(JHVEPhoto/Shutterstock)
The SAS Innovate convention drew a worldwide crowd this week in Orlando. The analytics pioneer unveiled new R&D funding plans, new and continued partnerships and integrations, and a benchmarking examine that explored the efficiency, scalability, and prices of its SAS Viya AI and analytics platform.
A principal theme of the occasion was decreasing prices and inefficiencies to optimize enterprise: “Over the previous few years, we now have been examined in sudden methods, from a pandemic, mistrust in provide chains, workforce shortages, monetary instability within the banking sector. And thru that have, we gained a brand new appreciation for resilience. We noticed firsthand the significance of working with velocity and agility. Organizations want expertise that permits them to profit from the information, fashions and workforce they’ve,” stated SAS CEO Jim Goodnight throughout the opening session.
Investing $1 Billion in Analytics
The corporate introduced it can make investments $1 billion over the subsequent three years to proceed growing superior analytics options for particular industries together with banking, authorities, insurance coverage, healthcare, and retail. These options will run on SAS Viya, the corporate’s cloud-native, massively parallel AI and analytics platform.
SAS CTO Bryan Harris chats with CEO Jim Goodnight on stage at SAS Innovate in Orlando. (Supply: SAS)
The newest funding builds on a earlier $1 billion funding in 2019 and consists of direct R&D and {industry} advertising and marketing. The {industry} options will allow knowledge scientists, statisticians, and enterprise analysts to leverage AI, machine studying, and analytics to be used circumstances like fraud detection, public well being and security measures, enhanced affected person care, and manufacturing optimization.
“The billion-dollar funding in our options is basically about how we assist the unimaginable work for our knowledge scientists, statisticians, and builders who’re actively working with {industry} consultants like consultants, system engineers, cloud architects, advertising and marketing, and others. It’s about bringing the entire bundle collectively in order that our clients get the advantage of our funding to assist them cut back time to worth for our software program,” stated CTO Bryan Harris in a press convention on the occasion.
Increasing Partnerships
The corporate featured new and continued partnerships and integrations throughout the convention.
“SAS is partnering in another way than we ever have earlier than with expertise corporations that complement our core AI and analytic capabilities. And for us, it’s about two corporations working collectively to ship the precise capabilities for our clients, on our analysis {dollars}, not in the course of a contract with the shopper. It’s about us working proactively getting the integrations proper and delivering unimaginable outcomes from these two applied sciences coming collectively,” Harris stated throughout the press convention.
(Supply: SAS)
SAS says its companions will play a key function within the rollout of its newest industry-specific options. One instance accomplice is Microsoft. SAS Viya on Microsoft Azure will embody Azure confidential computing capabilities because of AMD EPYC processors with safe encrypted virtualization expertise.
In an interview with Datanami, CTO Harris stated this partnership happened when SAS observed a big intersection between its buyer base and Microsoft’s cloud clients, significantly within the monetary companies {industry}.
“One of many greatest dangers of transferring from on-prem to the cloud is that this concern of compromised knowledge, unencrypted knowledge, particularly in that house,” stated Harris. “So, Azure confidential computing permits knowledge to be fully encrypted all the way in which. For the uncommon second you need to unencrypt and compute one thing, it decrypts and calculates after which re-encrypts.”
The brand new confidential computing capabilities will quickly be out there within the U.S., the Netherlands, and Eire, with a wider rollout deliberate for later, the corporate introduced.
One continued partnership highlighted on the convention is with SingleStore, supplier of a cloud-native, distributed SQL database that helps each transactional and analytic workloads in the identical engine. SAS Viya with SingleStore permits the superior AI and machine studying analytics of SAS to be executed immediately towards relational database tables in SingleStore. Harris defined the mixing prevents pricey and inefficient knowledge replication throughout the computation of analytics workloads. A streaming protocol permits knowledge to be streamed into the in-memory engine and repeat streams as wanted with out storing the total set in-memory.
(Supply: SAS)
“We’ve constructed this extremely modern streaming protocol that permits us to not have to duplicate the information and in addition execute analytics and AI at very, very excessive speeds. They will additionally, like SAS, run on-premises, which is necessary standards for why we partnered with SingleStore, as a result of they permit us to satisfy our clients the place they’re with the place they’re on their cloud journey,” Harris stated.
Harris went on to clarify: “We’re additionally transferring our analytics down into the database. So, if you wish to rating knowledge in a database, you need to rating it, pull it out, carry it to the engine, take a look at the rows on it, rating it, perhaps put a price to it. Properly, what we’re doing is we’re taking our analytic fashions, transferring them down into the database, after which truly scoring native to the precise compute node of the database, which suggests we reduce all this knowledge motion, and we truly can rating it and batch at very, very giant scales.”
Benchmarking the Full Analytics Lifecycle
The SAS Innovate convention additionally unveiled a brand new benchmarking examine that discovered the SAS Viya platform is, on common, 30 instances sooner than industrial and open-source alternate options and scales higher with bigger and sophisticated knowledge. The examine was carried out together with The Futurum Group which in contrast SAS Viya to an unnamed industrial, cloud-based knowledge and AI platform, in addition to open supply alternate options like H2O and SparkML.
The benchmarking additionally discovered that Viya is, on common, 49 instances sooner than a Spark-based knowledge and AI platform throughout all algorithms, knowledge sizes, and workloads, and even discovered that in some circumstances, Viya is as much as 326 instances sooner. The examine additionally confirmed Viya scales higher than alternate options with bigger and extra advanced datasets, and that for mannequin coaching on greater than 300 million knowledge factors, Viya outperforms the alternate options and was the one platform to proceed scaling. Lastly, cost-effectiveness was additionally thought-about, and the examine concluded that customers can decrease prices by greater than 86% with Viya in comparison with the examined alternate options.
(Supply: SAS)
Information scientist and AI knowledgeable Marinela Profi is the product advertising and marketing supervisor for knowledge science and open supply at SAS. She sat down with Datanami to interpret the benchmark examine and defined how organizations transitioning to the cloud usually face enormous prices because of the coaching, improvement, and implementation of AI fashions. These prices are affected by the selection of algorithms and mannequin coaching options, and this examine proves that some algorithms are sooner and extra scalable, Profi asserted.
Profi says the primary differentiator of SAS is its containerized structure based mostly on Kubernetes which is cloud-native and cloud-portable, in addition to the corporate’s in-house R&D that’s centered on optimization for the cloud. Profi advised Datanami that the Futurum Group is presently engaged on a technical report within the curiosity of creating this benchmark fully clear and repeatable.
Profi commented on the design of the benchmark, noting that it takes under consideration the whole analytics lifecycle from knowledge to selections, which she sees as a key distinction in comparison with different benchmarking: “The way in which that the Futurum Group determined to design the benchmarking was to think about every part that occurs from the second you load the information to the second that you simply rating the mannequin,” she stated. “We did the whole analytics lifecycle. We loaded the information, we did some pre-processing of options, we skilled the mannequin, we noticed the accuracy after which we scored that mannequin on new knowledge.”
(every part potential/Shutterstock)
Profi says SAS is presently exploring additional funding into the benchmarking world with plans to develop the benchmark to cowl decisioning and person productiveness in each mannequin improvement and UI. For instance, measuring effectivity for various customers and the variety of on-screen actions required to execute duties.
“I consider that that is the start of an thrilling benchmarking space that we’re setting, and I’m very pleased with how daring SAS is being in making this primary step into this benchmarking and I can’t wait to do extra,” Profi stated. “It’s a accountability for all of us, particularly industrial and open supply platform suppliers, to repeatedly show in a clear means their capabilities and the way they’re higher– not simply with the messaging, but in addition validating it. So, I’m actually enthusiastic about this benchmark. I consider it’s step one in the direction of one thing actually huge and gives our clients a tangible device to choose.”
Keep tuned for extra protection of SAS Innovate.
Associated Gadgets:
SAS Touts One-Click on Entry to Viya on Azure Cloud
SAS Declares Partnership with SingleStore