We’re thrilled to be joining SC17 this month, the international conference for High Performance Computing, networking, storage and analysis, where we’ll be showcasing how we’re accelerating innovation in Big Data and Fast Data.
Big Data and Fast Data – Building Better Workflows
The rapid advances in our ability to collect information from sensors, satellites, medical instruments, high definition video devices and computer simulators are increasing the size of data repositories to capacities never seen before. Analyzing this massive data is no simple task. It is both a performance challenge and a management challenge that requires monitoring complex storage systems and workflows.
The pace of innovation is often a direct result of how quickly and how well your data are collected, analyzed, and transformed. At SC17 we’ll be showing how you can simplify your HPC data workflow, increase storage optimization, enable better collaboration and preservation across global locations as well as reduce overall costs of ownership. Here’s what you’ll be able to see at our booth #643.
Big Data Insights with Object Storage
One major revolution that’s happening in HPC is the adoption of object storage for Big Data workloads. Object storage is purpose-built for Big Data, and is designed to deliver a cost-efficient and highly scalable solution for petabyte scale environments. Advanced object storage solutions such as our ActiveScale™ system can deliver several game-changing benefits for HPC workflows:
- Consistent access to large amounts of data – you can finally move away from tape and have ALL of your historic data accessible and online.
- Cost efficiency and optimization – move infrequently accessed data from expensive tier-1 storage to object storage. You can do this by using automated rules rather than a complex HSM layer.
- Simplify your architecture – aggregate disparate storage into a single global namespace. Move away from the limitations of traditional architectures and convoluted file hierarchies and adopt simple network topologies.
- Easy integration – with support for standard file and block protocols, you can easily integrate object storage into your existing environment.
- Cloud native – easily share data internally or with other scientists and organizations using cloud-native, S3 protocol.
- Extreme data durability and integrity – ActiveScale delivers up to 17 nines of durability, site-level fault tolerance, and automatically runs data integrity checks to ensure your data will be there when you need it.
Fast Data Actions with Flash and Platforms
HPC applications handle large amounts of data that require CPU-intensive processing and high data access rates, often accessing multiple servers. Whether climate modeling, protein and materials simulation, aerodynamic and thermal modeling or machine learning, fast data means getting more insights for faster action.
At SC17 we’ll be showing our latest in flash-based drives and platforms – from new innovations in NVMe™ PCIe SSDs to solutions that can provide cost effective performance. Using the right storage is critical to project success – here’s how you can get better results:
- The power of NVMe acceleration – whether it’s acceleration of GPGPU codes for visualization, rapidly storing critical simulation data, or minimizing IO latency of single-threaded workloads, the ultra-low latency of our new HGST Ultrastar® SN260 PCIe/NVMe SSD helps you maximize performance at incredible speed.
- Scale out with more efficiency – Scale-out storage deployments require a cost-effective solution to make Big Data faster and help Fast Data get bigger. The new Skyhawk Ultra PCIe/NVMe SSD is a great choice that can also remove bottlenecks of local scratch space.
- Up the stack with performance-optimized storage servers – With integrated CPUs and NVMe SSDs, the new HGST SVR2U24 Storage Server provides a performance-optimized platform for your Lustre® file system deployment. This platform was built for speed so your can take advantage of Fast Data with ultimate flexibility of software-defined storage.
Big Data, Fast Data, From Edge to Cloud
Big Data and Fast Data are the foundations for HPC workflows. Whether leveraging high-density storage devices, such as our helium-sealed drives, fast data SSDs, preconfigured storage servers with a mix of flash and hard drives, or ActiveScale, our booth will be a great place to learn which solutions best fits your data workflow model and needs.
Join us at SC17 in Booth #643
Come see us at SC17 in booth #643 to speak to our experts, see our demos and join theater sessions. Our four demos will allow you to learn more about:
- NVMe Acceleration: See how you can accelerate data analytics with NVMe drives
- Why Object Storage: Learn how we’re enabling Big Data insights with petabyte-scale storage
- Partner Ecosystem: meet a powerful ecosystem centered around you
- From Edge-to-Cloud: see the breadth of our portfolio
Our in-booth theater will host regular speaking sessions by several of our ecosystem partners, including:
- Globus –unified access to research data, across all systems
- ArcaStream Ngenea – seamless data management between object storage and IBM Spectrum Scale® parallel file system
- iRODS – open source data management software for simplified discovery, automation and secure collaboration
- Versity Storage Management – Exabyte scale data archiving management
You Can Trust Your Data to Us
Whether your data resides in the cloud or on premises, it is likely you are already taking advantage of our industry-leading storage devices. Just last month we shared some of the innovation that will lead the capacity hard drive market for the next few decades. Our expertise in manufacturing every level of the stack – from the hard disk drive components to NAND flash, storage servers and storage systems – delivers you with industry leading reliability, significant cost efficiencies and a data forever architecture. You can rest assured your Big Data insights and Fast Data results will be accessible when you need them.
Linda is Director of Machine Learning and Analytics Solutions at Western Digital and holds a B.S. in Computer Science.