Issue: Volume 38 Issue 2: (Mar/Apr 2015)


Autodesk’s Memento Reaches Beta

The ability to digitize reality into high-definition 3D just took another step forward as Autodesk Memento, previously available as an Auto­desk Labs technology preview, became available as a free public beta.

Memento is an end-to-end solution for converting any reality capture input (such as photos or scans) into 3D in the form of high-quality models. This 3D mesh can be cleaned up, fixed, compared over time, and optimized for further digital use, for physical fabrication methods like 3D printing and for 3D interactive experiences on the Web – from the all-in-one application.

Memento can handle more than two billion polygon meshes, yet no CAD expertise is required, making it easy to use for digital artists as well as scientists.

The beta release includes a new website with links to learning material and user case studies, and soon a 3D Online Gallery where 3D digitized models can be viewed and shared interactively. Beta users will experience Memento first as a cloud-connected desktop app that will initially be available on the Windows platform, followed by a Mac version later in the quarter.

Nvidia Supports Deep Learning

At the Nvidia GPU Technology Conference (GTC), CEO and Co-founder Jen-Hsun Huang introduced three new technologies that will fuel deep learning. One was the GeForce GTX Titan X, an advanced GPU that features more than 3,000 CUDA cores, has 12gb of onboard memory, and costs $999. The Titan X is built on the Nvidia Maxwell GPU architecture. The single-precision floating-point performance will reach 7tflop. In double-precision floating-point performance, the card can reach 200gflop, allowing it to process very complex data quickly.

Nvidia also revealed the DIGITS Deep Learning GPU Training System, a software application that makes it easier for data scientists and researchers to create high-quality deep neural networks quickly. The company also revealed the DIGITS DevBox, introduced as the fastest deskside deep learning appliance. Powered by four Titan X GPUs, the box is built to handle deep learning research and costs $15,000.

Nvidia Updates GPUs, Iray

With updates to the Nvidia Quadro professional GPUs and Iray rendering technology, now for the first time, interactive, scalable, physically-based rendering will be a viable and practical tool for deadline-driven projects.

Beginning this month, Iray 2015 will be available via new plug-in products for the most commonly used design software. Quadro Editions of each will run free of charge for anyone with a qualifying Quadro GPU (the new Quadro K1200 and up). When paired with Nvidia’s newly announced Quadro M6000 GPU, Iray 2015 dramatically reduces the barrier to entry for interactive physically-based rendering.

Iray 2015 is scalable via its Iray Server distributed rendering software, so that a local pool of users can harness additional power from each other’s machines when not in use. The Iray 2015 can be scaled by using the Nvidia Quadro Visual Computing Appliance (VCA), which has been upgraded with eight M6000-class GPUs. The Quadro M6000 features 12gb of memory and, like the GTX Titan X, also uses the GM220 Maxwell GPU.