For those who know our history, Plunify was born in the cloud. Our very first prototype of InTime was developed, executed and tested on Amazon Web Services. During those days, we only had to contend with EC2, EBS and S3. We even had to implement AES encryption in S3 ourselves (nowadays it is a checkbox!). So it is with a sense of déjà vu that we found ourselves coming full circle as a vendor partner for the AWS F1 instances.
Enter a new age of FPGAs in the cloud
Fast forward to 2017, AWS announced their EC2 F1 instances preview in Q4 2016 and have recently announced the public release of these instances. An F1 instance is a fascinating combination of both FPGA hardware and the cloud. From the website, "Amazon EC2 F1 is a compute instance with field programmable gate arrays (FPGAs) that you can program to create custom hardware accelerations for your application. F1 instances are easy to program and come with everything you need to develop, simulate, debug, and compile your hardware acceleration code, including an FPGA Developer AMI and Hardware Developer Kit (HDK). "
The new F1 instances enable ambitious individuals and startups who are prototyping or creating new products to quickly and cheaply test on hardware before scaling their implementations. Big data analytics solution providers are also a great fit with their FPGA-enabled solutions for optimization applications and deep learning algorithms.
"Double dose" of performance with InTime
The key difference between Plunify back in the day and now, is that FPGA giant Xilinx has made Vivado (their flagship FPGA tool's) licenses available in the FPGA Developer AMI for any registered AWS user. With ample compute power and license availability in the cloud, we can now deploy InTime in the cloud and make it available to every interested FPGA designer browsing the AWS marketplace. Any AWS user can now fire up a Plunify AMI, run InTime and close timing on an FPGA design. All the necessary licenses and software configured and ready to roll.
For InTime, this is a perfect storm. InTime is a machine learning software that studies FPGA compilations. With virtually unlimited compute resources and FPGA licenses, you can drastically cut down your time-to-market when tackling timing closure or optimizing design performance. No matter what your FPGA application is -- Image processing, Controls, Networking, etc. , InTime takes in a XPR project file or DCP (checkpoint) and gives you a bitstream with better performance at the end. (Read more about InTime)
Scale up your capabilities (at a fraction of the cost)
FPGA design managers should be asking themselves - what is the best way to take advantage of this product that will cost a fraction of the amount compared to owning the compute infrastructure themselves. The answer is -- there are many ways to do it! The obvious one is you can look into simply running more compilations in the cloud to alleviate your resource crunch. Another way to increase your productivity by using InTime to rapidly optimize your design. If you can close timing in days, why spend weeks or months doing it?
Limited Time Offer - Sign up now!
As part of the general release of F1 instances, we are providing a limited time offer to evaluate and run InTime on the cloud free. Just visit this page to find out how to utilize this free offer. This AMI is still in beta and we welcome any feedback on how to improve it. Here is a guide on how to start an instance with InTime and run it step by step. If you have any thoughts, let us know.