The Xelera Suite software accelerates workloads in data centers, public clouds and edge clouds.
A software stack to leverage off-the-shelf accelerator hardware in tomorrow's data centers and cloud hostings.
Edge Computing service providers place the cloud resources nearer to end devices at the far edge in order to reduce the access latency and promote data locality. Here, acceleration hardware and software are mission-critical: They enable service providers to manage the increasing demand for latency-critical applications, such as live over-the-top video, gaming and real-time analytics/inference.
True real-time processing of large data volumes, for example in real-time decision processes and live predictive analyses in areas such as banking & insurance, consumer, health & pharmaceutical, IT services, logistics, manufacturing, retail & wholesale and telecom.
Bringing the performance of accelerators into commercial applications by ensuring efficient and safe plug-and-play use in the data center.
Enabling application providers to leverage the performance of accelerators and offer acceleration to cloud applications in an “as-a-service”-manner.
We have joined the O-RAN Alliance (https://www.o-ran.org/) as a contributing member. We have started working on hardware acceleration support for open and virtualized Radio Access Networks infrastructure.
Lukas supports our software engineering team with machine learning expertise.
Today Xelera announces a new feature release for Xelera Suite: The SAP HANA integration. It enables users to run analytics algorithms on the data in a very simple way, similar to SAP’s own Predictive Analytics Library but utilizing the power of hardware accelerators. We conducted a first benchmark showing the advantages. If you would like to know more about it or have a look at the complete results, check out our website:
Meet us in booth 232 at Supercomputing 2019 in Denver on Nov 17-22.
Based on the Video & Speech Streaming Analytics plugin, Xelera provides a real-time audio analytics application, which separates the audio recorded during a discussion of multiple persons into individual audio streams according to the active speaker.
Frankfurt - Today at the ISC High Performance Computing 2019, Xelera Technologies announced the launch of Xelera Suite for Edge Computing, enabling companies such as telecommunications service providers to use Xelera’s flagship software acceleration platform – Xelera Suite – to manage the fast growing demand in low-latency compute acceleration in Edge Cloud infrastructures. For more information read the full press release here.
Joining Xilinx at their booth at the Embedded World 2019 in Nürnberg Xelera presented a first live demo of their new Deep Learning Accelerator for FPGAs.
Xelera has been ranked among the top 50 startups in Germany.
Meet us at Supercomputing'18 in Dallas, booth 4346, on November 12-15 to find out about our data analytics/machine learning acceleration solutions.
Xelera presents Apache Spark acceleration at the Spark an AI Summit in London on October 3-4. Visit us at booth #64.
Just after the presentation of the newly released Alveo cards Xelera showcases up to 50x acceleration of machine learning training using the FPGA platform:
The Xelera Suite Middleware powers LYNX AIR, a cloud-based video processing solution by LYNX Technik AG.
We are proud to announce that Xelera has received the prestigious award for digital innovations in Germany 2018 by the German government. The jury acknowledged that business process acceleration in true real time will be mission critical for almost all companies.
We are proud to announce that Xelera is now part of TechBoost: The Deutsche Telekom start-up program. TechBoost fast-tracks selected technology start-ups with € 100,000 credit for the Open Telekom Cloud as well as discounts for Telekom products and tariffs. Furthermore, in case of extensive cloud usage, TechBoost start-ups benefit from valuable sales and marketing support – Boosting our business!
Xelera demonstrates up to 80x speed-up of the training of a random forest classifier (part of the Spark Machine Learning Library) at the International Supercomputing Conference in Frankfurt. The solution runs in the AWS F1 cloud and on-premise.