Cloud Astronomy

Cloud Astronomy is developing cloud-based solutions for astronomical image processing. Image processing can be demanding in its use of system resources. Many amateur astronomers experience hours waiting for results on their home PC because they lack the resources to allow the timely execution of complex processing algorithms. The cloud can provide massive resources to amateur astronomers as an alternative to purchasing their own hardware.

Here are the current system recommendations of the popular image processing software PixInsight:

Processor

  • Minimum required processor: Intel Core i5 or equivalent. Current 1.8.7 and 1.8.8 versions of PixInsight require a CPU with at least SSE4.2 instruction support on Linux and macOS.
  • Minimum reasonable processor: An AMD Ryzen or Ryzen Threadripper CPU, or an Intel Core i9/i10/i11 or Xeon, with a minimum of 8 processor cores.
  • Recommended processors: AMD Ryzen Threadripper 3960X/3970X/3990X, AMD Ryzen 9 5900X/5950X.

RAM

  • Minimum required amount of RAM: 8 GiB on a 64-bit machine and operating system.
  • Minimum reasonable amount of RAM: 32 or 64 GiB, depending on the amount of data to be processed, on a 64-bit machine and operating system.
  • Minimum recommended amount of RAM: From 64 GiB to 1 TiB, depending on the sizes and number of images to process, on a 64-bit machine and operating system. The recommended practical minimum for production work is 64 GiB of RAM.

Hard Disk Performance

PixInsight demands fast hard disks. We recommend working exclusively with solid-state disks (SSD) on all platforms and preferably with high-performance NVMe drives.

… Starting from version 1.4 (released in early 2007), PixInsight uses parallel disk I/O operations to generate and maintain temporary swap disk files. When two or more physical disk drives are available, PixInsight can be configured to spread swap files on all disks (no specific limit), and read/write them through parallel execution threads running concurrently.

… The performance gain that can be achieved thanks to parallel disk I/O can be spectacular. This allows working with huge images in PixInsight.