Intel will be hosting a live Webinar with GigaSpaces on September 20 at 11.00 am EST (8:00 am PST) to discuss Bridging the Memory-Storage Gap to Accelerate Fast Data Analytics, giving you a chance to learn about evolving trends and our technology and products. Click below to attend:
About this webinar
The convergence of analytical and transactional workloads, through a unified fast storage and memory infrastructure, is giving way to a new breed of fast data architectures that support real-time analytics and just-in-time data science. To enable this vision, organizations are capitalizing on recent innovations in fast storage that combine volume, velocity, and variety facets of big data all in one data tier. The combination of in-memory data grids and storage-class memory solutions (such as Intel Optane SSD/NVMe) delivers performance that removes the need for competing operational and analytical strategies for accessing data.
In-memory data grids that leverage multi-tiered storage with SCM empower businesses with the capability of unlocking high-value insights from real-time and historical data. In this webinar, we discuss the forces driving this insight-centric trend as well as reference architectures using Apache Spark and GigaSpaces XAP to address particular design challenges for streaming analytics, fast data lakes, and continuous machine learning data pipelines.
- Webcast Live Date & Time: September 20 at 11.00 am EST (8:00 am PST)
- Duration: 60 mins
- Presenter: Ali Hodroj, VP Products & Strategy at GigaSpaces
The recording will be made available on demand.
GigaSpaces benchmark with Intel: Closing the memory-storage gap to accelerate data processing and analytics
Just last month we announced our collaboration with Intel to integrate and benchmark its IMC solutions with Intel® Optane™ SSDs and NVMe SSDs. The integration creates a distributed data fabric that can store data across RAM and SSDs transparently, accelerating access to data and maximizing processing, for extreme transactional processing and fast data analytics; supporting hybrid transactional/analytics processing (HTAP) initiatives.
Results of the benchmarking tests showed dramatic improvements across read/write workloads, as much as 150-300% across various tests.* The low latency, high endurance combination with high throughput will fuel the growth of new fast data and cloud based applications for customers.
Key benefits of the integration between GigaSpaces and Intel include:
- Enablement of enterprise-grade fast data applications– The resulting data fabric delivers mission critical real-time response and supports a heterogeneous data model (Objects, JSON, Spatial), full SQL support, and end to end Spark analytics API.
- Intelligent storage tiering – Enterprise systems are no longer limited to the choice between memory for real-time data or storage for historical records. They can now leverage a distributed, multi-tiered, elastic data grid which delivers high performance while intelligently tiering data processing and storage between RAM and SSD.
- Lowering TCO – Enterprises can store, process and analyze data-heavy applications in real time with less RAM footprint, eliminating cluster sprawl for Terabyte-scale footprints.
“For years, IT architects and decision-makers faced a performance dilemma between memory and storage, but innovative technologies, such as Intel® Optane™ technology-based solutions, disrupts the trade-off when incorporating in-memory computing,” said Ali Hodroj, VP of Products and Strategy at GigaSpaces. “The combination of Intel® Optane™ SSDs with our In-memory computing solutions, XAP and InsightEdge, opens up a new world of exciting opportunities in hybrid transactional/analytics processing. Data engineers, application developers, and scientists can operate on the same data set, right when it’s born, and at massive scale.”
* For Intel® SSD DC P3700 Series for PCIe (NVMe) benchmark results, visit the Intel® Storage Builders Program website. Intel® Optane™ SSD P4800X results will be soon be published. Results are available via direct requests.