performance-icon   PERFORMANCE

It’s time for real time

The pivotal battle scene between the hero and the villain needs to be finalized today. It’s looking good but it needs to look great. The character rigs are complex, dragging the authoring application to a halt and your animators just can’t iterate fast enough. Fabric Engine can take your rigs from painfully slow to genuinely real time. This means more iterations and more reliable artistic choices, leading to a better final result.


Powerful Core Execution Engine

multithreaded

Multithreaded. GPU Accelerated

Fabric minimizes the complexity and cost of building high performance custom tools. Faster tools. Better results.

cross-platform

Cross-Platform

With Fabric, you only need to write your code once, saving the cost, time and pain of targeting each platform individually.

JIT

Dynamically Compiled

The Fabric core execution engine leverages LLVM, JIT-compiling your code and allowing you to rapidly iterate.

“We’re seeing many ways Fabric could be used to speed up workflows and to update older tools to be more efficient. Fun times ahead.”


Eric Thivierge

Rigging lead – Hybride

“Being able to visualize the work that you’re producing as you’re producing it is really important. KL was already as fast as C++ on the CPU and now it’s as fast as CUDA on the GPU. That’s really, really exciting for us and why we’re working with Fabric – it’s the future of computing.”


Damien Fagnou

Global head of visual effects – MPC

Multithreaded. GPU Accelerated

Multi-core CPUs are common. Programmers skilled in writing code that takes full advantage of them are not. Rarer still are those who can leverage the power of GPUs. At the core of Fabric is an engine that does both.

Build Better Tools Faster

Fabric Canvas is a visual node and connection interface that lets artists and programmers quickly build new tools with immediate feedback. Not only is creating and modifying a tool fast, each node in Fabric Canvas is expressed in KL “under the hood” so the resulting tool runs at speeds comparable with compiled C++.

Let Fabric Do The Heavy Lifting

With Fabric Engine, you stay focused on tool creation and not on low level optimizations. The Fabric Core provides easy-to-use, high-level primitives for both task- and data-based parallelism. Instrumentation methods start and stop collection of data to help you profile time spent in user code (KL operators), the Fabric Core or external to Fabric.

Use ALL the Power

Now you can build new tools to run on either the GPU or CPU to get the highest performance possible. You don’t need to master CUDA or OpenCL because Fabric KL code runs without modification on CUDA GPUs. Your tools are automatically optimized for maximum performance on the chosen hardware.

Cross-Platform

Writing code optimized for multi-core CPUs that also harnesses the power of the GPU is hard enough. Ensuring that code runs efficiently across Windows, Linux, Mac OS X and a plethora of graphics cards is harder still. Fabric Engine handles this with aplomb – performance-sensitive applications can now be written once to target multiple hardware and OS configurations while still getting the best performance out of each platform, saving the cost, time and pain of needing to target each platform individually.

Dynamically Compiled

Whether you’re using Fabric Canvas to visually connect nodes to create a tool that simplifies and streamlines your workflow, or whether you’re coding in KL (or both – feel free to mix and match), you’ll get immediate response to your changes thanks to just-in-time compilation. Under the covers, Fabric Engines leverages LLVM, a collection of modular and reusable compiler and tool-chain technologies that takes care of a number of difficult problems, freeing you to quickly iterate and test your tools with the assurance that they’ll run wherever you need them.

FE_logo_losange-75

Get Started with Fabric

Looking for a better, faster, cheaper way to build high performance tools?

Get Fabric