I've always found the image processing stage a bit of a chore and convinced myself it was because the computer couldn't keep up with my amazing PI skills.
I started considering options, inspired by Rick's Threadripper build. There was a temporary setback when I managed to repair my Dell laptop power plug and restore it to it's full i7 8 thread glory however the spark had ignited and there was not going to be any stopping it now.
While a Threadripper was never going to fit within my budget, AMD continue to develop multi thread options and they seem to be seriously competing with Intel on price. After what seems like months of research I convinced myself I'd earned a new PC since the last desktop I've owned was in 1999 however I've built a number of gaming PC's over the past 6 years for my son and his friends so I felt quite well qualified to build an image processing beast of my own.
Roll forward to last weekend and it is done.
Picked up the parts from Computer Alliance - they are sooooo well organised and efficient at packing your order and have got Covid-19 covered - kudos to them
Assembled all the pieces in the box, connected the cables and powered it up and it came to life with a beautiful beating heart of the Ryzen Wraith cooler pulsating through the window of the case.
Couldn't get PI installed fast enough and ran a side-by-side benchmark with the Dell laptop for the benefit of the manager of household expenditure (and me)......it wasn't a fair match. The new build was done with the benchmark test in 10 seconds and was already shaved, showered and dried before the laptop was done. The project received a glowing report in the stage 5 look-back, so I'm calling that a good investment. Anyway on to the specs:
Ryzen 9 3900X
X570 motherboard
32GB DDR4 3600MHz RAM
860 Evo 500GB system SSD
970 Evo plus M.2 PCIe swap disc with room for a second one to be added later
RX570 GPU - not relevant to PI processing but you need to plug the monitor into something and Troy is trying to get me started on CS Source
Anyway, here's the current state of play. More work to do to understand any bottlenecks and further optimization options.
