I have a little problem. I have two computers plotting
The first is a Ryzen 3600 with 1 NVME rocket sabrent 1TB pci3.0 Disk. This monitoring PC manages to copy at maxium 1400mb/s making the plots. (monitoring in HWinfo)
The second computer is a dual xeon e5 2620 v3 (with Chinese motherboard), which only gets to copy at 600mb/s with a Samsung 980 (not EVO) 1Tb pci3.0. In the CrystalDiskMark test it writes at 2663.48Mb/s. This computer takes to make 3 parallel plots almost twice as long as the Ryzen3600 having twice as many threads.
Would someone know how to guide me where the problem is? Could it be because of the speed of the RAM? on the Ryzen3600 it is 3200mhz while on the DualXeon it is limited to 1800mhz.
(sorry for the english
clock speed of cpu and ram will slow you down too.
Thanks for answering
The ryzen 3600 = 3.8 Ghz
e5 2620 v3 OC = 3.2 Ghz
I understand that it affects, but how much? the e5 2620 has twice the number of cores
The exact amount I cant give you cores wont increase speed, only ability to do more at once.
I think it is not a problem. Write a large amount of data refers to sustained sequential write speed but not maximum write speed. Benchmark with default settings tests write speed with small blocks.
Look at this:
writing 10MiB samples
and writing 100MiB samples
The current implementation (1.1.6) is almost completely single thread CPU bound. more slower threads will help you very little compared to a single fast thread.
okay! Thank you all for answering, more or less that is what I knew but I doubted since in the ryzen I have reached a maximum of 1400mb/s in Hwinfo, while in the dual xeon at most it has reached 660mb/s, I wanted to find out if it could be due to some type of driver error or similar.