Best time for a single plot

Hi Guys,

This is my first post - Just wanted to thank you all for your valuable info.
Secondly, I’m interested to see what speeds you guys are getting on a single plot to give myself a comparison to see if Ive done a good job or not.

Im running an Intel Core i5-10600K @ 4.10GHz with
Adata SX8200 Pro series 1Tb drive
32GB Memory
Offloading to a Crucial BX500 480GB SSD

Started a plot at: 16:11
2021-05-18T16:11:42.397 chia.plotting.create_plots
Ended at:20:30
Time for phase 4 = 496.344 seconds. CPU (93.830%) Tue May 18 20:30:26 2021
4H:19M

Cheers

This is my best so far

Intel Core I7 - 6700,
Windows 10 Pro
Temp: Samsung Evo 870 (sata),
Final: Synology NAS

Just playing around with stuff I have idle

best

1 Like

i got much the same with a similar set up but to an hdd with my first plot!

i added another 1.6tb u.2 ssd (on a pcie card) for plotting and do 8 at a time now, time per plot is about 6-7 hours. slows things down but gets more plots a day.

i have other plotters with -d drive as an ssd and it’s super quick to finanlise.

The best I have personally seen is 4.5 hours, that’s with i9-9900ks on a 970 pro target.

Since Linux is 10% faster, and there are faster CPUs than the i9-9900ks the very fastest systems should be able to approach 4 hours for a single plot.

Plotting is mostly CPU limited in my experience, you can search for ramdisk topics here to see why I think that :wink:

But not all CPU’s seem to scale equal for Chia

from: Tested CPUs for Plotting Chia, Intel vs Ryzen. Intel Wins? Ryzen 7 2700 Ryzen 9 3900x Intel I9 9900k - YouTube

2 Likes

Hmm that is an interesting graph

It seems to indicate that Intel has an advantage in plotting? But those CPUs are fairly low end aren’t they? 2700 and 3900x? Isn’t 5950x the fastest? I have not been keeping up with AMD stuff…

The Zen 3 series (5XXXx) is definitely a great improvement. As for fastest it depends. The Ryzen 5900x and I believe even the 5800x beat the 5950x in some workloads like gaming.

I’ve never done 1x at a time but with 4x I’m right around 4.5 hours average. Threadripper 3970x, 980 Pro Temp (4) and 18GB External WD (4) for Destination. Only running 4x the system is still asleep… I can load the system about 50% before I run out of SSD, which puts me just over 80/day. Trying to tweak everything to get more though. So hard to tune these as there’re so many variables and so many hours between tests:

My HDD’s are full so i’m filling my ssd’s with plots. These went really fast. Don’t know if it is because there were only 2 running or because of my new 570 motherboard with more gen 4 support.

5800x with oc / ram on 3777 / 980 pro 1 tb / 970 pro 1 tb / gigabyte aorus gen 4 1 tb.

Thats pretty quick! I thought my 4H19 was fast but youre ahead!

It looks like the clock speed plays a big role, what is yours OC to?

4.5 Ghz on all cores. Tomorow a new 18 TB disk will arrive and I will be able to write more plots simultaneously. I’m curious what my speeds will be when I will have more running parallel. On my B550 motherboard I was getting 6,5 / 7 hours per plot with 8 or 9 running at the same time.

3.7hrs was fastest single I did with 5950x iirc. Wasn’t even with a fast gen4!

storage_jm’s recent video for chiadecentral is worth a watch on the this topic,

1 Like

MacBook Air M1 8c/16t – NVMe OWC Aura P12 via Thunderbolt 3 Envoy Express (plots 4t/4GB)

  • Total time = 13894.995 seconds. CPU (140.040%) Fri May 21 21:31:14 2021
  • Total time = 13547.966 seconds. CPU (144.550%) Fri May 21 19:31:49 2021

Averaging between 3.7-3.8 hours total per plot.

1 Like

Just watched this - 2.6H!

I think its worth overclocking the CPU. Did not realize the difference.
Going to give it a try in between next plots on my 10600K.