Nearly 25 years after value-based care models began, Change Healthcare’s just-released survey showed 100% of respondents are saving money using those principles.
The survey of 120 insurance company payers is Change Healthcare’s third effort since 2012 to understand what works — and what doesn’t — when it comes to value-based care models.
The universality of savings was a surprise to Andrei Gonzales, M.D., director of value-based reimbursement initiatives at Change Healthcare, a provider of payment, data analytics and a variety of other healthcare platforms based in Nashville, Tenn. “We expected medical cost savings,” Gonzales said, “but to have 100% say they achieved savings and with almost 25% having cost savings of over 7.5% was surprising. We’ve seen savings in our own work, but we wanted a more objective and quantifiable view of the impact of these programs.”
But Gonzalez wasn’t surprised payers were struggling to innovate and to move quickly. Just 21% said they could roll out a new value-based care model in three to six months, while 13% said they needed a full two years.
“Agility in healthcare is difficult, but it is possible,” he said. “With more experience, this should get easier.”
Andrei GonzalesM.D., director of value-based reimbursement initiatives at Change Healthcare
Payers were also frustrated with the technology underpinnings of their value-based care models, Gonzales said. More than half of those surveyed said their analytics, automation and reporting capabilities just weren’t doing the job. In many cases, these tools were custom-built specifically for the value-based care models, Gonzales explained. But as more and more initiatives roll out, the services can struggle to keep up.
The other struggle is to get hospitals and physicians on board with value-based care models, he said. “It’s not surprising to us how difficult it can be to engage providers in a value-based care model,” he said.
Busy doctors and staff can do a good job with patient care in their particular silo, but have a hard time looking across the spectrum of care, he said.
“But once they understand the problem and can see the data that they’re not doing as well as a competitor is doing and that they have to compete for patients, then they start to understand. That’s really where the work is right now.”
DataDirect Networks has refreshed its Storage Fusion Architecture-based ExaScaler arrays, adding two models designed with nonvolatile memory express flash and a hybrid system with disk and flash.
In a related move, the high-performance computing storage vendor acquired the code repository and support contracts of Intel’s open source Lustre parallel file system for an undisclosed sum. The Lustre file system is the foundation for DDN ExaScaler and GridScaler arrays.
The fourth version of DDN ExaScaler combines parallel file storage servers and Nvidia DGX-1 high-performance GPUs with Storage Fusion Architecture (SFA) OS software. SFA 200NV and SFA 400NV are 2U arrays, with slots for 24 dual-ported nonvolatile memory express (NVMe) SSDs. The difference between the two is in compute power: SFA 200NV has a single CPU per controller, while the SFA 400NV has two CPUs per controller.
The arrays embed a 192-lane PCIe Gen 3 fabric to maximize NVMe performance. DDN claims the dense ExaScaler flash ingests data at nearly 40 GBps.
DDN also introduced the SFA7990 hybrid system, which allows customers to fill 90 drive slots with enterprise-grade SSDs and HDDs.
AI and analytics performance driver
Adding NVMe is a natural fit for DDN, which provides scalable storage systems to hyperscale data centers that require lots of high-performance storage, said Tim Stammers, a storage analyst at 451 Research.
“NVMe is going to help drive performance on intensive applications, like AI and analytics. It makes storage faster, and in return, AI and analytics will drive the takeup of NVMe flash,” Stammers said.
Data centers have the option to buy DDN ExaScaler NVMe arrays as plug-and-play storage for AI projects. The DDN AI200 and AI400 provide as much as 360 TB of dual-ported NVMe storage in 2U. The 4U AI7990 configurations scale to 5.4 PB in 20U.
The AI turnkey appliances include performance-tested implementations of Caffe, CNTK, Horovod, PyTorch, TensorFlow and other established AI frameworks.
Customers can combine an SFA cluster with DDN’s NVMe-based storage. Lustre presents file storage as a mountable capacity pool of flash and disk sharing a single namespace.
The DDN ExaScaler upgrade provides dense storage in a compact form factor to keep acquisition within reach of most enterprises, said James Coomer, vice president for product management at DDN, based in Chatsworth, Calif.
“At this early stage, customers don’t necessarily know where they’re going with AI,” Coomer said. “They may need more flash for performance. For AI, they need an economical way to hold data that’s relatively cold. We give them a choice to expand either the hot flash area or augment it in the second stage with hard-drive tiers and anywhere in between.”
Recent AI enhancements to the SFA operating system include declustered RAID and NVMe tuning. Declustered RAID allows for faster drive rebuilds by sharing parity bits across pooled drives.
Inference and training investments planned
DDN’s Lustre acquisition includes the open source code repository, file-tracking system and existing support contracts from Intel. Coomer said DDN plans to make investments to enable Lustre to support inference and training of data for AI workloads. The open source code will remain available for contributions from the community.
DDN is a prominent contributor to Lustre code development, and it has shipped Lustre-based storage systems for nearly two decades.
“DDN says they’re going to make Lustre easier to use,” Stammers said. “What they’re banking on is that it will lead more enterprises to use Lustre for these emerging workloads.”
BOSTON — Recent advancements in the iOS development community aimed at simplifying AI models with Swift have opened up the potential of mobile app machine learning.
Users have come to expect mobile app machine learning to be a part of every technological interaction they have on their phones, and developers should consider implementing machine learning into their apps’ features. Here at this week’s SwiftFest event, developers discussed how machine learning reached this point, the future of mobile app machine learning with Apple’s Swift development language for iOS and how it can be applied.
Why does machine learning matter for mobile?
By creating a machine learning model, mobile app developers can create applications to do more without developers individually programming every action and reaction. Machine learning technology can see the rules that shape a pattern and predict future events, while people are limited by their own imaginations, said Ray Deck, CTO of Element55, a time-tracking software provider in Cambridge, Mass, in a session.
The challenging part of creating an AI model in the past has been collecting the quantity of examples a machine needs to correctly identify what it is seeing. For example, if the technology needs to correctly identify one person from an image, it would need to collect a proportional number of images to the size of the neural-network model developers were creating.
About five years ago, a breakthrough in organizing neural networks — deep learning — created a faster way to write models more accurately. This opened the way for computers to begin to accurately predict patterns or identify subjects through machine learning models.
“If we just try to write these models ourselves, we won’t get it right,” Deck said.
Why could Swift be the future of AI?
The future of machine learning may be on the side of Swift developers. With the release of several new software development frameworks — Swift for TensorFlow and Apple’s own Core ML 2 and Create ML — developers do not need to know as much to incorporate mobile app machine learning.
“Machine learning is more accessible with the latest releases of iOS that they have been doing, and it invites me to explore more and try to use some of that technology in our apps,” said Jaime Santana Ruelas, a software engineer at Cisco.
Ray DeckCTO of Element55
In March, the Swift for TensorFlow team at Google announced its open source project. Python has been leading the way in TensorFlow, despite TensorFlow being written in C++ — a variant of Objective-C, which lends its runtime library to Swift. Creating models with Python is slow, however, and with Swift for TensorFlow, developers can have more creativity when building AI models, Deck said.
“You get that high-level language experience of Swift and that compile performance associated with the runtime, creating a more natural connection, because you are compiling straight into [TensorFlow],” he said.
This month, Apple announced Create ML and Core ML 2 to simplify the creation and implementation of app machine learning models. Create ML enables developers to create machine learning models more easily in Swift through more of a drag-and-drop experience. Plus, developers don’t need to have as much technical knowledge to use Create ML. Core ML 2 boasts faster processing speeds and a smaller model size to implement AI models into apps.
“Swift is defining a new golden path of usability for consumption and creation and potentially advancing the vanguard of automatic differentiation,” Deck said. “The most powerful models may yet to come.”
What can mobile app machine learning do?
In an interview after the session, Deck said app machine learning has been growing based on two factors: the supply increasing quickly due to better techniques developed to create AI and the demand users have for the promise of AI.
“The promise of AI is that we’re carrying not just a camera, but an eye in our pocket, [for example], and being able to have software make decisions based on what we see or an advanced understanding of it,” he said. “It helps people make better decisions.”
People already use AI technology in their fitness watches. Mobile apps could take this further by aggregating data to predict health risks and warn users if they are following a path that models previously predicted would lead others to be taken to the emergency room. In the enterprise, mobile app machine learning could help business travelers get a ride or put email messages in spam folders.
App machine learning can also allow devices to respond to people’s voices. Martin Mitrevski, a technical lead at Netcetera, a software company in Switzerland, works with AI to create conversational user interfaces that can complete tasks, such as creating a list from voice commands.
“Anything you can imagine can be made smarter with AI,” Mitrevski said. “Pretty much any industry will be disrupted with AI and machine learning.”
Currently own a GTX 660 and want to upgrade to the following models only:
GTX 780 Ti
Please let me know what you have.
This message is automatically inserted in all classifieds forum threads.
By replying to this thread you agree to abide by the trading rules detailed…
For Sale due to upgrade, I have a number of large NAS specific drives for sale:
2 x 6TB Models – under warranty until October 2018. £160 each
4 x 8TB Models – under warranty until 4th April 2019 £240 each – NOW SOLD
1 x 10TB Seagate IronWolf Pro. Brand New unit, purchased as I was about to give up waiting for the WD Red 10TB units… then they arrived! These come with a 5 year warranty as standard, 7200rpm, 256Mb cache and 2 years data recovery. http://www.seagate.com/gb/en/internal-hard-drives/hdd/ironwolf/£320 – NOW SOLD
Price and currency: See Below Delivery: Delivery cost is included within my country Payment method: PPG (Preferred) or BT Location: Bromsgrove Advertised elsewhere?: Not Advertised elsewhere Prefer goods collected?: I have no preference
______________________________________________________ This message is automatically inserted in all classifieds forum threads. By replying to this thread you agree to abide by the trading rules detailed here. Please be advised, all buyers and sellers should satisfy themselves that the other party is genuine by providing the following via private conversation to each other after negotiations are complete and prior to dispatching goods and making payment:
Landline telephone number. Make a call to check out the area code and number are correct, too
Name and address including postcode
Valid e-mail address
DO NOT proceed with a deal until you are completely satisfied with all details being correct. It’s in your best interest to check out these details yourself.