19 June, 2017
Cloud computing is becoming ever more prevalent in the engineering world. Using the Cloud can provide businesses with access to high performance computing on a cost per core hour basis, without the need to invest in hardware that will most likely be considered dated within a few years of purchase. Cloud computing can help reduce time-to-market by removing time taken waiting in queues for use of hardware from the equation.
With the release of 6SigmaET R10, users could benefit from the advantages of cloud computing for the first time thanks to our partners Rescale. Users could export jobs from 6SigmaET and upload them to the Rescale platform to be solved before downloading and importing results. This unlocked a whole new world of fast solving to those who were previously limited by the hardware available within their organisation. But now, we’ve made it even better.
In Release 11 it is now possible to submit/retrieve jobs to and from Rescale directly from the 6SigmaET user interface, as well as monitor job progress. This means a user no longer has to endure the process of exporting jobs and importing results. All Rescale settings, such as selecting core type, are incorporated in the new interface. The user interface was designed to be as familiar as possible to existing users and is based on Batch.
Initial set up of this new feature is simple once you have an account on the Rescale platform. You simply need to copy and paste the API key from your user account into the set up wizard in 6SigmaET. When you first click ‘Cloud Solve’ the software will provide detailed instructions to guide you through this process.
There are also a whole host of other improvements including: the introduction of new core types and faster start up times. The introduction of new core types can, at one end of the scale, provide more performance and at the other end cater for more budgets; per core hour costs can be as low as $0.08. No matter which core type you use, optimisations to our solver have dramatically reduced start up times. In testing we have seen start up times more than 3x faster.
By: Matt Evans, Product Engineer