When attempting to compile architectures in the Intel® FPGA Deep Learning Acceleration Suite using the Intel® SDK for OpenCL™, some of the architectures may require up to 256 GB of working memory.
For the shortest compile times, add additional RAM to the development system.
If additional RAM is not possible, create additional working memory on the hard drive.
In Linux, increase the size of the swap partition or add a swap file.
In Windows, increase the paging file size.