-n is just the total number of plots it will do 1by1.
If you want to run parallel, you need to run each in separate console/terminal
1 Like
How are you using 30 or even 24 threads on a 10 core / 20 thread CPU?
I was just trying out settings. Right now I’m using -r 20 and the plot time is exactly the same as when using 24 or 30. Meanwhile I changed to 128GB RAM 2666MHz mounted as tmpfs for temp2 and the plot time reduced to 29.6 minutes (30.5 min when copying the previous plot in background). Also tried with both 2 NVMe RAID-0 and a single NVMe and no difference.
In any case CPU is at 80% at most so most likely my bottleneck now is not the CPU or the disk (because RAID-0 makes no difference, I’m assuming the bottleneck may be the RAM speed or the motherboard bus).
If you only have 20 threads then you could put that number at 24, 30, or 3000 and it won’t make a difference. Just trying to say it’s not worth your time testing anything over 20 because your cpu only has 20 threads.
I was not spending time testing, just changing configurations everytime I started plotting to a new disk.
Regarding the ‘r’ parameter and the number of cores/threads, I don’t know if that’s so linear, because according to the documentation, that number is more like a multiplier, thus maybe even a lower number could lead to a similar result, although I haven’t tried.
Quick Q about xmp if someone in the know could answer I’d appreciate it.
If my mobo / cpu only supports 2133,.and I bought 2133, is there any point in enabling xmp?
Or is it just for ram that is a higher speed, to take advantage of that higher speed.
TIA.
Ah, nevermind, thought I’d just try it.
46 min down to 36 min, hoorah.
Hello, guys. I have i9-10850K 10C 20TH, 32GB DDR4, Micron 7300 Max U2 1.92TB and Samsung 970 Evo 2TB. Using Mad Max plotter i have 60 minutes. Bad result i think.I use only 1 disk.
My log:
Final Directory: p:
Number of Plots: 1
Crafting plot 1 out of 1
Process ID: 1168
Number of Threads: 16
Number of Buckets P1: 2^8 (256)
Number of Buckets P3+P4: 2^8 (256)
Working Directory: p:
Working Directory 2: p:
Plot Name: plot-k32-2021-07-06-21-32-2c43d0b860461348620608210a4d2a5e767a23573419b66aca3898356c04badd
[P1] Table 1 took 27.0144 sec
[P1] Table 2 took 186.483 sec, found 4294865369 matches
[P1] Table 3 took 297.075 sec, found 4294739167 matches
[P1] Table 4 took 352.932 sec, found 4294475469 matches
[P1] Table 5 took 355.905 sec, found 4293914248 matches
[P1] Table 6 took 327.575 sec, found 4292928118 matches
[P1] Table 7 took 227.662 sec, found 4290665257 matches
Phase 1 took 1774.87 sec
[P2] max_table_size = 4294967296
[P2] Table 7 scan took 16.1005 sec
[P2] Table 7 rewrite took 42.8584 sec, dropped 0 entries (0 %)
[P2] Table 6 scan took 76.7238 sec
[P2] Table 6 rewrite took 116.553 sec, dropped 581601759 entries (13.5479 %)
[P2] Table 5 scan took 61.5318 sec
[P2] Table 5 rewrite took 95.3363 sec, dropped 762291915 entries (17.7528 %)
[P2] Table 4 scan took 72.3943 sec
[P2] Table 4 rewrite took 112.862 sec, dropped 829088088 entries (19.3059 %)
[P2] Table 3 scan took 67.99 sec
[P2] Table 3 rewrite took 105.758 sec, dropped 855234158 entries (19.9135 %)
[P2] Table 2 scan took 86.6581 sec
[P2] Table 2 rewrite took 120.774 sec, dropped 865695938 entries (20.1565 %)
Phase 2 took 981.019 sec
Wrote plot header with 268 bytes
[P3-1] Table 2 took 78.5047 sec, wrote 3429169431 right entries
[P3-2] Table 2 took 65.5422 sec, wrote 3429169431 left entries, 3429169431 final
[P3-1] Table 3 took 105.49 sec, wrote 3439505009 right entries
[P3-2] Table 3 took 66.7394 sec, wrote 3439505009 left entries, 3439505009 final
[P3-1] Table 4 took 80.5635 sec, wrote 3465387381 right entries
[P3-2] Table 4 took 97.0116 sec, wrote 3465387381 left entries, 3465387381 final
[P3-1] Table 5 took 93.1583 sec, wrote 3531622333 right entries
[P3-2] Table 5 took 67.0325 sec, wrote 3531622333 left entries, 3531622333 final
[P3-1] Table 6 took 89.58 sec, wrote 3711326359 right entries
[P3-2] Table 6 took 69.2851 sec, wrote 3711326359 left entries, 3711326359 final
[P3-1] Table 7 took 88.8933 sec, wrote 4290665257 right entries
[P3-2] Table 7 took 95.0749 sec, wrote 4290665257 left entries, 4290665257 final
Phase 3 took 998.767 sec, wrote 21867675770 entries to final plot
[P4] Starting to write C1 and C3 tables
[P4] Finished writing C1 and C3 tables
[P4] Writing C2 table
[P4] Finished writing C2 table
Phase 4 took 91.4316 sec, final plot size is 108777914987 bytes
Total plot creation time was 3846.15 sec (64.1025 min)
Whats wrong? Thank you!
Linux or Windows? Which SSD are you using of the two mentioned? Have you tried different bucket sizes?
Windows. In log i use only 1 ssd Micron with write speed 1900MB/s. I haven’t tried diff bucket.
The SSD obviously the bottleneck. Try RAID0
Or probably try the 970 Evo as single temp drive first…
Thanks. I have 2 Micron 7300 Max, will try in RAID0. But i want to try ramdrive in Windows. What will be better: ramdrive or raid 0?
P.S. I think what i can’t create raid0 using 2 micron in pci-e ports. Right?
Ramdrive in windows is bad, but you don’t wear out your flash.
I see you have only 32 GB ram, which isn’t sufficient for ram as tmp.
For sure you can!
1 Like
Sure you can, just use windows storage pools and you should be fine
1 Like
Made RAID0, but its no difference. I think that pci-e x4 ports limit speed. Single disk - 1900MB/S, RAID0 - 1900MB/S
Sorry for asking what I’m sure is in this thread, but it’s got long
With windows and mad max.
- Is it worth playing with bucket numbers?
- Does changing buckets affect nvme writes, and potentially shorten their life?
Thanks.
I’ve got 11900k, 2 x nvme, 64gb ram.
Currently at 35 min plots.
Next try should be switch to ubuntu linux.
When Chia comes out saying now new plots will be final, what would be the syntax for MM? Something like this?:
-n 1 -r 16 -u 512 -v 128 -t T:\Temp\ -d R:\Final\ -p *** -c ***
Edit, seems more like this?:
-n 1 -r 16 -u 512 -v 128 -t T:\Temp\ -d R:\Final\ -c *** -f ***
Did it got resolved and can you able to run it
Is the same code after edit is that works with you ?