Case Studies

CPU, GPU and FPGA

We are computer technology experts who specialize in producing solutions for environments where throughput and latency are critical. Our previous work has seen us deliver solutions specifically optimized for server farms, GPUs and recently FPGAs.

Our latest work is principally focused on automatically generating efficient implementations of deep learning algorithms on FPGAs for use in server farms and autonomous vehicles: drones, automobiles, and robots. The goal of this work is to drastically reduce energy consumption and increase the range of fields where these algorithms can have a real world impact

Our expertise is applicable across a wide range of industries and below are selected highlights of our work.

FPGA

July 25, 2016

Myrtle's state of the art DRUM compiler enables quick deployment of low-latency, high-throughput FPGAs solutions

Custom GPU Compiler and DSL

June 30, 2015

Myrtle authored a custom llvm-based compiler, with CPU and GPU back ends, that provided a full data driven post-processing pipeline for real time graphics contexts

Visual Effects Work

January 31, 2013

For over five years Myrtle worked with leading visual effects house DDPI developing and supporting the next generation of its Academy Award Winning, CPU based, simulation and rendering software

Transformers: Dark of the Moon

November 28, 2011

Myrtle provided new software and full support for all the volumetric elements on all the Transformers: Dark of the Moon shots executed by DDPI

SPEAK TO THE EXPERTS

Contact us