Jump to content


Community Developer
  • Content Count

  • Joined

  • Last visited

  • Days Won


jbartlett last won the day on August 6

jbartlett had the most liked content!

Community Reputation

155 Very Good


About jbartlett

  • Rank
    John Bartlett
  • Birthday 07/20/1970


  • Gender
  • URL
  • Location
    Seattle, WA
  • Personal Text
    Foster parent for cats & kittens

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Probably could use a change to the default label to add something like "non-UEFI only" to prevent against confusion.
  2. Sorry, I commented about the wrong motherboard. Deleted post. The Designare board is good but the PCIe passthrough proved to be problematic for me without the PCIe override enabled. I haven't tried in the past year though. It's been running rock steady though with a 1950X otherwise. The board I'm working with now and in love with is the ASUS ROG Zenith Extreme Alpha X399.
  3. I'll have to ponder this. My use case is video overlay which Livestream Studio seems to do in memory vs the GPU and exporting video over NDI. Once I'm done building my Ubuntu VM on the same box which takes that NDI feed, possibly an average CPU utilization with it spread out or on the same node will be telling.
  4. I'd take your unraid stick and plug it in your regular PC to test it for errors.
  5. <numatune> <memory mode='interleave' nodeset='0,2'/> </numatune> You're right, spreading the memory over both nodes did increase the average latency from 92.7 to 125.6 (26% increase). However, the read/write/copy scores improved under this "RAID" like effect by 27.9% / 45.5% / 31.2% respectively. Seems like a good trade off.
  6. With Hyper-threading detected in the VM, Cinebench R20 gave a 1.89% improvement in average scores (1/16/2) but the scores were a lot tighter. I was able to get the variance down to 50 with the high/low differing at 23 & 27 after dropping the three scores that varied farthest from the average (which were on the high side) with only three runs.
  7. I think I'm going to take your stance there. I don't need to squeeze out every MHz of processing possible.
  8. Holy CPUs Batman! This got me hyper-threading in my VM! Of course, it had to go through a Blue Screen first and then setting up my devices but yuh! Another set of Benchmarks to compare apples to apples coming up....
  9. Is there another section? I added it to my XML, set my cpuset, and on saving I get: unsupported configuration: Cannot find 'iothread' : 1
  10. Looks like I'll be installing Topo before I open my mouth again. haha Unless you already have the PNG....
  11. I had my CPU's isolated (CPU Pinning in the Settings makes it easy) but I didn't know about nohz_full & rcu_nocbs. Just added those. Good point on the task scheduler wanting CPU 0 because that's pinned to the main VM. I used to pin the emulator to the whole core but I've never seen the Dashtop show any usage on the 2nd one.
  12. The 2990WX has four nodes, ram is attached to notes 0 & 2. I pinned all the CPU's on nodes 0 & 2 to the VM so I didn't have to worry about getting memory from other nodes. Node 0 Node 1 Node 2 Node 3 Total ------ ------ ------ ------ ----- Huge 0 0 0 0 0 Heap 0 0 0 0 0 Stack 0 0 0 0 0 Private 7171 0 5199 0 12370 ------- ------ ------ ------ ------ ----- Total 7171 0 5199 0 12370 I can't get CPU-Z to work at all if I don't use the stock template, gets to the CPU detection part at 10% and just hangs. 3DMark, in the past, was questionable as it would hang at the collecting system info after a test or two. I'll try it again. I'll download AIDA64 and retry 3DMark. I don't know how my CPU will handle your 1950x's tweaks since the 1950 only has one numa node. It's promising because my main NAS us running the 1950 and I have a Win10 VM there that is my main driver for email & plex.
  13. <cpu mode='host-passthrough' check='none'> <topology sockets='1' cores='4' threads='2'/> <feature policy='require' name='topoext'/> </cpu> Topoext is automatically enabled in guests if the host supports it. Adding it to an AMD host that doesn't support it doesn't force it on. The CoreInfo utility still showed no hyper-threading.
  14. I found a domain feature which may enable hyperthreading in AMD CPU's that don't support it. <feature policy='require' name='topoext'/> I will experiment with it. Utilizing the following gave a 19% improvement in RAM scores with negligible differences between 1/32/1 & 1/62/2. VM had 12 GB of RAM assigned and this evenly spread it between the two numa nodes that had the memory attached to it - VM was likewise pinned to those nodes so the VM matched the physical. <numatune> <memory mode='interleave' nodeset='0,2'/> </numatune> <cpu mode='host-passthrough' check='none'> <topology sockets='1' cores='32' threads='1'/> <numa> <cell id='0' cpus='0-15' memory='6291456' unit='KiB'/> <cell id='1' cpus='16-31' memory='6291456' unit='KiB'/> </numa> </cpu>
  15. Already seeing a 19% improvement in the memory score with this set. I'll also test hugepages.