#off-topic-tech
1 messages Β· Page 51 of 1
well shit i cant even get port fwd to work with ddwrt now π
Did you manually install the right drivers?
was the deveice
device malfucntioning
im at step 7/8
installed lineage
Yup, that specific error means there was communication error when Windows asked the connected USB device what it is and what it can do.
"Device Descriptor Request Failed" part
Can be because of bad connection, bad cable or just the device firmware sending crap for some reason.
The device was being a bitch. (Fixed that.)
Yes, saw that it was solved, but as general information for others in future.
And I successfully created installer config for Office 2021 LTSC that doesn't install Teams, Onedrive or Skype \o/
(I hadn't bothered to try last time I had to install it)
Well, Spotify does it too in latest version... DNS, SSPD, mDNS.
So seems to be Chromium.
Including acting as responder part.
8GB of RAM absolute minimum, 16GB strongly preferred.
Just that as long as you have 8GB of RAM and 4GB of VRAM, you can run the game with almost anything from last 10 years.
But that 8GB of RAM will limit save size very soon.
btw does any1 know how i can solve it? even after i installed the newest driver this still happens. o,o
And did you verify the game files like I told you to?
Or check the other stuff I mentioned about general system stability?
If the game file verification didn't help.
well yea i verified it and they were all there
but this happens with many games on steam
Then you almost certainly have unstable computer.
idk until like this week i never had the problem
OC can get unstable with time.
And even stock can get unstable with time, if there is hidden fault in some component that then develops into real fault with time.
Curve Optimizer is OC, setting RAM to the marketing spec is almost always OC, and so on.
so what should i do to prevent it from happening again? ;w;
Need to find what is causing the instability and fix it?
yea
idk the instability since theres almost 0 pattern, it just decides to crash midgame sometimes and then other times decides to sleep for 10h+ straight
And how is that relevant?
You don't test with games if you can avoid it.
But you have not told anything about your computer except it has problems with some games and has some model of non-XT AMD GPU.
I mentioned two common causes for later instability (CPU OC or undervolting, RAM at marketing spec and not auto)
Another is just RAM going bad for any speed.
Depends on what your system is set to on what is next step.
Test without OC, test RAM, something else.
thank you for answering : D
Either is damaged right away, has hidden fault that surfaces usually in under 6 months, almost always in less than 12 months.
the one im looking for only has 8gb of ram but expandable to 32gb
it worked fine for 4y so far o,o
And the old age dying depends a lot on voltage etc. stuff is run on.
Higher the voltage, earlier the death on average, for same thing.
And DDR4 default is 1.2V for example, most kits sold for gaming computers run at least on 1.35V if not higher.
And that also causes extra stress on the CPU silicon.
so should i just get new ram in that case, cuz i think the ram is the reason since i never overclocked my cpu, and the ram tends to go to 70% without doing anything oftenly
But again, no specs, no info what you have set in BIOS etc.
No, you should test the RAM?
And RAM capacity usage doesn't matter.
Again, is the RAM set to XMP profile or like?
To make it work at the marketing spec?
If yes, try without that first, with the RAM on all-auto.
id have to check 1s
If that doesn't help, then run Windows Memory Diagnostics (worse but builtin) or Memtest86 (better but needs to be run from USB stick).
Its set to XMP2, idk what u mean with the other tho
Meaning RAM makers programmed XMP type profile in slot 2.
Which is what that RAM can do on golden MB & golden CPU system when new.
Not that it can do it on YOUR system.
So disable that and allow the RAM run at actual JEDEC stock SPD settings and see if it is stable then.
Or run memory testing with XMP on first, and then if that shows errors, without it.
(both XMP profiles that can be in the RAM)
so I should test XMP2,1 and without? but I don't got no USB Stick so I should use Windows thing yes?
XMP is way to set the RAM to the marketing spec (tested on that golden part system) easily.
XMP1 and XMP2 are such auto-OC specs programmed to the RAM sticks.
Set that thing to disabled/auto to let the RAM run at JEDEC stock.
And see if the games still crash or not.
it would crash occasionally so I'd have to wait like 5h for a chance for it to crash.
But RN I'm testing with it on first like u said earlier
btw with the windows one should I use standard setting or advanced
Standard for first at least?
wdym for first
That depending on what that first run says to decide what to do next.
Ok
The normal seems to run mid coverate testing once with cache on.
And the advanced side allows to run three different test coverates (basic, standard, extended), and specify how many times the tests are run before it returns (taking more time), and if the CPU caches are enabled or disabled for specific tests or not.
With XMP2 it's currently at 52 with no errors, so it's taking its whule
With XMP2 it hasn't found any problems. Should I still try without?
?
May I ask full system specs @queen pendant
And next crash what's the error.
As those bsods often mean something
@twin dew Couldn't I just use this for the pump? https://youtu.be/tCJ2Q-CT6Q8?si=LZFl-g60s0057hDX
This video explains using MOSFET as a Switch or as a PWM driver to control motor.
Here is better version for Higher voltage or Higher current https://youtu.be/eqXaqRFAWrA
The best thank you for me is to not skipping ads and Subscribing π to my channel, and clinking π the video. It is greatly appreciated. π π.
I will reply to all Subscriber...
Says input signal 3.3V
Looks like easy PWM speed control, no?
Within the specs of that, yes.
So 36V, 15A, 400W max for the motor.
And that rating might be with extra cooling.
Would this be considered a good network architecture?
Or just buy 3.3V to 5V converter board to put between the Pi output and the Motor PWM controller you bought, so the 555 can be completely bypassed.
Designing a potential network for my home
Weird that I had such a hard time finding it then
It's very cheap and small
20pcs/10β¬
I understand that I can't reverse with it(don't need it as it's for a pump), but theoretically can make a H-bridge with 4 of those for bi-directional current.
I don't understand why it's so hard to come by when searching for ways to control a motor with raspberry pi.
@edgy hazel flashed lineage easy enough. No more issues with my pixel
I wonder what else I can do now with all this
I Think my battery life has gotten hella good too
44% will last me 18hrs.
VS like 6 on stock android
depends what is "flex mini" and how they are connected with each other in "edge-router X"
by default you can try to group as much stuff into single switch as possible
then connect that with each other
then if you need network isolation or physical isolation, split it
then adapt for performance/bandwidth
noting that having vlan aware switch can help isolating things even if you keep the single physical switch
The u6+ is an awesome AP
Just a "dumb" switch from ubiquiti. I need them to have one in every room since there are not enough ethernet jacks on the wall
Connecting everything to a single switch is going to be problematic because of house wiring
depending on your cabling, you could eventually connect them with each others, depending on how you plan to use "edgerouter"
I mean if its better to not do it then its better to not do it, you just have to start with that idea (which you did at least now)
Edgerouter is going to be a central point where a fiber line is connected. From there, I will pull cables to 2 different rooms where I'll put flex mini in order to get more ethernet ports for devices
but what kind of device, I mean how you will route your room
if you want a single network you can put a switch here as well
if edgerouter is a switch then that's it
if its simply a router you may have problems with transparently communicating between other switch
if it has plain sub/integrated switch then its ok
Flex mini is just going to be an unmanaged switch in order to get more ports for devices
the most "powerful" cable must be between first switch and edgerouter
there will be load between the wifi endpoint and first switch if you want to be able to acess nas from other device
as well pc will probably access camera
and also wifi devices
and even tv,
tv may connect to pc
so the frirst switch <=> edgerouter
must be the most robust one
I currently have a somewhat similar setup where a cable from router leads to a room with unmanaged switch where I have my pc and NAS. So far, it works without any issues
Don't have two switches thought atm
I could connect tv and nintendo switch to AP but I like wired connection 
can be intersting for switch to be cable yep
Can't wait to buy U6+ and test it compared to all-in-one wireless routers I've been using so far
I just meant that for the whole network, if you have a better cable
put it between :
- first switch <=> edgerouter
2-3. nas <=> first switch
2-3. pc <=> first switch - wifi <=> edge router
Oh, you mean cat6?
one big upgrade of having dedicated ap
is to have real single network (not have wifi logically loading the router chip to reroute/proxy the wifi to the integrated switch)
multi gigabit ports /cat
Won't be a problem. Will just need to buy 2 cat6 cables and that's it. Cables connecting different rooms are already cat6
And everything connected to the first switch is using cat6 too
everything 1Gbe ?
Only the network equipment atm is 1 gbit/s. All cables are cat6 afaik
?
cat != speed
with a short bad cable
and 2 beefy ports you can still have 10Gbe
but long bad cable with 2 shit ports over 200m, and it's even if you could pray for it to even connect
Oh, the longest cable is 15 meters
that's why I was asking port speed independently of cable quality
that's why I said both
put better cable on better speed ports
and put better speed ports on first switch first
so I guess you meant everything is 1Gbe max
Yeah
Since network equipment is the bottleneck in my case
But I think 1 gbit/s should be enough atm considering current setup works so far. The only problem I have atm is bad wifi coverage
I don't transfer files between my pc and NAS often so I don't mind not having 10 gbit/s
1Gbe is good but for :
- core networking if you will have mixed heavyloads often going on (not your case)
- between your pc and nas if your job is working on big files
- between core and nas, if big enough network, and the whole network consume data from nas (and nas is not bottleneck)
ok
then you're fine with 1Gbe
1Gbe is very fine with everything usual consumer side
it can deliver typical HDD data bandwidth
Yeah, if I was transferring files to/from NAS often and they were large, I would definitely go with 10 gbit/s
Maybe things will change in 5 years and I might move to 10 gbit/s. It's hard to tell what the future brings 
for instance I have core network core pc and core server sharing virtual storage volumes as well as multiple kind of network loads, and file sharing, and it have to be able to loopback efficiently
so I have :
pc <=10=> core <=2x5=> server
cost of 10g is going down, but its still complicated
fibre is going down too, but its still even more complicated
(for everysingle task 2.5g could maybe fit, but its all at the same time, so its very not enough)
especially big file and storage volume sharing
My biggest worry is changing serial number and ploam password of gpon sfp so I can avoid using ont and plug the fiber directly into edgerouter
But I've seen guides on the internet for the isp I plan to use so I guess it should work
Need to make sure sfp module is compatible with edgerouter
make sure the edgerouter is :
- flexible enough to tweak it to your needs (potentially even scripting it)
- performant enough so it's worth trouble of having direct fiber connect
some isp likes to have fancy stuff such as unusual dhcp keys
you must be able to configure it
2nd point is optional obviously if your main concern is control and privacy
Just want to get rid of ont since I don't want a combo of ont+router
Planning on buying this (people have confirmed it works perfectly for the ISP and edgerouter)
It's basically ZyXEL PMG3000-D20B and it's compatible with edgerouter
Just need to access the web interface where I can change the serial and ploam. People are saying to avoid buying huawei sfp since it's a nightmare to configure
Zyxel is the easiest one to configure in my case
huawei is also not nation security friendly
make sure you have proper firewalls in place if you're gonna have your NVR connected to internet
I'd use MAC whitelisting for it if it supports that
aaaaa im an idiot. ended up checking if port was being used via netstat and it wasnt. reinstall of server resolved the problem
first pc i built i managed to forget to plug in the cpu power, & it took an embarrassingly long time to figure out what the beeping & GIANT FLASHING RED LED right by the connector was trying to tell me
heh
did that once cleaning a radiator hehe
i knew someone who forgot to take the plastic off the heatsink base ...
wouldent matter until you launched a game but yea thats probably the most painfull fail you can have
system suddenly jumps to 100C
been there done that
π
not the plastic but the 100C
heck, my 14 series does that all on its own no matter what cooler is on it lol
rad or no rad intel cpu's dont care π₯
lol right ... "im too hot, i'll just throttle"
loss of thermal controls is always a fun way to test when some plastics become a bit more plastic
But the code part would be exact same for the other solutions too.
Don't worry. You can only remotely access it through OpenVPN π
I could also add mac whitelist on nvr
Time to start disassembling the old main computer...
Here I am cable-managing the secondary case, when I was only supposed to do that once the new main machine was in operation...
And I'm blind...
This MB had both native TOSLINK output and S/PDIF header...
And I went and installed TOSLINK output addon to that header...
Savun hΓ€lvettyΓ€ tarkistamme kytkennΓ€t...
Booted after clearing CMOS.
So the MB doesn't work if I disable CSM...
As in cannot boot to UEFI setup.
Setting Intel IGPs (V)RAM allocation to 1024MB doesn't work... Boot fails on PCIe resource allocation.
And OS installed, time to start tuning...
maybe this channel should be renamed to "general-unsatisfactory" lol
but why.
looks like a horse
i hate u
what the bloody
i typed and pressed enter on that message in the general-off-topic channel... how the fuck did it end up in here ?
the magic of electron apps
something applied physics to it and it fell down into this channel instead π
prolly just a fucked cache
... reloading discord just in case
still here ... so yeah just got sent to the wrong place
i blame macOS
i mean, Adobe does for all their bugs in their program
(even though many of the bugs still exist on the windows version LOL)
bro fuck adobe
but ohh nooo... its 'apples macos is causing the problem'
sure let me just wait 10 seconds to look at a pdf
until this fucked up PDF reader start
hahahaha you use adobe reader ??? i use Preview.app
i do a lot in pdfs on my mac ... but yeah, i guess on windows its a bit reliant on the shit adobe software
hat to use it for a signature and it kept bothering me everytime I want to open something. It's gone by now
or how about Adobe Air for an application to run π
and don't get me started on the sell-up they keep trying to push
everytime I click somewhere wrong HEY YOU WANNA BUY ADOBE PDF VIEWER PRO???????????
JUST 200β¬/DAY TO LOOK AT A PDF π
lol
it's really time for the entshittification
i had t ouninstall creative cloud cause it kept putting shit in my sidebar
and not like something useful ... but a folder with a name so long its like wtf "Creative Cloud Documents And Fonts folder"
any app that puts something in the sidebar is in violation
"oh they wont mind...." ... then when the user takes it out, a week later "oh, it seems to be missing from the sidebar, lets add it... they wont mind..."
it's just very aggressive "hey come into our ecosystem!!"
i also uninstalled it cause it kept updating a ton of js and css files in its cache or something, which would then show up in my active folder of "Recent Code Changes" ... and since macos smart folders are not smart anymore, i can't EXCLUDE a directory to auto-search in
macos has smart folders
"smart" lol
my life has changed
what the fuck I've been using macOS for like 7-8 years now and never knew this
and if you accidently edit or touch the search criteria, it lights up this button:
however there is no 'undo' ... and if you close the window and try to re-open your smart saved search ... its all fucked up still
seriously ... i hate apples auto-save-on-anything bs they have been implementing
ive ruined so many documents by accident because of that, and just closing the window, saves all the messed up shit
winget is pretty nice.
winget is great
yes, I don't see how someone could regret any ----X3D CPU, such good performance love my 5800X3D
Im not sure if the 7900x3d is as good tho
Thats true, and the 5600X3D didnt seem great but it was only in bundles which were decently priced so still no complaints
It's a 7800X3D
So go for it cause thats the best price/performance :D
anyway, need to start replacing parts so I can keep my PC from crashing.
and if I am going to replace my CPU to start with, it might as well be an upgrade π€·ββοΈ
thi channel too
Rip
Was responding to previous statement
sudo enters the chat
anyone know how to fix this
i think its because i had the mod launcher before update 8 and havnt played sines but i deleted it and reinstalled i have no idea what to do
you clearly dident uninstall it because its still trying to launch whit mods that arent suported
uninstalling the mod manager doesn't uninstall mods. please come to the modding discord for modding support
ty i fixed it
still, if you have any more issues please ask for help in the modding disc because this disc is oddly hostile to modding sometimes
outside of this channel
Well, in turn I'm only hostile to Satisfactory Modding talk on this channel and the other offtopic π
@edgy hazel olives again
ew
But why has Fractal Design failed so badly with Define R6 front panel connectors...
The cable can just get to the MB header on full-size ATX MB when routed via the "clean" route.
And even then it should be few cm longer.
So just tiny bit too short for that red route they are using even in the marketing pic, even with full-size ATX.
So with ATX that only uses two first screw locations it is too short, same for any kind of mATX.
And need to do the orange route instead.
Fortunately I could do the red route with little corner cutting near the MB with full-ATX, but still...
agreed
Hopefully finished for now.
Need to check if I get smoke or not after eating.
At least booted to UEFI Setup...
And just realized I connected the monitor to the IGP and not the dGPU...
Quick to fix, but had to turn the computer off for that.
But the case reset button has died.
The connector at the button PCB end had started to pull the traces off the PCB.
Checked that it still worked before assembling, but has died after the check, probably while reassembling the covers for it.
So no Direct To UEFI button for me π
But anyways, off to old test Windows.
But how is my cable management π₯²
Need to convert the HDDs sector format from 512e to 4Kn.
And read little more and not doing that, as there doesn't seem to be any benefit if the stack is 4k aware, like everything is today.
π
yea just build aligned partitioning and with 4k blocks
that's what the "e" mean in 512e
512e is 4k physical with 512 conversion layer
I know, logical 512 sectors emulated in firmware with 4k physical sectors.
Which is what would even have made that conversion possible.
To tell the HDD to just not bother with the emulation.
But there was still small chance to brick the drive during the one minute or so it processes the request.
If the bricking chance wasn't there, I would have done it even without real benefit.
Far better than average
you do not wanna see my nzxt h1 cable management
And the IGP connection thing was because that was where the monitor lead went for the past several months.
And I just cannot get Win10 or 11 to install on AMD chipset NVMe RAID or RAIDABLE single drive...
The drivers load fine, the drive detection works, but writing just craps out.
In the installers, works fine in OS.
Weird...
And I'm not the only one with the problem.
Just no solutions anywhere.
So going to try the bypass.
Installing OS to SATA SSD, then cloning the partitions to the NVMe RAID.
Seems that the driver craps itself if the Windows management utility isn't running, and it cannot be running during the install.
On AM5 boards.
Possibly only if both SATA RAID and NVMe RAID are enabled?
Got much farther with SATA drives disconnected and SATA in AHCI mode, but still failed...
One last try with SATA controllers completely disabled.
And then on to cloning.
My cable management skills are average at best...
Need to remove dust soon
All the mess is hidden behind 
Which is why I took pic of that side too π
I'm not doing that lol
Well, not until I do some cable management
Hopefully getting nvme soon in order to replace old ssd's since there is not enough capacity (not to mention higher speed). I will also use those old ssd's in laptops
Since I am getting fiber soon and there is no fair-usage policy afaik, I could switch to cloud gaming for really demanding games
Still extra latency.
Need to test in on some non-demanding older games to see how it works out
Before moving to more demanding ones
Btw, I hope there are multiple servers on cloud gaming so I can at least lower latency
now thats my kind of thing just hide it until your willing to show it haha
At least it's much easier now since I have modular PSU
Seasonic β€οΈ
The store was selling both 650 and 750 W for the same price and I took the advantage of it and bought 750 W
Not sure if it's normal for stores to sell same PSU's with more W for the same price
might be an overstocked item they are trying to move
whats with all these reddit posts about something called "Copilot" ...
Have you not heard about microsoft's newest dumb program?
Its chatgpt but its in ms office and windows now
=]
ok, i've never crapped myself in my adult life, but this kind of thing might do it
united, San Francisco to Boston, 757 diverted
"delamination" of the leading slat
we love airplanes
"delamination" is term for when the layers in composite material separate from each other.
Which makes the material much weaker, and has caused it to break in this case.
As the slat is made of three layers, top and bottom outside layers and honeycomb-material between for strong but light solution.
Slat is the moveable surface at front edge of wing.
The shown position is for when the airplane is optimized for higher speeds.
And it can be moved downwards to change the aerodynamic shape of the wing to allow for lower stall speed, combined with the flaps at the back for lower speed at landing.
And takeoff.
just terrifying to see
it's like when you see all the spoilers come out when landing, looks like the wing is coming apart
Anyways, last try to get Win11 on AMD RaidXpert NVMe RAID-1, then going to try to do it with Windows Dynamic Disk Mirror.
How would one debug 'lag' in a server
How would you determine if it be network issue or computer
Really fascinating post
Well, seems I will have the EFI partition on the 58GB Optane, and the 980 Pro:s will be in Storage Spaces mirror with OS install there...
Win11 doesn't support Dynamic Disk Mirror booting anymore.
No, the EFI couldn't boot the second stage...
More testing...
If it's in an internal network, it's most likely not network. If it would be network, you'd notice.
But yeah a tracepath would most likely tell you if it's network
And the installer is trying to put the boot stuff on the Storage Spaces too...
And not on the separate normal drive that already had the partitions on it from another Windows install...
it still can be network settings
And back to having the System partition on Windows Mirror.
Just with some extra steps to have non-mirrored dual EFI partitions etc.
?
I spent like 5 minutes trying to understand that
To trust Storage Spaces "Mirror" or not, and to trust ReFS or not...
Any words of advice?
Ah, yeah, just remembered it doesn't give any alerts for anything...
And ReFS doesn't really help without Storage Spaces either...
Anyways, the two 2TB NVMe SSD:s are now in Dynamic Disk Mirror, with 0.5TB for OS and 1.3TB for game installs etc. with NTFS for both.
But need to decide what to do with the two 512GB SATA SSDs that will have the Windows Profile, and the two 6TB HDDs for Stuffβ’οΈ.
What to use for mirroring and if I want to use NTFS or ReFS for each...
File contents checksumming would be so nice to have
Seems it is finally starting to be stable?
As in Server 2019 was still somewhat messy but with Server 2022 ready for primetime?
Seems to finally work?
Even just 2 years ago it only detected bit rot etc. on "primary" disk, and didn't check the other if the primary copy was OK.
But now seems to check both?
And no more "Windows Update turns the drive to unformatted RAW" either?
So going to test if I can run SS+ReFS on both SATA sets.
Will see how many programs that install themselves under AppData fail because of that.
Need to do the profile partition as non-case-sensitive, but the HDD stack ReFS can be case sensitive.
I meant like peer connection settings
even if network is good it can be badly managed by app depending on its networking settings
that's what the pathtrace is for.... and no like 99% of the time a wrong config leads to nowhere/error (and a badly managed app isn't network)
So next I need to read again how to create Storage Spaces pool in PowerShell, as the GUI for that just creates pools with bad settings...
Wandows.
And even the pool creation needs to be redone with PowerShell...
Because it fucks up PhysicalSectorSize when SSD lies about that being 512 bytes.
Which will cause the write amplification to go through the roof.
My NVMe still seems to be falling off the bus sometimes. Yay..
But at least I now know it's only the one storage. Shame it's the one with the root filesystem..
Ok, so Storage Space Pool set up with the two HDDs, with 4KB physical and logical sector size set.
Then Mirrored Virtual Disk created on that 64KB Stripe size with fixed allocation.
And then ReFS volume created on that with 64KB Allocation Unit size and Integrity Streams turned on by default.
But seems I can do just 4KB Allocation Units for the SSD, as the two-disk Mirror type doesn't care so much about the Interleave to FS Allocation Unit size interplay as the higher disk amount things care.
Robocopying the whole old C-partition from single SATA SSD to the HDD Storage Space, hitting up to 300MB/s with 200MB/s or so as normal speed.
But as the source is inconsistent from the sync mode with various size files, that is completely normal.
Connecting to host 192.168.1.69, port 5201
[ 4] local 192.168.1.106 port 57344 connected to 192.168.1.69 port 5201
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.00 sec 113 MBytes 949 Mbits/sec
[ 4] 1.00-2.00 sec 112 MBytes 942 Mbits/sec
[ 4] 2.00-3.00 sec 111 MBytes 934 Mbits/sec
[ 4] 3.00-4.00 sec 108 MBytes 907 Mbits/sec
[ 4] 4.00-5.00 sec 106 MBytes 893 Mbits/sec
[ 4] 5.00-6.00 sec 112 MBytes 937 Mbits/sec
[ 4] 6.00-7.00 sec 111 MBytes 927 Mbits/sec
[ 4] 7.00-8.00 sec 106 MBytes 889 Mbits/sec
[ 4] 8.00-9.00 sec 107 MBytes 895 Mbits/sec
[ 4] 9.00-10.00 sec 109 MBytes 912 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.00 sec 1.07 GBytes 918 Mbits/sec sender
[ 4] 0.00-10.00 sec 1.07 GBytes 918 Mbits/sec receiver
iperf Done.```
testing internal doesnt show any problems π
satisfactory sometimes "hiccups"
but cpu usage never over 20%
32gb memory on that system so shouldnt be that either
Total CPU usage doesn't show anything meaningful with many-core CPUs in most loads.
And neither does the averaged per core readings, as the threads jump around too fast.
Need to look into per thread CPU usage with semi-specialist software to see if there is CPU bottleneck.
For SP or Host you can check the "stat unit" ingame console command output.
But as client or with dedicated server, no such easy way to see.
And this is wrong place to talk about Satisfactory anyways.
π€ lets say its palworld
Then I have no idea 
I'm absolutely clueless. Way too new a game
13 SECOND average response time at spikes, HDDs are "great".
Now to bed.
Tomorrow last backups from old drives, and then to finish the new systems Windows installation.
13 second average response time???? is it just a really slow drive or what?
with latest kernel ?
there has been multiple ssd/pcie/usb-c bus falloff fix in in 4-7 last kernel versions
especially when going in and out of sleeping
I'm using fairly new kernels, 6.6 right now. I've had the issue since somewhere around 6.2 or so.
But it's happening during active usage. I'm playing a game or writing code and suddenly boop, gone.
And entirely unrelated but, YouTube, what?
Is this A/B testing gone wrong somewhere?
ungoogled chromium ?
Firefox.
oh
It was happy until about an hour ago.
why don't they provide full version number
it's frustrating
x.x is just feature
not patch level
Not every distro or kernel distribution keeps with the mainline kernel versioning. They may pull from different commits and have it suit their packaging.
I use Xanmod-packaged kernels for instance, so I have a 6.6.10 kernel, but the .10 refers to the version packaged by Xanmod.
Ah nope, it does refer to mainline at least.
But then my distro will have packages like 6.5.0-18, where they sit on a single kernel and only add patches.
Linux π
Apparently the bug is related to Adblock. I have no idea if it's an incompatibility between a YouTube update and existing Adblock settings, an adblock change breaking things, or YouTube deliberately trying to degrade the experience for adblock users (which wouldn't make sense, given there's no indication of adblock being the issue)
So I guess I'll just write my own patch for now and wait for it to be fixed.
Linux good
I concur
had this happen aswell but not as extreme
in fact i have it right now
It's fine. When this stuff happens, it just disincentives me to use it. π
I've got a ton more stuff I can watch, play, etc.
Like Stargate SG-1.
yea myn is just a little skooted over nothing that crazy
At what resolution?
1440P
maby zoom because im at 150% zoom
It looks like some kind of fix is rolling out at least, so yay.
Really glad to see this, in my mind that's like charging an auto manufacturer because one of their vehicles was used in crime
https://arstechnica.com/tech-policy/2024/02/court-blocks-1-billion-copyright-ruling-that-punished-isp-for-its-users-piracy/
No, just being hammered with certain types of writes.
HDD:s have lot of seek latency.
IronWolf Pro NT 6TB 7200 RPM NAS HDDs.
This is the reason why HDDs are so bad compared to SSDs as OS drives.
As they just suck at small reads or writes being hammered on them.
Five star rating system comeback?
yes please
Moving 1.9TB from one HDD to another takes so long...
120MB/s?
Yeah, read is happening at about that speed.
Now that it is doing large files.
Should be less than hour remaining now.
should take like a day
About 4.6 hours if constant 120MB/s.
Ok, so little longer, as the robocopy has been running for 2-3.5 hours now.
Started with another partition copy at 6:50 or so, and might have started the current partition copy at 08:30 or so local, and now it is 10:46.
This is just YouTube lagging, either wait a bit or try to quit fullscreen and then go back.
Need to get the backup done, so I can install and nuke second SATA SSD, so I can put two of them into SS Mirror with ReFS, so I can set Windows to put new profiles on that partition, so I can create my actual user account.
Which needs to be done before I can start setting up the OS properly.
I really miss the two-slot-high taskbar...
But the regedit to get that in Win11 doesn't work anymore in 23H2.
regedit is really hit and miss
Microsoft doesnβt really seem to document it properly.
The spot is still there, you can edit it, but it reverts to default 01 automatically now.
Most I kno that does work or doesnβt is by trial and error
As in there was specific byte in HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Explorer\StuckRects3 that allowed higher taskbars, and second that allowed it to be moved to different edges, but neither works anymore in 23H2, when they worked in earlier Win11.
I donβt like to mess with regedit anymore. Just a clusterfuck to keep track off what works and doesnβt.
Best is to always backup and backup.
Must have been unintended and patched.
Or they unintentionally patched it π
It's not.
And just remembered to add the registry key to prevent automatic device encryption.
And now need to document it.
Documenting all the various settings changes I do on newly installed OS.
Settings, Control Panel, Regedit, Group Policy etc.
Finally done, took 4h 54 minutes.
And seems that the warnings that Win11 doesn't work right if the basic profile stuff isn't in C:\Users\ was true.
Even when the move was done by pointing Windows to having different default profile storage place for new profiles and then creating new one.
So people at MS have coded stuff to not use the %profile% type shortcuts for some of the "Modern" style system apps and instead have hardcoded to C:\Users\<profilename>\
5 hours for 2T across 2HDD?
From one HDD to another.
i guess thats not that bad cuz it takes about 1 on most m.2's
but yea HDD are for the waiting game
No, I just had written "F:\" instead of "F:" into the new profile location registry entry.
And that fucked things up.
As the final profile specific reference ended up being "F:\\Baldur" and not "F:\Baldur"
I did the same. Took me days. But I don't regret it. I had to do multiple clean OS installs and my little guide really helped me setting the device up the way I wanted. Quicker.
It's a bit of effort, but it can pay back.
Been doing it at the same time as I have been configuring the systems.
As I have just done three OS installs.
And when I needed to troubleshoot, I just went over the list to see what changes I made to system.
Need to do the final touches on this third one on the main computer.
And then port the new stuff to the other ones too.
So did I. Hence it took days. Needed to restart the installation multiple times because of mistakes etc.
It's only 277 lines of text π
Baldur, you know if I can use these two together?
Need extra part between for that exact combo.
That converts the 10mm stud on the red part into 4mm banana socket for the cable.
sjiiiit, can't I just pull out the clam from the wristband and plug it into the earth bonding point? π
There are also yellow ones with banana connectors.
this one?
This thing as adaptor is needed.
Still banana on the cable.
Would need this type of plug
The black ones are 4mm bananas and the top is 10mm stud.
wellt he plug wasn't meant for wristbands
meant to be used like this.
and it's the only one the shop has.
You can get wrist straps with either 7mm or 10mm stud on the strap.
And then the connector onwards on the cable can be 4mm banana (most common) or 10mm stud.
And you can get that kind of banana to stud converters.
in theorie, I could just clam it on the pin in the socket.
but that seems so unstable, I will yonk it off pretty easily.
I don't understand why all these wristbands ship with these clams? Is it really that good π
But either you need another kind of cable for the wriststrap, converter like I picced, or another kind of power socket plug thingie.
and in my local shop they laughed at me when I asked if they had it. They don't believe in wristbands.
got it.
Now if I just remembered what was the key combo in the "select user" login screen to switch to "enter username" version for one time...
So another shittiness...
Dynamic Disks don't support TRIM passthrough.
So need to keep eye on if AMD fixes the issues in their NVMe RAID driver and reinstall OS at that point.
And ReCaptcha showed image of a scooter moped as "Select squares with motorcycles"
And System Restore doesn't support ReFS?
yea.... just like accidentally making a 2.2M bitrate nuke on my HDD once lol
i accidentally recorded in no compression instead of minimal compression and because i was recording at 1920x1080 172fps it quickly became very fat very fast like 700GB per hour of write/read kinda fast
minimal compression for camparison is generally about 25GB per hour at 1920x1080 172fps
ewww system restore
gonna defedently remember to double check compression next time i try to record slowmotion footage
The "Previous Versions" functionality that is on top of that is nice.
A limitation which applies to System Restore in Windows versions prior to Windows Vista is that only certain file types and files in certain locations on the volume are monitored, therefore unwanted software installations and especially in-place software upgrades may be incompletely reverted by System Restore.[18] Consequently, there may be little or no practical beneficial impact.
To recover previous user file versions.
However, beginning with Windows Vista, System Restore monitors all system file types on all file paths on a given volume, so there is no issue of incomplete restoration.
In a professional setting, sure I guess.
When you just accidentally deleted wrong file with shift+del.
Or game fucked up with its saves, etc.
backups to nas
Once I get it running.
Already enabled File History once per day to the HDD stack, but still need to add AppData to the directories it handles.
the reason I am not a fan of these OS recovery systems is because when I needed them, they failed me.
and so far, my plain backups to my NAS have provided more flexibility and solutions than TimeMachine or windows recovery.
Was just so easy to just get the directory properties and Previous Versions in there.
As the first step to try.
Yeah, but it's also kinda of "all or nothing" since it reverts multiple files and settings.
No?
So you don't know what was wrong, you just know something was wrong somewhere and reverting fixed it.
That is the point.
The "System Restore" does full.
"Previous Versions" can browse that datastore and get single files or directories.
yeah, I am talking system restore
Just that having System Restore for the volume enabled is the prerequisite for the Previous Versions.
previous version is a cool feature, but most cloud solutions also offer that. Like my NAS. I can set it up to always keep X versions of a file if I want too.
As both use the same Volume Shadow Copies service data.
I see, well. Multiple ways, same solution I guess.
Just automatic snapshotting.
And two things running on top of that datastore.
One to restore the system to earlier state, one to just get random files from there.
I meant, that snapshots can be done by OS (like you seem to do) or by external software like Synology NAS.
Just pick your poison or whatever floats your boat π
I don't like to be dependent on the OS for recovery anymore.
win7 win10 win11
always landed partially reverted bullshit broken system in every single times I saw it attempted
Was just the first quick level.
And again, not talking about System Restore as the system to restore the OS to earlier state.
But the stuff running to keep data about the old file versions etc.
Yeah, but the fact you need SR enabled to able to revert file versions π
Just about this thing
Also, I rather have external storage taking care of it.
More room for games and screenshots π
Because it is about if the Volume Shadow Copies is active and how much space it is allowed to use at max.
Or one could add extra SSD just for that, they are not that expensive.
It uses less if the drive gets full.
if you really need to save, then use backup
that is my way of doing it then
and unless there is a "manual snapshot" that provides similar functionality it's no use for me
How? Does it clean up older data?
Yes?
It could be that it started to cherry pick files to reduce space.
Or clean up older data. But have drawbacks.
No, just removes oldest points.
Yeah, just like timemachine does.
And yes, time machine has file history.
and yes, time machine can also (or should be able to) restore full OS. π But this is the point that failed for me. It wouldn't do a install from Timemachine.
So it just a glorified file restoration π
Ok, seems that the Previous Versions also looks into the File History data, so as long as I enable that for the AppData too, it will be fine.
I am so tired of crashing when I alt tab π’ (hopefully fixed soon)
Making backups of my minecraft world whenever I can. Changed auto save to every 1 minut too.
and I wonder if I will ever know what the correlation is between CPU and alt + tab, I assumed this was more GPU bound.
is minecraft crashing or what?
nah still first issue, full PC crash.
It happens a lot when I am gamging and I alt + tab. Also happens other times. Lowering FLCK gave some stability, but not solve the issue.
what if youre doing like a gpu/cpu stresstest and alt tab, does it still crash?
that's the main issue, can't reproduce it when I want. But it keeps happening when I don't want it.
Doesn't happen during a normal CPU stresstest (10 minutes).
neither during a memory test.
ordered a new CPU, so in a month or 6 I will know if the new CPU fixed it 
what new cpu did you get π
7800x3d iirc
The way I see it, I see it as an upgrade
So if it doesn't work, atleast I got better CPU than 7600X
you're using factory settings in bios?
100%, besides on.
for some weird reason, factory resets FLCK back to 2000MHz.
Which is troublesome, 1733MHz is more stable. And according to Baldur, 1733 is default.
what form of crash? I had shutdowns after i put in a new 6800xt until i changd out the power cables
like smash to poweroff
Because "All CPU should be able to do that", and ASRock is doing insanity because of that.
So we are not sure why the motherboard is defaulting to a wrong FLCK setting.
HW induced reboot, no notification to OS.
So should be MB/UEFI seeing some uncorrectable error or other dangerous situation and rebooting the system.
And FCLK is the only one that should count.
And the specific CPU gets worse as FCLK gets higher.
Uhu uhu, we think the CPU has been on bad settings for a while.
While I didn't do anything in the bios π
So the leading theory is VSOC damage from too old BIOS originally.
That is now hitting critical stage.
And then we reverted everything to auto, it helped.
consider remove, clean contacts on everything, if you have 4 ram sticks, reduce to 2
The last springs thingie.
refitted my RAM already.
Symptoms match.
refitted my GPU.
And any of those shouldn't cause reboots without warning.
I didn't refit my CPU π
If you don't have ECC RAM.
I guess so? Honestly, I didn't know this was even possible π
I though all errors in this day and age would be captured by OS or BIOS or something.
It fucking sucks to be in the dark.
With ECC RAM clocked too high you get the exact same, random reboots without warning, when you get uncorrectable error.
As AM5 AGESA doesn't support FCLK/IF or ECC RAM error reporting yet.
I did RAM test. Was all fine. So in my case, I don't think it's the RAM.
You don't have ECC RAM.
So MB wouldn't know there was error.
So no reboots.
Yeah, doesn't mean it does't suck 
You would get BSOD or app crash instead.
π
Major oversight from AMD in my opinion.
Lot of stuff missing in the AGESA at this time compared to AM4.
Lot of CBS directory entries that don't have anything under them yet.
I don't understand why, besides money.
Releasing a half finished product to cash in on it and stay competitive.
IIRC it took few years to get full ECC RAM support for AM4.
At first it was in similar state, with just enable there, but no settings and no reporting.
Not enough skilled enough coders to do the AGESA type coding.
In so many business it wouldn't even be possible to produce a half assed product.
Most MB makers too only have few UEFI/BIOS coders.
As the amount of people who can do the job is very low in world, total.
And not lot of training for that going on.
And by the time they have AM5 finished/perfected. They release AM6. Idiotic.
And for lot of the stuff, the server CPU line gets priority.
I am just pissed. A simple error with: "You fried your CPU" would be fine. I can work with that.
It would safe me time. And I had a reason to send my CPU back.
I don't care if it doesn't work.
Dropbox doesn't like ReFS.
what do you mean by this?
They probably did it risk analyses. How many people will have trouble with FLCK? 0.1%? Oh, that's fine. We can lose 0.1% of the costumers 
"No need to focus on catching FLCK errors, just make a cheap CPU".
I don't know ... just unlucky I guess
or AMD is less quality
My school did something good for once they finally enabled 5G on the wifi network
no
my high school experience was the best experience I had.
Total freedom + it's your own responsibility to study or not. Loved it.
But seems I can finally play something tomorrow again.
After I move the new system to right spot under the desk and connect right monitors etc.
Nice π
Was hoping to get there already today, but didn't get there.
its on its 5GHZ band now but 5G whit 400mbps download/upload is probably not gonna work great
so far it seems still as unstable as before its just quicker so the speeds are the same but its more responsive it reacts quicker essentially
but thats not too suprising 400mbps isent a lot once it gets split across like 2500 devices
im fairly certain its also capped at 15mbps per device even if theres enough network to go higher
especially for modern standards...
I actually prefer work than studying. This might sound weird but school and college was more stressful for me than work lol
Oh absolutely, I think that's true for most people
And also love when you earn some money and you can decide on what to spend it for
Especially since money is usually really tight when you're in school
College was really stressful for me because my parents paid a lot for it and they would beat me up really bad if I failed the year
I guess that's what was keeping me in order
But perhaps the worst part is when you are a fresh graduate. Looking for my first job was a nightmare and I almost suffered from depression and thought I couldn't get a job because I was completely useless and dumb.
Once you get some work experience, things are much easier
Yeah, that's a very typical scenario to end up in
Working on making industry connections prior to graduation can help a lot
All the OC stuff is off right now.
But pretty unbalanced system currently.
I need RX 7900 XTX!
Madness follows:
Need to sort all that old stuff on G later.
Sorry not what I meant.
Of all the schools I did, my last experience (hogeschool) was the best. Basisonderwijs and *middelbaar * weren't that enjoyable. No freedom
So I am speaking of the higher education.
ASO/TSO/KSO/BSO - Primary school and pre-school were the worse.
Wish I was back in the last year of high school when I graduated and college started 4 months later. I had the longest vacation ever π
wasn't your whole adventure to get your system stable π€
im sure this is probably a good place to ask how tf do you make a windows batch file?
open notepad
write stuff
save as .bat
because it seems like it has changed from windows 10
dosent work just renames the text document
how?
brother select all files in the dropdown instead of txt and add a .bat behind it
otherwise just save it as txt and change it in your file explorer
that also dosent work
thats why im hella confused
well then i guess all 3 of my windows 11 copies are broken
that's not the thing being broken
A better editor doesnβt hurt
it did change slightly you cant replace them anymore you have to make a new one
but i got it
Notepad++ is what it sounds like
uh i got it whit just normal notepad but i do have notepad++
Download notepad++
i think this is the first time i have seen a air cooled gpu bellow room temp
the thermodynamics implied in that statement fascinate me

i suspect the thermometer location for the room's temp is giving an inaccurate measurement
they just need to enable show filename extension to do it though
its probably inacurate but its chilling in the 50C exhaust of a 13700K so not like it makes a difrence lol
my cpu sits at 38 usually, gpu 50
Going to start re-enabling the OC settings over time.
Just didn't bother after last BIOS update.
So going to set the Curve Optimizer and FCLK today at least.
Then start re-enabling RAM in a week, and do that over time.
As no stress testing can find everything and actual usage can cause issues.
And as the AGESA doesn't do those FCLK/IF and ECC RAM machine check error notifications currently...
But the unbalanced mostly meant the CPU, RAM, drives. Compared to the GPU.
2000+e other components with 230e GPU.
Where even the MB costs over 2.5 times what the GPU costs.
I really wanted RX 7900 XTX π
What's the avaliability like over there?
Problem is Money, not availability.
I see
Hopefully in next 6-12 months...
Currently tight on cash as well, so I get that π
Waiting on ebay refund for a faulty piece so I can try to replace it again
We love inflation
Das so many cores
Nvidia has really been on money grabbing spree:
https://investor.nvidia.com/news/press-release-details/2024/NVIDIA-Announces-Financial-Results-for-Fourth-Quarter-and-Fiscal-2024/
Q4 2022 to Q4 2023:
GAAP Revenue from $6.1B to $22.1B (+265%)
GAAP Net income from $1.4B to $12.3B (+769%)
Y2022 to Y2023:
GAAP Revenue from $27.0B to $60.9B (+126%)
GAAP Net income from $4.4B to $29.8B (+581%)
Record quarterly revenue of $22.1 billion, up 22% from Q3, up 265% from year ago Record quarterly Data Center revenue of $18.4 billion, up 27% from Q3, up 409% from year ago Record full-year revenue of $60.9 billion, up 126% SANTA CLARA, Calif., Feb. 21, 2024 (GLOBE NEWSWIRE) - NVIDIA (NASDAQ: NVDA) today reported revenue for the fourth quarter ...
For those who don't know:
Revenue is how much you sold/had total income (gross profit).
Net income is how much profit you have left after all expenses (net profit).
GAAP is specific US accounting standard so you cannot fudge the numbers so easily.
When everyone thinks they need AI in their businesses
Remind me why did you need that exact mobo?
Slot layout, 10G LAN, video outputs.
But I did buy the stuff last spring knowing I might get fucked with GPU at least in short term.
In most AM5 ATX MBs, you don't have space to install both big GPU, and two PCIe add-on cards (Sound card & 10G LAN).
Because almost everyone has made it so that 7th slot is PCIe, then M.2 slot, then another PCIe at 5th slot.
But that PCIe is too close to the 2nd slot with the 16x for GPU.
And as I wasn't sure if MST hub would work properly, I needed two DP outputs from the MB for the upper monitors.
Why sound card tho? Just get a good USB DAC
Which I would somehow needed to bolt to the bottom of the desk or like?
Internal sound card is still just so much easier for space usage and tidiness.
Apple's USB-C adapter is a good DAC and is just a cable extensions
And don't know any good one with good headphone HRTF.
The Creative external ones are crap.
I'm using Optical in, Optical out, mic in, headphone out.
Apple adapter accepts TRRS - tho optical is an issue with that
And it isn't good DAC, it is just cheap usable quality one.
Except the unit I got for another use was complete crap.
Might have been faulty, but didn't bother trying to RMA.
The amount of noise on the mic input...
As in very high noise floor.
-30 or -40 dB?
It's as good as you need unless you're after high end stuff
But why is the one M.2 slot between the two PCIe slots, and not other way around...
GPU, M.2, M.2, PCIe, PCIe.
In the cases when the middle PCIe isn't connected to CPU and is only 4x from chipset.
Or just 1x in this case, with 16x physical slot.
Already had 100+e PCIe sound card before started planning the new build.
Price β quality, but yeah a pre-owned card that you want to reuse changes things
oh, also: EU apple adapter is shit
"Cheap end", so Creative Sound Blaster Z SE.
Mainly Creative for the HRTF.
Even when the one in that is worse than X-Fi:s had π₯Ή
Everyone else has even worse or not at all.
Anyways, now starting to play Control again for second time.
This time with patches and DLC.
Before heading to Alan Wake 2.
Friends got me hooked on Helldivers 2
Too bad it crashes constantly
I have completed 2 missions and the tutorial so far - 3 more missions were ended prematurely through a game crash
tbf tho, one of those three I would've failed anyways - but the point is, I have a 40% mission completion rate due to instability
Which is shitty.
It is Creative Sound Blaster Audigy RX.
But it uses chip from before X-Fi series, which does forced sample rate conversions internally.
Internally in the chip everything is done on 48kHz, and everything is converted to that, then to whatever you set the output to afterwards.
The chip used is from early 2000, 2005 at latest.
For the Audigy RX.
Usable for RX. Problem is that chip internal stuff that causes issues if everything into the card and out of the card isn't 48kHz.
And lot of stuff still use 44.1kHz.
And it isn't really upgrade from proper current MB sound chip in most cases.
Z SE was the first "worth it" Creative card.
Asus Xonar SE 5.1 not available?
Or higher up Asus Xonar AE?
Also you could try various drivers for the MB sound chip.
The one MB support pages have.
The one Realtek pages have.
The one Windows Update installs for it.
Often not all work the same.
So probably the second from top download for 64bit, or fourth for ZIP version.
No idea...
Much more expensive Asus STX II seems to suck in implementation when tested in 2018.
Ok, moment, need to reread.
Ok, that was not worth the money, but no major issues except at -90dB or lower.
No idea about the cheaper Asuses.
So should be better than MB in most cases, but for same price well selected USB sound card would probably be better.
Also doesn't have connector for use with case front audio headers.
Check how much it would be direct from Creative?
Sound cards are long dead, yes
Because idiots think that moving the sound stuff out of the case helps with interference that comes with the ground and power.
And still using the ground and power from the computer...
There are lot of USB sound solutions that have gigantic interference problems from those.
Yes, it helps againts radiated interference, but protecting against that is easy even internally.
And it is much lower power.
So if the USB sound solution doesn't have separate power supply, it doesn't help in any way in protecting against interference from anything in the computer it is connected to vs. internal card.
And the actual solutions to get rid of that interference are exactly the same.
I mean usb is also more plentiful than pcie
But the secondary computer:
With the CPU, MB and RAM of the previous main.
And rest from the previous secondary.
are you gonna put a gpu in it or just leave it as is? i guess if you dont really need one there isent really a need to put one in
Running with the Intel IGP.
so your just not doing anything graphically intense whit it im guessing?
Just for office type usage.
Usually not even video playback.
ussually that works but somethimes the Igpu's are such a disapointment they cant even do that lol....
That is Generation 9 GT2.
I have Generation 7.5 GT2.
About same Gflops when the clock speed and power differences are taken into account.
If youβre just using headphones, usb-c to 3.5mm is all you need. For example VE Abigail (pro) or jcally jm6 are good.
Abigail has included usb-a to usb-c adapter in the packaging, donβt think thatβs the case for jcally.
yea the difrence is one is a capable desktop Igpu whit a capable cpu one is a incapable laptop igpu paired to a 2 core 2000MHZ cpu and 4gb of ddr4 single channel ram
i was gonna upgrade this thing to 8gb of dual channel ram a whatever speed i could get but i decided not too because this cpu is a disapointment
the desktop version is ok because it has a turbo clock of like 3400mhz or something but the laptop varient dosent have a turbo clock so its just yet another pointless dual core cpu
in that sense the Igpu would probably be tolerable if it wasent on a lacking cpu and ram config
efficient idles generally does basic tasks ok
And if you got into the BIOS and set the IGP static allocation to lower amount from that 2GB.
its a laptop bios you cant do shit other than battery efficiency and like some display check or something
if i would be able to i would have a while ago by now
Or maybe you are blind again and haven't checked all the places as you didn't think them relevant?
Like IIRC last time.
What is the computer model?
Ok, wasn't in there according to the manual.
Have you reset the BIOS to defaults at any point?
If the specific computer had more RAM at some point?
no its in its factory configuration and i havent reset the bios only updated it
I had forgotten how great Ahti is.
With all the word for word translated sayings from finnish.
And in some cases using wrong translation.
Reindeer = Poro
Palaa poroksi = Burned to a crisp.
Part of one Ahtis line: burned to a reindeer.
Arctic P14 (PWM)?
I just replaced all the stock Fractal fans with P14 PWM:s in this upgrade.
Had the stock Dynamic X2 GP-14 3-pin DC fans up to now.
P14 are inexpensive, and better than the stock Fractal ones.
AFAIK still one of the best general use 140mm fans if you can do without bling.
If you hear it tick, there is a chance it will go π₯
Speaking from experience π
was to say, do it from bios 
You cannot do it via BIOS in most MBs.
Both my asus and this asrock MB can turn on/off fans in BIOS.
You probably talking about something else?
Are you mixing the low speed warning limit with speed control?
No, I go to my bios. I see all my fans and there speeds/thresholds. I turn off a fan. I turn it back on.
fan curves are also in my BIOS, for both mobo's.
In my current MB, for PWM fans the minimum is 20% speed, for DC fans the minimum is 50% speed IIRC.
You cannot set the control lower.
huh ...
You can disable the fan control, which runs the fan at 100%.
And you can set the low speed warning to be disabled.
But neither stops the fan.
In older MBs the minimum speed for DC fan was often so low that they wouldn't start with it.
But would try to turn at times and immediately stop.
Yes.
Hidden by the shroud in most pics.
I was able to stop my fans somehow though π
Looking at the fan configuration in the bios is a bit confusing.
My CPU 1 setting is silent. MY CPU 2 is set to standard. And all my chasis are set to silent by default. Why is my CPU 1 fan set on silent by default π€·ββοΈ
So in this case with "BIOS" we have meant the UEFI Setup / UEFI Configuration Utility or like.
And even with BIOS the config screens were just small part.
So what is the different between "CPU FAN 1 Setting" and "CPU_FAN2/WP Switch"? π
Water Pump.
Is the output for CPU fan or Waterpump?
First is probably lower max amperage with control, second is higher amperage without control.
all air cooling here.
I mean it switches the output mode of that fan header.
Many current MB:s have two linked CPU fan headers and separate WaterPump header.
Yours seems to have CPU fan header, then combined CPU fan 2 header and water pump header, with BIOS switch between the modes.
And isn't fixed that the CPU2 follows the CPU1.
Also, I only have one CPU fan so π
But anyways, if you want to try to stop a fan, you need to switch the Fan Setting to manual or like and then set the speed very low.
Ah well, I figure it out when I install the new CPU.
Most fans stop under 5-10% pwm
In DC mode.
And some PWM fans have fan stop mode once the PWM goes low enough.
or else I just unplug the fan π€·ββοΈ I mean, if you want to find what fan is making noise.
Unplugging is the easy way.
For that purpose, I think it's pretty safe to unplug fans yes.
The fans have protection for it I think, vaguely remember someone mentioning it. So if you block the fan, it should be ok.
wow, who would have expected that 
Probably not from the fan being in the way, just from less forced airflow.
eh, it wouldn't surprise me cheap knock off fans don't have anything.
yesterday, when I couldn't sleep. I suddenly started to think. What if my motherboard is fucking over the CPU. If I buy a new CPU and put it in, then I have two fried CPU's π
I am slowly losing confidence in Asrock π
and from GN I know it's possible that a motherboard fries other hardwre.
Yes, as the MB controls the voltages to the other components.
Even when you expect that a motherboard has protection build into it. It was not the case. Even the top line of Asus motherboard didn't have proper protection.
So it's still possible the motherboard is faulty.
OV vs. OC protection.
Asus was lacking the second in early boot.
Which doesn't matter if the CPU isn't toast already.
But yes, broken VRM regulation can kill parts.
Or shorted components on the MB.
But this doesn't seem to be either at this point.
Would need to measure the CPU rail outputs from the VRM with multimeter to get basic idea if the software reported values are sane compared to the actual VRM output.
But need to remember what differences the measuring point can make.
I am not good with multi meter.
And that the MB would have VRM feedback that works, but is offset is very remote.
Usually it just breaks, or works.
And the SVI3 values are measured(?) inside the CPU.
Ok. Let's hope π
So you can compare the SVI3 values in HWiNFO and the MB section values for the same rails.
Should be pretty close at low load and then the spread should get higher with high current.
As usually the MB value is measured from either the VRM output, or from before the socket.
And the voltage droops after the measurement point.
Good to know. But I am not going to do that. I really struggle with a multimeter. And I don't want to damage the motherboard.
Meant the MB reported values in HWiNFO
and with struggling I mean, I don't know what I am reading. Or how to use it.
Losses to the board, socket, and inside the CPU before the actual die(s).
Oh like that. I once had HWinfo file with everything logged to the point my PC crashed. But I didn't find anything back then.
Ok, I am following you now. Sorry for the confusion.
The multimeter could have been compared to the MB reported values (not the CPU reported ones) to see more closely if the MB reported ones are correct or not.
But you can also compare the CPU reported and MB reported values for VCore, VSoc and VMisc.
Just remembering there will be drop from the MB reported to the CPU reported because of resistance between the two measurement places.
And the difference gets larger higher the current is.
I think there is a slight difference between the two
if I am reading correct values.
Can be to about 0.05V for VSOC and 0.1V or higher for VCore under load easily.
But as long as the CPU reported values aren't higher than the MB reported ones, then everything should be fine.
And the MB regulation is working correctly.
Bottom graph is my motherboard. Top graph is what my CPU is logging.
So 0.015V difference where MB is higher, absolutely normal.
You see the big dip in the CPU sensor, but you don't see it on the motherboard for examdple.
Different averaging, and that kind of thing happens when load ramps up.
The voltage getting to the die droops.
As there is latency in VRM upping the current delivery.
But the HWiNFO just doesn't get good enough data to look into that kind of results.
Would need oscilloscope.
Understandable.
I wonder, if this data is available, shouldn't it be possible to monitor this and build in warnings? Into the driver of AMD or something? So if the CPU is logging it's values. And the motherboard is logging it's values. Can't "something" (driver, OS, ...) monitor these values and act when something seems fishy?
Or isn't this data, like you said, precise enough to pull it off?
HWiNFO is getting one momentary sample per X ms (configurable to degree, default one every 2 seconds), from sensor that can have done anything to the data.
For things that change at GHz speed.
Current VRM often work in 300-500kHz range, and CPU load changes at clock frequency.
So you need low MHz range data to get something useful to diagnose possible VRM output dips at minimum.
Aka a very pricey scope
I see, I remember our talk about MHz and I understand the short comming of the inbuilt sensors now π
No, you cannot have HWiNFO ask the board sensors for new values 1 million times per second.
The sensor isn't checking anywhere that often anyways.
1MHz, 1 million times per second.
Not fully related, but I was watching some videos about airplanes that crashed. And one plane had all four engines failing. It was because of vulcanic ashes in the sky at night. Back then they didn't know it could cause major harm to airplanes. Now the interesting part was why the weather system, in the nose of the airplane, couldn't see the ashes. And it was because the bandwith it uses, has too big of waves and can't pick up the tiny particles. Was interesting.
The weather radars are tuned to work on frequency that water reflects.
Exactly. Maybe not today anymore. Since they learned volcanic ashes are deadly for airplanes.
Still are, you cannot easily have multiband radars.
You just limit flights when there is possible danger.
When the last great icelandic ashcloud was that closed the european airspace for weeks, military jets ran flights with sample containers to determine how much ash there was in the air.
Not even military radars etc. could get anything usable.
And that was tens of years after the case you talk about.
The ashes contained particles of glass, they perforated the engine blades and also, because of the heat of the engines it melted and got stuck to everything inside the engine. Thus the engine failed. After a while, it cooled down and the pilots were trying to reignite the engines, this probably broke most of the glass inside and eventually the engines worked again. The pilots climbed to higher altitude, back into the ash cloud (it was night so they didn't know) and again a engine failed. That was the moment they realized they had to fly lower. The correlation between altitude and engine failing.
The ash melted, stuck to the parts and blocked cooling air outlets.
Oh it did way more.
Yes, it acts as sand blaster too.
But the way low concentrations cause problems is the blocking of cooling air, causing the engine to overheat and melt.
Plumes of volcanic ash near active volcanoes are a flight safety hazard, especially for night flights. Volcanic ash is hard and abrasive, and can quickly cause significant wear to propellers and turbocompressor blades, and scratch cockpit windows, impairing visibility. The ash contaminates fuel and water systems, can jam gears, and make engines ...
That specific case was much different.
And the engines restarted after getting out of the cloud.
Just at much worse power.
That's one of the planes.
It's the worst single thing I done this week. Watching videos of what can go wrong in airplanes π
But the problem for civil aviation is that even much lower concentrations destroy the engines.
Not via sandblasting, but that melting of the ash.
If I am not mistaken, it depends on the altitude. A volcano has two types of smoke. A dark one, and a white one. The dark is mostly just ash, rocks etc. And don't go up very high. The white smoke is the one that hit the airplane. It goes very high and spread far and wide.
Because the turbine blades after the combustor have this kind of shit:
Depends on eruption.
Some can get the ash to 20km.
Most don't.
All I remember from the documentary is that there are two parts of smoke π
With ofc, different elements in it. The glass is in the white smoke.
That specific icelandig eruption got it way up.
This recent one didn't.
Which I talked about, that closed almost all european airspace for any flights for weeks/months.
btw, there are also watercool solutions for the oil. So that the oil actually cools of the engine.
Not for the turbine blades.
But this is done briefly during lift off or other instances. To get more power.
And it wastes a lot of oil, so you have a black plume when the plane lifts off.
No, that was just water injection.
Oil cooling is very common in aviation and/or other vehicles
No oil involved.
According to the documentary it wastes 40% of fuel.
It just spits it back out at the end.
Fuel yes, oil no.
Sorry, my bad π
You inject water after the compressor to cool the air, so the combustor can be run with more fuel without melting it or the turbine.
In my language, we use those words interchangeable.
But then you lose lot of the extra fuel used to heat the water.
Yup, pretty cool stuff π
Oh weird, those are very different words
Indeed, I had a brain fart.
We have a lot of words that have so many meanings and are based on context.
But I'm off to bed, was planning to go hour ago already.
Ah, ok. Good night π π
Thatβs early
for you maybe
Yea my day is half over
i came across that today when i was trying to see how i would get into the tire discount store tomorrow ... lol
Wait. That thing is real?
it was on google maps street view
that's what you call an american mini
Oooh!
smallest american car
lol
Low FPS google street π
ive seen some weird mashups on street view π
I don't visit it that often
i do when im heading to something ive not been too before, and i kinda wan a mental picture of just how crazy the roads are near it
"should i be heading this way on the road? does it even have a turn off?"
Oh, yeah. That's when I use it too.
i cant trust GPS to route me the right way into a parking lot around here lol too many medians or block walls
my gps still doesnt realize there is a cut across a median getting in my mothers neighborhood .... it keeps trying to route me 2 miles out of my way around the back side π
I love to play games with my GPS. Who knows best, the GPS or me. So sometimes I take different routes to beat the GPS time π
haha
but I turn it on often because of traffic jams. Although I know the way.
a couple places it wants me to turn and go a different road, but im like 'naw, im staying on this road till further down' ... and when i pass the turn, it 'recalculates' and then the 'estimated arrival time' lowers by a minute or two LOL
laughs in german
doesn't that feel great? 
No street view for you
You whish.
doesn't that get auto censored? π€£
It's kinda cool to see the whole world.
google street view spreads like a virus over it π
now i know to find the blurred house in germany and one of them ius yours
have fun looking for it
yes the one house in a country that was infamous for blurrying so many houses that google just stopped doing it
look at it, what is that? 30-40%?
Yeah, Google won some other court case.
Was a dope time, Google thought they won when a court ruled "streetview is allowed unless the residents make a petition to block it" - and then German paranoia hit
plague inc
Odysseus successfully landed
hopefully we'll get the images from the camera watching it land
Why would you be on f*ndom anyway when minecraft has a proper, independent wiki?
There are search filter extensions where you can block all fandom sites from appearing in search. Highly recommend it. Fandom is horrible
I am so confused. I just had a website that was unable to load when I am on ethernet. But instant loads on wifi. And now the problem seems to be fixed for ethernet too.
But I could visit other websites with ethernet. Almost as if the specific website was able to load on WIFI network but not ethernet network.