Missing Link Electronics

MLE Supports Amazon EC2 F1 With Binarized Neural Network Inference

Users interested to witness the acceleration potential of FPGAs in the cloud for Machine Learning applications can now use an exemplary implementation of a reduced-precision Binarized Neural Network inference for CIFAR10 image classification, running on an Amazon EC2 F1 Instance.

Presentation at SNIA Storage Developers Conference 2017

MLE and Xilinx present "Heterogeneous Multi-Processing for Software-Defined Multi-Tiered Storage Architectures" at the SNIA Storage Developers Conference 2017.

 (download slides) or visit our technical publications.

New FPGA Solutions from MLE Listed by Xilinx

In recognition of MLE's strong committment to support the FPGA Acceleration ecosystem, new solutions for Network Protocol Acceleration, Key-Value Store (KVS) & Memcached Acceleration and Storage Protocol Acceleration have been added to the Xilinx Alliance Program.

1st Workshop: Programmable Processing for the Future and Autonomous Car

Xilinx and Missing Link Electronics are proud to support Professor Anestis Terzis from the Ulm University of Applied Sciences in organizing and performing the 1^st Workshop in Ulm: Programmable Processing for the Future and Autonomous Car - From classical FPGA to heterogeneous MPSoC on the 11th of July all day in Ulm (University Campus Albert-Einstein-Allee)

Servicing Amazon EC2 F1 Customers

Amazon Web Services (AWS) recently introduced the Amazon EC2 F1 instance, a compute instance with FPGA for custom hardware acceleration. MLE's mission is to de-risk the adoption of FPGA technology in new markets and applications by offering expert Design Services.

To fully support customers in their Amazon EC2 F1 projects MLE has been enabled by Xilinx and Amazon with sufficient seats for the required toolchain, and documentation to fully support users interested in FPGA acceleration in the cloud! Click here for more information.

MLE Releases New Zynq SSE

With full support for Xilinx Vivado Version 2016.4 and PetaLinux Version 2016.4, including a new Linux kernel 4.6 device API, this new Zynq SSE offering also comes with a more customer-focused license fee structure. This reflects MLE's dedication to the growing needs of embedded applications using Xilinx Zynq-7000 All Programmable SoC with SATA connected Solid-State Drives. Click here for more information.

Presentation at FPGA4GPC 2017

MLE presents first OS metrics for DTO-based DPR based on exemplary implementations of programmable accelerators for the Linux Kernel Crypto-API, implemented on the Xilinx Zynq 7000 platform at the FPGA4GPC (International Conference on FPGA Reconfiguration for General-Purpose Computing 2017) on 2017-05-09 & 2017-05-10. These metrics can guide future research on OSs forGPC on FPGAs. FPGA4GPC focusses on novel work in the field of reconfigurable systems optimized for the use in a general-purpose environment.

Robert Milton Gardner, January 31 1943 - April 11 2017

On April 11, MLE's President & COO passed away  after serious illness. As a co-founder, role model and mentor at MLE Bob took a significant part in shaping the team with his leadership style, his deep interest in semiconductor technology and strong personality. We will always keep his memory in honor.

reVISION_Alliance Program Member

MLE joins the Xilinx reVision ecosystem. As a Design Service Partner to the Xilinx ecosystem, we have been actively supporting key innovators for computer vision in the automotive and industrial space in the US and in Europe.

Binarized Neural Network Application

MLE presents an Embedded Vision Robotics System with Accelerated Image Recognition at the Embedded World 2017. This industrial pick and place demonstration utilizes Binarized Neural Networks for machine vision recognition via programmable logic. This FPGA-based implementation accelerates image recognition while increasing recognition accuracy and reliability necessary for high performance, mission critical vision-enabled applications.