incredible from a supercomputing perspective 10 or 20 years ago,” Benton says. “If you think about supercomputers, they’re used for climate modeling, physics simulations – big scientific problems. Not everyone has big science problems. But that same massive amount of computing capacity can be made available for other use cases.
Instead of just tuning queries to save a few minutes, organizations can reduce the time the entire analysis process takes from start to finish, thereby increasing the speed of network, ingestion, query and presentation of data.
“What’s happening now with technologies like SQream that leverage GPUs with CPUs to transform the way analytics are processed is that they can harness the same immense raw strength and power that GPUs bring and apply them to traditional analyses. The impact is of an order of magnitude.
Accelerating the Data Science Ecosystem
Unstructured and ungoverned data lakes, often built around the Hadoop ecosystem, have become the alternative to traditional data warehouses. They are flexible and can store large amounts of semi-structured and unstructured data, but they require extraordinary preparation before running the model. To address this challenge, SQream turned to the power and high-throughput capabilities of the GPU to accelerate data processes throughout the workload, from data preparation to insights.
“The power of GPUs allows them to analyze as much data as they want,” says Leff. “I feel like we’re so conditioned: we know our system can’t handle unlimited data. I can’t take a billion rows if I want and look at a thousand columns. I know I have to limit it. I have to sample it and summarize it. I have to do all sorts of things to get it to a workable size. You completely unlock this with GPUs.
RAPIDS, Nvidia’s open source suite of GPU-accelerated data science and AI libraries, also accelerates performance by orders of magnitude at scale across all data pipelines by taking the massive parallelism now possible and enabling organizations to apply it to accelerate Python and SQL data science. ecosystems, adding enormous power under familiar interfaces.
Unlock new levels of information
But it’s not just about speeding up these individual steps in the process, Benton adds.
“What slows down a process? It is communication across organizational boundaries. It’s even communication between people’s offices. It’s the latency and speed of the feedback loops,” he says. “That’s the exciting benefit of accelerating analytics. If we look at how people interact with a mainframe computer, we can dramatically improve performance by reducing the latency when the computer provides responses to the human, and the latency when the human provides instructions to the computer . We get a super linear advantage by optimizing both sides.
Moving to sub-second response speeds means responses are returned immediately and data scientists remain in flow state, remaining as creative and productive as possible. And if you take that same concept and apply it to the rest of the organization, in which a wide range of business leaders make decisions every day that drive revenue, reduce costs, and avoid risk, the impact is deep.
With CPUs as the brains and GPUs as the raw power, organizations are able to realize the full power of their data: queries that were previously too complex, too long, are suddenly possible, and from there, anything is possible. Leff said.
“For me, it’s the democratization of acceleration that’s really the game changer,” she says. “People are limited by what they know. Even on the business side, a business owner trying to make a decision — if the architecture team says, yes, it will take you eight hours to get this information, we accept that. Although it might actually take eight minutes.
“We’re stuck in this model with a lot of business analysis, saying, I know what’s possible because I have the same database that I’ve been using for 15 or 20 years,” Benton says. “We designed our applications around these assumptions that are no longer true due to this acceleration that technologies like SQream are democratizing access to. We need to set the bar a little higher. We have to say, hey, I thought this wasn’t possible because this query wasn’t finished after two weeks. Now it ends in half an hour. What should I do with my business? What decisions should I make that I couldn’t make before? »
To learn more about the transformative power of data analytics, including an overview of cost savings, a dive into the power and insights that are now possible for organizations and much more, don’t miss this VB Spotlight .
Agenda
- Technologies to significantly reduce time to market for product innovation
- Increase the efficiency of AI and ML systems and reduce costs, without compromising performance
- Improve data integrity, streamline workflows, and extract maximum value from data assets
- Strategic solutions to transform data analytics and innovations that drive business results
Speakers:
- William BentonSenior Product Architect, NVIDIA
- Deborah LeffChief Revenue Officer, SQream
- Tianhui “Michael” LiTechnology Contributor, VentureBeat (Moderator)
