#off-topic-tech
1 messages Β· Page 34 of 1
INNER JOIN? more like inhale joint amiright
Where did that come from?
I had to learn more SQL today π¦
Oh wow, I'm sorry
left join, right join, outer join, inner join π
yeah I hate it
I just have libs handle most of the SQL legwork
I JOINED the SQL hated club
yup, my machine is below average, not even in the bell curve...
And I can google how to write the queries I need π
Userbenchmark is very bad
3dmark is userbench? doh
wait untill someone fucks up the DB and you gotta run some SQL to fix it. You will love SQL.
Oh, mixed up the UI
@charred pewter did you forget to enable XMP?
yeah I live in the richest country in the EU so I still have to do my tests on paper obviously
Oh, didn't realize this was for a test
I thought you were actually using it
says its on
Oh no I will at work next month
but at work I at least will learn to use it how it's actually used and not in an isolated enviroment on paper
my experience, I have to use SQL so little, that when I do, I use google to refresh my memory.
a.k.a. I don't use it enough, to memorize it all. And why else do we have google π
select * from DB
drop DB; 
What flavor of SQL are you learning?
Since the syntax/features can be different depending on which SQL you use
MySQL
ah π¬
So the most useful shit ever
And also because the school doesn't have to pay for it
Would be a shame if we actually had to invest into our future generations
mariadb works too, have pcre and is free
I used h2 for a bit.
in memory DB. - h2database.com/html/main.html
This is what I love about it:
So you can just access the DB with a webbrowser to check data, modify it etc. No need for client.
we use phpmyadmin
Just put H2 in your web-application. Spin it up. Navigate to url. Enter password etc. And you good to go π
And it's free. You can use it in your application and distribute the application.
it does have drawbacks. Like it's running on the same server as your application (since it's embedded). A lot of SQL you need to hire an extra server π Or make sure you can run a DB on it.
put an index on the db
odd... soooo ... i guess my old 3dmark results where i was on the high end of the average curve with a 3080, was just because no one serious about overclocking ever bought a 3080 (or they couldnt actually overclock them cause they thermal'd out even on STOCK) ...
but the 4080, ive not overclocked at all, and everyone who reports on 3dmark, HAS... which is why im so below the average
i guess i need to get crackin =p
are you sure about that?
I did lot of stock runs with my 7900xt which is very fun to OC
i duno
i cant figure it out ... ive got all the latest drivers, bios, settings, removed g-sync and vsync as it says, ran it, ran it ... and i cant get anywhere near the curve (but i've YET to try overclocking)
because stock will be about here
thats what i woulda thought .... my prior rig was:
π€· ... im fine if its "because im not overclocking" ... but i just want to make sure ive not got something broken in here killing my results
i need to dig up an extra SSD and do a clean wipe bare install of windows11 and try again
(just to rule that mess out)
heh
(also a lot of peeps with multiple entries which is kinda cheating
You mean on the leaderbord?
the leaderboard I don't understand π¦
they should only put highest per single user on the leaderboard.
timespy normal
It shows my score at the bottom too which had me confused
Time to run time spy extreme and check what I get with that OC
Yeah I'm 23 with normal time spy on the leaderboard
CPU score gonna probably drag me down...
that's a new one
I think he didn't use a case? He OC'd more, lower temperatures π
Like the CPU is still running hot, but GPU only 37, while yours was at 47. And I assume you used your current watercooled rig.
hm, someone with the same specs as me said they ran at stock with 3dmark and was getting the 28k scores .... so something must be wrong with my windows, or hardware.
im gonna dig up a sdd to do a test clean windows/nvidia/3dmark install and see
sounds like a 'tomorrow problem' π
I wouldn't be able to do that. I would be too curious.
D: my rpi appeared to be frozen I couldn't connect to cups or open remote desktop.. does linux keep any logs that i could check as to why
congrats. try not to fry your gpu as you try for higher scores
I'm not tryna beat that tbh π
For real, see if you can sell it some some super serious benchmarker for a markup and buy an XTX
Or even just a straightup swap
yeah I'm too iffy about used GPUs and idk if I could sell that
Stipulate a new XTX that you could register.
SOmeone out there might want the inside track to a top 10 score and would pay for it, people do crazy shit to get on leaderboards
for real?
I've heard of crazier things done for the sake of benchmark scores. Holy shit, you should see the insanity in laptop Discords.
Because then you need to acquire the whole board if you want someone's golden sample.
How do you prove that the benchmarks was produced by that exact hardware? Like it's not photoshopped or a different machine?
No idea π€·
It's in combination with a 5800x3D
Which holds it back a lot. I currently don't know the real "power" this gpu has in timespy extreme
Benchmarks rarely benefit from the extra cache
Which is a shame tbh, cuz that means you'll only see them shine in games
I wonder, can't they make a benchmark specifically for that?
It exists and it's called "Microsoft Flight Simulator"
with 3dmark, its being sent as some sort of key encrypted data packet to the machine.... so its really dificult for someone to spoof send a data packet with bogus details
unlike say... leaderboards in games where you see some jackass with a 'level completion time of 3 seconds' =p
cool
well, this is embarrasing ....
so i ran from a clean windows11 install, fresh nvidia drivers.... got a score above the dashed blue line "average" ...
soooo clearly my current windows11 has some crap in it that is hurting my results
well, you know the fix.
peanut telling me "I TOLD YOU SO!!!" ..
fixed the problem on my win11 install:
Burning interest demands an explanation! you say??? well, ok... here goes:
So RTSS has this global setting to limit FPS (so some games dont go flying off the handle trying to render 5000 fps in menus for example). I had it set to "144" because my monitor is 144hz. Ahem. Well, that apparently makes a difference in 3dmark when you finally get a gpu that goes above 144fps in the tests (my 3080 never seemed to hit that to show the score hit). So, when I turned it off, and re-ran the benchmark... im now landing 'in the middle of average' where I would have expected to be all along. sigh
hey yall are smart and im ont sure where to ask, do any of you know a tool that can detect duplicate images?
i have a ton of pictures on my phones and i want to have them all on my laptop and remove duplicates
some duplicates are cropped versions of the original tho and sometimes i downloaded it frmo discord again to send it to someone else so some have gone through compression too
is there an easy way to detect likely duplicates?
depends on your patience
I recommend you start by gathering all the images in one root folder and use a tool to find exact duplicate files. Toni arts easy cleaner has built in duplicate finder. there are others.
that sounds exactly like what im looking for
once actual duplicates are found and eliminated, you need a tool that can analyse image similarity. I had good results with Awesome duplicate photo finder, but 1. It will take literal age to run on bigger folder and 2. interface is clunky and requires manual assesment. there may be better tools out there.
well i want to decide manually on each possible duplicate it finds, dont want to accidentally delete stuff i dont want deleted
and i dont mind letitng it run over night if necessary
the first step (actual duplicates) is easy and safe to run and you are sure the files have same content. for step two, take care to not remove the better quality version of the image. I recommend you backup whole folder to zip archive before starting work.
good idea, thanks
I have made own tool for finding exact duplicates, but dont have any release ready versions. Unless you are fine with running exe files from random stranger on discord.
i think iyll use this here, sounds promising: https://github.com/jhnc/findimagedupes
I strongly recommend you use programs with actual gui, not command line tools.
normally the originals are in DCIM folder or similar
the edited in Edited or similar
edited are in dcim too for me
you could find by name normally then
like 476533_edited.jpg being edited off 476533.jpg
if you really mean multiple similar shots then it's all up to your human eyes
the programs i found calim to detect similar images automatically
the gtx 1660 sucks ass
Its a low end card. Unfortunately its price hasn't really gone down enough to make it worth buying. You'd be better off buying an rx 6600 or hell even a 6500 XT
I have a 6700xt in need of repair
I'm using the GTX 1660 S atm
And even with a 2.1ghz oc
It's still ass
Idk I sent a follow-up email as it's been 5 days since I last heard from asrock
Fuck I want my nice gpu working again
I don't wanna buy another gpu
@charred pewter Have you checked what HWiNFO says about low power state "time spent" percentage (C-state Residency on modern AMD), and core effective clocks.
As the normal frequency "idle" frequency doesn't mean anything anymore with AMD, not sure about Intel.
As the power saving core frequency and power gating is much more granular now, and the Windows reported core frequency is absolutely meaningless on AMD.
My unused cores "idle" at 3600MHz, but the effective clock is 50-100MHz, and they are 90-99% in C6 power down state.
The effective clock is "high" and C6 residency numbers are "low" as I'm running 5 core load testing.
Probably, but that is the reason for the ping in first.
But anyways, stopped the loads.
Windows reports 4.2-4.5GHz core clock while idle, HWiNFO reports 3.6GHz Core, and 40-50MHz Effective.
With 0.5-1.0% C0 residency (full power core on), 2.0-2.5% C1 residency (frequency gated?), and 96-97% C6 residency (power gated?).
With average 0.035W Core power usage in HWINFO.
This with Balanced power plan, iwhen I set to High Performance, only changes were that the HWiNFO reported Core clock jumped to 4.2GHz and Task Manager shows now 4.4-4.8GHz jumping.
Actual power usage while idle is exactly the same.
So for AMD, the reported core frequency doesn't mean anything, as the CPUs from them now gate below the "old" layer at very fast timescales and the changes aren't reported to OS.
Haven't kept up how Intel now does the same thing, so not as sure about them.
This was the reason AMD wanted people to use Windows Power Plan (that Ryzen Optimized one that came with the chipset drivers) with CPU set to 100% to 100% load, as in Zen1 and Zen2, that fast system got confused when Windows tried to set core clocks.
Zen 3 and Zen 4 don't care anymore and just ignore what OS tells to use as frequency.
what settings would u use?
Why are you trying to contact them?
And in what way?
Rma and the US branch Said they forwarded the issue to asrock canada
So might be that the US branch forwarding was done wrong.
I sent a follow up today let's see if it has any effect
MLID apparently has a big Blackwell rumor video out. So be prepared for the rumor mill to go into overdrive as all the usual websites churn out their SEO content.
Also, another leak to War Thunder Forums, about Bradley this time.
Again "just" export-restricted stuff.
I think Russia captured a Bradley in Ukraine. My mind immediately wondered if it was someone who has their hands on that one.
Nah, just someone in US getting the stuff (legal for US persons to have) and had posted it onto elsewhere in the net this month.
And someone used few pages for some argument on the War Thunder forums.
Because what's a little national security when you've got an internet argument that needs to be settled. π€
Got a brilliant suggestion from my mechanic on how to "cheat" on car emission test: Change your fuel filter before the test if possible and drive for at least 15 minutes on 3.000 RPM (highway would be an ideal place)
I can achieve that by driving around 90 km/h on a highway in 5th or 6th gear
And my 14 year old car passed the test btw 
Just normal thing to heat up the engine.
The people doing the testing will rev the engine for time if the initial result isn't ok until it is or the numbers don't change anymore.
In here I mean.
Because both the engine, oxygen sensors and the catalyzers need to be heated to operating temperature before the test is valid.
And that last one takes time and heat.
Here, they will test the emissions without revving the engine so there is higher chance of failing the test and you have to retake the test and pay again....
Because profits 
They only tell you to leave the engine on during the test but that's not enough imo
No, the actual testing even specifies what RPM the engine should be on, and idle isn't it.
IIRC somewhere between 1500 and 3000 RPM.
And most cars idle at 750 or so after the engine is hot, 1000-1200 before when it is running fuel rich before the oxygen sensors work etc.
I lied
iirc C6 is fully parked
it will only jump into other state for processing
it is mostly disabled
Yes, the power usage is very minimal in C6, but the question is what is actually happening behind the scenes.
But it can jump between C6 and C0 faster than the old system changed between frequencies.
Much faster
So I wasn't completely sure, but I did say it is probably (almost?) complete per core power gating.
there is more than gating
a core is not an engine it is a system with its inner resources
basically setting up its registry in parking state
to shut down any useless back processing
All the other stuff is needed to do the power gating, which removes the power from the parts of the core that can be completely shut down.
And still be able to restart quickly.
I mean hardware ones
Older CPUs did power gating too, but the restart was SLOW.
And clock gating doesn't give anywhere near same level of power savings.
I mean gating in itself is not effectively disabling anything
Power gating stops the whole voltage input to the specific core.
So everything behind the power gate is completely disabled.
even for power gating
you reduce preload prefetch and forward caches stuff
oh sorry
I got the word in the wrong way then
Clock gating disables feeding clock to the core, so it still gets power but doesn't do anything.
that gating ok
But the parts clock gated still use the static power, just not the dynamic part.
Clock gating keeps the state, power gating doesn't, in the gated parts of the CPU.
So the C6 should be that the important parts of state are just clock gated, and the rest of the core is power gated, probably.
And C1 is probably just clock gating both parts.
as I was following OC on GN YouTube channel
I noticed the engineers was not expecting any power controlled prefetching preloading and forward caches behavior
those plays huge part on power efficiency
and so I felt bad for the cpu
feels like amd CPUs could have even more potential at low level performance
power efficiency is not only about power draw
but also about what you do with the power
Yes.
But this was just about the power saving modes the current CPUs do when idle.
And why the frequency reported by many softwares is still high.
Even when the CPU core(s) are actually in very low power state.
looks like amd only relies on caching speed catchup when cpu speed goes down to increase efficiency
I'm not talking about frequency
that's the point, it's not about frequency
but about how far you push for potential miss predictions
these cost power
If you check up, the whole thing was because Frigit last night wondered why two of his P-cores kept very high frequencies, with lot of spikes, in Afterburner(?) graphs, while the system was supposed to be idle.
Part of it is that the Afterburner does polling, so it wakes the CPU and it boosts high to process that request, part of it is because of these changes in how the CPUs idle.
I mean there is also windows stuff going on for the spike part
And was explaining what current AMD CPUs do on high level.
And that the frequency doesn't really mean anything on them, and that I didn't know how the Intel does the same.
Intel had marketing about doing the same years ago
except the Intel CPUs reported lower frequency to OS
I don't know today
I knew back at skylake
The much slower old style still in that time I mentioned.
That does report clock speed changes and is partly OS request driven.
tens of ms response time compared to ns timeframe changes in that new AMD system.
Point was that I'm not sure about the 11th-14th gen Intel how close they are to AMD style, or what they use, instead of the "old" system.
but you were also questioning what's going on in C6 state and other stuff
so on the topic about what's going on
I was rambling about amd CPUs not doing preloading /predictive behavior reduction for power saving
3600mhz with every single alu processing for nothing is no near 3600mhz with single queue fetching ops few alu power efficiency
looks like amd optimize the fet use very well
but feels like they still can do better at higher level
maybe the engineers are held back by marketing to have stuff to improve over the years?π
Meanwhile Thermalright: our engineering department is our marketing.
There's no better marketing than a good product at an even better price
They would really need a marketing department too, to actually get the stuff into retail...
If no-one is selling the stuff, it might not exist.
No matter how good it is.
Not marketing. Distribution
And they have third party sellers doing that job
Well, Sales Department.
Thermalright is baller period.
Grey market isn't Retail.
Difference is in warranty.
They cheap out on advertising and documentation (which is a good thing, keeps cost low), but they also cheap out on a proper retail structure
And when even the official retail partners just cannot get the stuff to sell?
Lack of proper retail structure
I just use Amazon tbh, return policy is quite generous
And it's all shipped by them anyway
So, in HWInfo, the effective clock speed for CPU, is just a theoratical one? effective value is the average clock across the polling interval used in HWiNFO.
Polls the tick count of the specific core.
And calculates how many clock cycles the CPU as actually been active since previous polled number.
I am trying to wrap my head around core clocks (BCLK?) and effective.
BCLK is Base Clock, just generic clock rate that everything else is derived from.
The there is the Core Clock, which is the speed the core "runs", but as it can be clock gates, and then doesn't actually run, it can do less actual cycles than the official frequency suggests.
That Effective Frequency looks at the actual clock cycles the CPU core has run between two moments in time, and calculates that frequency "rating" from it.
Runs at 1000MHz, but only actually has clock fed to it for half the time, so only runs at effective 500MHz.
And processes same amount of stuff as 500MHz clocked CPU would have.
So, does that means it runs at 1000MHz or at 500MHz (in reality?) What I understand from HWInfo forum, that the new one, effective, is more realistic.
Runs 1000MHz for half the time, at 0MHz half the time.
With very fast switching between the two as needed.
So, effective is a bit more accurate
Thanks Baldur, what frigidman posted yesterday got me confused.
So I digged into it a bit more and got more confused π€£
but now it makes sense, windows task manager also uses the core clock speeds and not effective one I think.
Yes.
That was the whole point, that the "Actual Frequency" doesn't really mean much anymore, and you need to look at the effective to see if the core is idling or not.
Yes, same for Afterburner etc.
Yeah, exactly what I was investigating
Does it still work to spot throttling and track boosting behavior under full load?
The Core Clocks is the max the core can run at that time.
So yes, if there is some limiter in effect that will drop.
The power savings just can make that the effective is below that current "max".
But for AMD, you also have "clock stretching", if the voltage droops, the core shuts down and the effective frequency drops.
So if you set too low voltage, or use too droopy VDroop curve, you lose performance to that on load changes.
Without getting errors or crashes.
Or even all the time in some cases, as the CPU sees the stable core voltage as too low, when it is asking for X but only getting less because of the intentional vdroop.
Partial idiocy from AMD that their default load line calibration is near 0...
So lot of spikes, and if you set more droopy one, the clock stretching kicks in.
I haven't used it before, but a friend recommended it
I use dupeguru.voltaicideas.net in a docker container on my NAS. Very good stuff for duplicated images π
dupeGuru is a cross-platform (Linux, OS X, Windows) GUI tool to find duplicate files in a system. Itβs written mostly in Python 3 and has the peculiarity of using multiple GUI toolkits, all using the same core Python code. On OS X, the UI layer is written in Objective-C and uses Cocoa. On Linux & Windows, itβs written in Python and uses Qt5.
any suggestions on good library to work with timeframes in Python? (or at least date-frames)
i just need something to solve i have a bunch of timeframes, some of them "touch", i need to know which ones are missing to have a complete target timeframe covered, and rn it feels like writing a small lib for trivial problem
the who?
why
cool put some nice pictures in there and put em on the wall
im so bored of this week bruh can winter break just arive allready so i can finally finish my setup instead of just pushing it back over and over
Who said you need to wait for break? Just do it after school
A really well cable managed build doesn't take more than three hours, depending on complexity/skill
Pulled pc-part trigger too, ordered another identical set of 2x32Gb DDR5 kit (F5-6000J3636F16GX2-TZ5RK). Will see how XMP works with those 4 sticks on Z690 Hero.
@night girder you take issue with my comment?
Considering I've had some week-long builds in the past its not exactly true but it also doesn't need to take more than three hours
I just find it funny. Took me a day to build this machine and get it up and running. And the cable management is a 4/10 I think π. In my defense: I read the manuals. So that takes time 
I read the manuals for stuff I need to know
And you don't know if you need to know something in the manual before you read it.
I'll skim the section in the gpu manual detailing exactly why it has the best fans ever and how they'll save the world
Yeah
I read the entire thing, but skip/skim some sections
@twin dew would you rather get a 13600KF and reuse old DDR4 RAM (likely 3200 cl16) or get a 12600KF with new DDR5?
No real idea right now what the performance difference between those options would be.
This is the CPU data that I have. But idk about the RAM portion
So no comment on which.
Easy to check the CPU side, but the problem is that RAM side, how large difference is there on any 12th-14th gen Intel CPU in games between the specific DDR4 and DDR5 speeds, including the primary latencies in part.
And I'm not going to go looking for that info right now.
There have been some tests about that, but don't remember more than that.
Might have been on Hardware Unboxed?
no, i've not. not sure it really matters so long as something is churning away 15% of core... its enough to keep kicking it into boost mode. until i figure out what is eating cpu time like that
@edgy hazel i figured out why I was getting such shit scores π #off-topic-tech message
yeah had a similar issue after updating my AMD drivers which enabled a framerate cap to 60 and I was so confused why everything felt like shit
doh!
this also explains why the 3dmark warning "turn off vsync and gsync" (and when i turned those off) didnt result in any change in score lol ...
it was STILL capping at 144 lol
and you checked W11?
this is why having a nice 'clean' install helps figure out stupid problems
depending on how you're set-up it's the fastest solution
I want to know if it was user or OS.
i actually ran a test set down to 1080p ... didnt help haha cause it was still capping at 144fps due to RTSS
oh it was user .... user forgetting about a setting that was made a long time ago π
user. got it.
i always used RTSS to do framelimits, as the one built into nvidia cp was hit or miss
Do you include desk cable management? Because I'm looking at mine and not seeing it done under 3 hours.
3 hours for the cables in a pc... yeah, doable. 3 hours for everything in the office? hAHAHHAHA
its been 3 years and im still working on mine =p
indeed
not even talking about office building. talking about my private desk in my home.
... that happens to have a desktop, two laptops, two screens, a tv, 5.1 crappy speakers and whatever else I'm forgetting...
and im always annoyed, cause when i need to go dust/clean... im like ... ya know, it would be better if these cables were this way ... oh wait, that messes those up... erg, i need 1 foot longer... time to order more
I have my cables in handmade velcro sleeves. keeps them together, only need to dust off the velcro... but now I need to replace some cables and uh, thats gonna hurt
then i see 3 cables, and 1 cable is a slightly different thicknes... now i want all three to look the same, so i need 3 whole new cables ...
I got a macbook, two monitors, a dac, two yamaha monitor speakers, a desktop, a USB hub, a ethernet adapter, a NAS and a switch.
all together 
I tucked away the cables under the desk for now, out of sight π
I have several cables dangling by my feet bc they were addedcafter velcro was zipped up
but the switch has 5 ethernet cables running from it and it's not clean
but that's a worry for another time π
unsolicited pc pic - the "birthcord"
there is 3 or 4 more behind desk, for each screen and laptop, going into common 'cable exchange tray'.
braiding it into the tray probably was not my brightest idea.
so now (after repeated failed attempts for windows activation) ... ive got this lower right corenr watermark saying to activate windows... AND that even shows up over games now too... and windows, and interfaces, and steam chat
whats odd, is my HTPC downstairs, i didnt register windows on it... just letting it run in freebie mode (i mean why bother?) ... but it has no watermark! and its been running for over a year straight like that
do you have microsoft account logged into pc?
yeah, prolly why huh, my htpc is just in local user mode, no msn account
I got my main pc auto activate licence by just logging in. Do you have a sticker for the pc?
your HWID is stored in a DB, when clean install, Windows checks DB with HWID, if you registered once, Windows automatically activates license.
dang this computer hates me
all I did was disable csm support to enable secure boot now it doesn't have image

If the specific MB has ever been activated before, it autoactivates again when it gets internet.
And if you have specific type of retail lisence for Windows, it is tied to your MS Account and activates computers in specific cases, not sure what the limits are how many you can have activated like that at one time.
well, from reading ... MS killed all the old windows 7 keys that were upgraded to win10/11 digital licenses... which is what my original key is, a win7 key that i had upgraded. but since they killed those.... my msn account wont re-activate this new motherboard
it sure is weird. I have moved several (two) sticker licenses between computers, they were supposed to be oem but had stickers and normally activated.
Once a system is activated, this activation cannot be removed because the license is stored in the Microsoft servers and not in the userβs system. MS checks the hardware ID (HWID) and if a license is found in their database, the system will automatically activate.
try going to change key and punch the original sticker key. it may work.
Because the old Win7-Win8.1 keys get converted into Retail ones for Win10 and 11.
Even OEM ones.
yeah, i tried putting the key in manual again... it churned (cpus 100%) for about 3 minutes. then was like "nope"
in august/september MS made some changes to activation of licenses.
so the license i HAD on my account, just kinda vanished with my prio motherboard... even though ive removed that HWID device from the account (which should have freed up the license to apply to another).
Just disabled activating Win10 and 11 with pre-Win10 keys.
Kinda. Since it still worked π I was able to upgrade W7 key.
i mean, this all worked this way LAST year, but this year.... ms changed something acouple months ago (according to reading on reddit... if thats a source to go by)
π
yah
you do not happen to have the 'generated' key written down?
i was going to be like 'meh'... but now this watermark is showing up on top of games now... so its like ... ok, i need to do something lol .... (what that something is, well... teehee)
you could try the media creation tool that takes windows key and downloads correct win10 version.
i have the digital 'proof' key, but ive not run across anything that lets me input that anywhere during the processes
and that system does not even works correctly btw
it very consistently failed with builds I touched
if you have the key itself you should be able to enter in windows activation menu
the original key is win7, and ms isnt accepting it anymore it appears
though it will likely give you a random error like ms know to do it π€‘
yup π its a nice 0x37403974092730293290 error
Oh no, that's an extra month or two
If it ever gets don't, that is
well hm
i do have another key in my notes... labeled "Windows 8.1" .... i tried entering it, and it immediately took it but then said "Upgrade windows?" ... but im already on windows11 ... so i hit 'cancel'. no idea what that key is gonna do LOL
dont want it trying to install windows 8.1 lol
Probably Pro key and you are on Home Windows.
ah, yeah i dont want pro ....
i thought pro includes a bunch of junk that starts getting in the way of gaming and simple use
like "use windows home, not pro" for gaming rigs
No.
Just additional features that you need to configure for them to do anything.
Just not NEEDED for personal computer for most part.
Group Policy Editor etc. are nice.
HyperV
HydroX
It includes vm support & remote desktop out of the box as well
hm
more features to disable ... i see π
whats the worst that can happen, i have to clean wipe start over?
Not really?
As they are all opt-in.
Allow you to configure stuff that is forced on Home mostly.
k
https://en.wikipedia.org/wiki/Windows_10_editions#Comparison_chart
Windows 11 is close.
None of the extra stuff is on by default AFAIK.
google seems to confirm that theres no resource drain for pro ...
lets see what this key does
Well, running the OS inside HyperV is default on both, just on Home you cannot run additional VMs too, like you can on Pro.
π€ i cant seem to find my hologram box with this key on it, so now im wondering where tf this key came from in my notes....... ugh.
i found my win7 box ...
moving on ...
cable management ....i need to fix my mess i have for my macbook ...
its just ... bleah all over my desk ...
also need to get that old viewsonic on a VESA mount so i can get the stupid stand off my desk too lol
M1/M2 MacBook π
its just the M1
but one with MAGLOCK ...
really wish that anker hub came with a longer cable =p
and i need to find a NEW ssd enclosure that comes with usb-c so i can ditch the convertor
my 'tech shelf' is overflowing π¦
how do you guys get rid of old tech ... ebay?? generally ok experience?
Apple OS sucks more and more. Just now my mouse just dissapeared. Had to cmd + Q everything to get the cursor back π
Disagree but ok
Yeah, I guess people who use OS for long periods (a decade) see the changes.
well yea true I've been only on it for like 7-8 years
bring it to a recycle park, can't find the english name. We have special places we can bring tech too, they get out components that can be reused etc.
All the metals they can salvage they try to melt too and recycle.
An urban mine is the stockpile of rare metals in the discarded waste electrical and electronic equipment (WEEE) of a society. Urban mining is the process of recovering these rare metals through mechanical and chemical treatments. In 1997, recycled gold accounted for approximately 20% of the 2700 tons of gold supplied to the market.The name was ...
somehow im not sure muricans have it ?
Pretty sure you have it. Just don't know it.
too much junk!! AGGG ....
π€·ββοΈ
you can reduce that to like 1/4 by removing the boxes lmao
we have recycling hubs.... where they take recyling materials...
boxes keep it clean and organized.
without boxes, i wouldnt be able to stack much of anything! LOL
each of those motherboards still have a cpu and ram stuck on them .. heh
er, except the z87a ... thats being used as the htpc right now
then everything gets more and more dusty over time π
Cabinet
how shitty are your cabinet doors
don't cabinet door shame
bro buys them made by tesla
Regular cabinet doors tf you mean.
okay explain to me how dust gets in then
No clue but it seems to get in anyway
yes if you have like 1cm gaps
I mean there aren't gaps. I guess the cable routing holes in the cabinet is letting dust ib
πΏ
heheh
bro got a dust spawner in his cabinet
unless the cabinent doors come with gaskets... shit inside gets dusty
Exactly
but very very slow
well, you cannot prevent dust. You can reduce it.
I donr have airtight cabinets it's still gonna get in anyway
and we also all know that boxes don't get dusty
heck, ive found dust INSIDE some boxes since they didnt tighly close all around the edges π
and then yeah, the layer of dust on the outside of the box
but if they are all stacked... only the TOP boxes get dusty π
I try to vacuum clean every 2-3 days. Ideal, I want a robot. But those will get stuck atm.
get one of those air purifiers
you think they work efficiently?
it's just a fan and a filter if you're concerned about power
No, I am worried about it consuming power.
and after a month, collect 5 dust particles π
but for dust, if properly specced you'll never have to dust ever again
(or only like once a year)
Ah, I should look into medical ones. They should be quality. And more QA.
Boxes are fine, provided the board fits inside with the stock headsink installed.
I use one to filter smoke/dust in the summer and it's quite effective for a smallish room
And the filters collect a lot f visible dust
Oh interesting. I will need to look into it. I would be happy if it could just collect dust around my PC area π
I always sleep with my windows open at night, so I assume dust comes in with it.
Wtf π
only one of my boards fits like this... but they stopped providing stock heatsinks now on newer intels i have π¦
gotta love windows...... (with a note on the screen "do not power down your computer until finished")
this is the moment where you pray your ups will last long enough
Another reason why I like to have control on when to update and install. If you have pro you can disable auto updates.
Yeah
My updates take 5min tops
I donr have a UPS tho
ups is honestly a must
I am buying one soon
eh, just don't update when there is a thunderstorm?
especially if you live in a rural area with frequent power flashes
I have a surge protector
still nice even if you're just playing a game
well yeah, depending on how frequently you have powercuts.
i have a ups on my pc and wifi router, so if the power goes out for a couple seconds i can keep playing like nothing happened
I also live in a village with pretty much stable power
I'm more scared of a bios update going bad
At any time
personally it seems a bit of overkill to me. But I don't know, it does make sense to get one if you got the money and room for it.
you can get a decent one for 40usd, and they aren't too big
theres a total of 3 in my house
oh, that's not much
though that's cause we have pretty consistent power outages and flashes
I probably will get one with the next build
oof that sucks π
As it's a much more expensive build
little bit of peace of mind never hurts
true
but like for a gaming PC + NAS and a long cut out, one would need a good UPS no?
my ups with a ps4 next to it
Like how long would they last.
with my pc and one monitor i can get maybe 10-15 minutes on the ups
though that's with the pc idling
if it's under load that's a whole different story
it also has a built in surge protector
The typical expectation is that they last just long enough for you to save work etc & shut down safely
If you want something that allows you to work for an hour or more, you'd be looking more in the 200usd range
or a generator π
the exact one i have is 80usd
True, just a pain for mantineance & fueling
(i didn't pick it or pay for it)
Yeah, I agree.
And also not permitted in some areas
yeah, ok. Now a UPS makes a whole lot more sense.
Yeah
1kwh battery lol
I've really considered it since I've lost probably several hundred dollars worth of time/work from power outages
would be good for my NAS tbh.
600VA / 330W Battery Power
23 Minutes Runtime at 100 watts
5 Battery Backup Plugs
2 Surge-Only Plugs
1 # USB Charging Ports
specs for the exact one i have
So you could get nearly half an hour of runtime at idle with reasonable power consumption from the monitors
but like my PC can draw 750W, if I am gaming and there is a power cut, the UPS needs to handle it no?
that's plenty of time to save what you're doing and shut it down properly
UPS wouldn't handle my next build wile gaming for very long
Worst case scenario, my PC shuts down I guess
Yeah
But even at 750w, that ups could theoretically last for 5-6 minutes
Long enough to game through a short power cut or to save & quit the game
Unfortunately, most of them don't have more than a 500w inverter unless you pay for a more expensive model
also not the best value ups
800w of power draw plus monitors and perif.
Oof.
Tho that's projected full load power
Yeah, I should check at my average consumption while gaming.
yeah, I was thinking worst case scenario. (theoretically)
2400W π
im topping around 520 w as measured between PC and Wall
not sure how much my NAS consumes
SHOULD be low.... like 10w low
my pc idles (at boost clock) around 80W ... since I cant get it to drop its clocks down consistently due to that oddball background idle activity on cores5/6
its consistent whatever it is! and its ALWAYS core 5 and 6...
thatβs legit bigger than my planned pc build
Seehs
@night girder a 1000w UPS would do fine though, as if you have a power cut you think will last more than 10 minutes, you can simply quit the game and power your PC down... then it would run a lot longer
without my displays running and pc.... my router/modem/switches will run off my UPS for ... uhh it says 48 minutes
Turbo Boost Max 3.0 speed applies to two favored cores. Itβs only achievable if the CPU is running below its power, current and temperature specifications.
So maybe those two cores are just working on whateverf your PC is doing?
π€· its different behavior than my 12600k ... which had p-cores and e-cores
Yeah, probably not it.
on the 12600, i would see SOME idle activity only when i would leave the pc alone and the display slept (like windows saw i wasnt using the pc, so it started indexing)... but that would go away when i bump the mouse, all cores would flatline since nothing is happening then
this behavior, this is like ... constant. always.... no matter what, and always those two cores
π₯Έ
I HAS MALWARE! thats it... has to be it .... ok, time to boot off my 'clean' win11 install again and see if i can figure it out
do a safe boot
how do i do that again?
it was *shift * during boot right?
ah yes:
On the Windows sign-in screen, press and hold the Shift key while you select Power > Restart .
its still doing it as well on my total clean windows 11 test ssd
so not software
well a clean install and still same behavior? + safe boot.
hm
does your bios have a way to check cpu usage?
not that i see
just overall temp currently while in bios
interesting, so this 'fresh install' of win11... decided to do its 'behind the scenes house keeping' since i have not touched the mouse/kb for about 15 minutes... im seeing that old normal 'weird activity' when the computer should* be idle. bumped the mouse and that odd activity ceased instantly.... BUT, that new always-going-core-5/6-activity remains
so the new thing, is not that typical win11 housekeeping .... interesting ... wtf is this π
check CPU with other mobo?
I mean, the system is stable so I wouldn't worry immediately
i dont know if the cpu would even work on the z690
and yeah, its stable ... its just got this WEIRD idle activity
wierd "new" idle activity that i never saw with the 12600k's p and e cores
im betting its some glitch and a bios update in a month or two will fix it
or even a windows11 update in a month or two will stop doing whatever the fuck its doing with scheduling π
Bro an autistic member of lapsus got sentenced to LIFE for leaking gta 6 clips. Until a doctor clears him which will be difficult due to his condition
Getting life sentence for leaking gta 6 clips is actually insane
rockstar aint messin around
seriously ?
The court heard that Kurtaj had been violent while in custody with dozens of reports of injury or property damage.
There is a bit more context to it.
He will remain at a secure hospital for life unless doctors deem him no longer a danger.
I mean there is hope. If the doctors are able to work with him and calm him down.
not in a prison hospital
do people not understand autism?
they are actually worse than a prison
I mean, it sucks. But so does getting hacked.
sometimes i think... boy it sure would be nice to be stuck in solitary for months ... not dealing with people, or work, or life...
pleading mental health is like the worst thing you can do
hacked Rockstar for 5 million dollar damage*, violent with dozens of reports, ...
I am also not surprised he got this sentence.
he wasn't even able to get represented in court
why
because he has a mental health crisis in custody?
or because he caused MONETARY damage to a million dollar company
cause the court is rational and cold. They put him in a hospital in the hopes to work on the mental health issues. If he is no longer a danger for society, he gets a trial.
if the hospital fails, that's not the courts problem, from their perspective, they did their job.
Is this a fair system? I don't think so. But I am not surprised.
it's not a hospital. Prison hospital and actual hospital are two very different things
Still, he isn't fully innocent either.
I've never been in an UK secure hospital
life for an 18 year old for monetary damage is insane
with a chance to get released
although I highly doubt it. Even if he clears the hospital. He still gotta go to trial I think?
there is no word on if they do the trail again. The original trial is done and he got sentenced
note to self: never hack rockstar
or anyone without permission
depends
Noone except shareholders got hurt. Life is crazy
write an email
I've already written another email today
do they have a support portal?
I assume you already went through the process with the store you bought it from?
Asrock us has a support portal and I had some help from them but they still need the Canadian division to talk.
Newegg has no other help other then give me asrocks support info
welp time to wait then
Talk to your provincial Dept. of Justice.
I got a check from Lenovo for filling out a web form with my State DOJ.
I mean it's been two weeks without a word from asrock canada
Asrock USA responds but Canada. Nope
Even on open office hours
Yes, contact your government, and tell them your warranty isn't being honored if that's the case. That got Lenovo's ass in gear for me.
Shit aight.
As this is attempt 2 to get warranty from asrock
So yeah I'll prolly do that here next week if it still hasent had any response from asrock canada
In my case, I wrote that I felt like Lenovo was trying to run out the clock on my 1 year warranty, to not provide meaningful service. Lenovo cut me a check for the cost of the 4 year extended warranty.
Mabey asrock is trying to do the same. Like I only have warranty till January iirc
I actually like this laptop so I did opt to spend the money on the 4 year warranty. because that turns into a 6 year warranty with my Citi Rewards+
That or I have 1 year I think it's 1 year
1 year is law on new products pretty much everywhere.
Bur it makes no sence why it's been several emails and now a phone call and still nothing from asrock canada
SO it won't be less than 1 year
If the warrany is this hard to get I won't be buying asrock again
And I kinda like their stuff
Ultimately all customer service is bad until you find a way to penetrate the first layers.
As much of a PITA as it was at first, Lenovo ultimately sent a tech to replace my laptop's mainboard and cut me that check. And now I have a history of issues, so that's hopefully a few skips in the line the next time something goes wrong, including them possibly just sending a newer better machine.
The mobile 5900HX and 3080 are holding up fine though, really what I want is just stability.
stinky weasel beams !
I have a $20 gift card for Newegg, but I just can't find anything I want. Anything that starts to tempt me, I'd pay just as much overall on Amazon due to the price difference, and often for a better product too.
now that's optimistic π
for all of three seconds i thought you were running it at that speed
those were a beautiful three seconds
anyway, even expo 1 doesn't boot... so I'll try manual settings
even at the rated 6000 the ram won't kick into gear. so maybe a bit more voltage or just relaxing the frequency/timings a bit will help
Watching YouTube for the first time in a while and in a 15-minute video there was three 20-second unskippable adds
the like/dislike ratio was a good indication of quality of how-to videos
fixed
Huh. Buy cheap get burned. Testing a noname brand (Lisen) USB-C cables
https://www.amazon.com/dp/B0CG1LGWR6
My friend got these, but in trying multiple battery banks and my laptop, they can't output much juice.
My Anker cables work fine.
Actually, hold up, this is getting weird. Those cables work some of the time. And either they all work, or none of them work.
But my Anker cables work 100% of the time.
We're just plugging and unplugging them over and over to test with the meter on my Anker hub.
The Anker cables shoot right to 91W~ when I plug them into my laptop, every time. The Lisen cables work only some of the time, and during the times they work, all of them work.
i bought a cheap no name right angle usbc cabe thats a meter long, says its good for usb3.2 and 8k video but it cant even do 20MB/s file transfer
https://a.co/d/3nUvKxR aint worth it
I still go cheap on all audio cables and adapters. At least there's no logic in play for those. USB-C cables need charging logic, so there's room for wonkiness I guess.
THese are the Anker cables, I got 6'
https://www.amazon.com/dp/B0CCXT4QZR
I just dont like buying anker
and I have a tb3 cable that i mainly use for everything, was just trying different cables to see why my external ssd enclosure wont pickup on my iphone
I ended up buying Anker incidentally here. I could find a better power strip USB hub combo. Then for cables I wanted 240W and anything above USB2 speeds. Only 5gbs, but I figure if I ever need a high data cable, I'll get one for that purpose.
I am looking at my portable touchscreen options, incidentally. It's probably best to just buy a whole Android tablet.
sadly
do these cables only do 5gbps?
Right
oh 2.0 only nvm I can read
Though I haven't tested them yet. The only way I would have to test whether they can actually do 5gbps is transferring stuff from my phone to my laptop
If anyone knows an NVMe enclosure they would recommend, I'm looking it upgrading my backup situation. This is another thing where it shouldn't be complicated, but this recent experience has made me paranoid.
I can reccomend the one I have, I can also give drive speeds from the ssd thats I have in it
Might also be too low VSOC or VDDP if you have those on manual and not auto.
Or MB might boost FCLK too high if it isn't manually set.
What you got? Bonus points if I can use that $20 Newegg gift card I have
should be the same as this https://a.co/d/cHB18r0
Turn your NVME SSD into a portable external hard drive! ORICO USB 3.1 Type C Gen 2 10Gbps NVMe M.2 SSD enclosure based on PCI-E NVME with UASP. When connecting it to USB 3.1 port, you can get a read/write speed 900+ MB/s. Specifications: Model: ORICO M2PV-C3 Color: Black Material: Aluminum Alloy+...
As continuation:
My current CPU was Core Cycler "stable" at:
-29, -23, -24, -20, -27, -29, -29, -26.
And it was all-core "stable" at:
-28, -19, -19, -14, -18, -25, -22, -17.
But still testing partial core loads and currently at:
-27, -18, -16, -09, -17, -23, -18, -12.
So very big differences on some of the cores.
And for the all-core stability, would only have needed to drop the CO of the worst core (8) to that -17.
But I was already testing partial core loads too at that point.
So if I had only run Core Cycler and all-core, I would currently be at:
-29, -23, -24, -20, -27, -29, -29, -17.
And would be complaining about random crashes in some programs.
Yeah makes sence
Worst core, as in the one asking for most voltage when loaded, even with that -17 CO.
Cause is that with more cores loaded, the voltage drops and spikes as load change are higher.
And as the AMD loadline is too tight, the core gets too low voltage when the load kicks in again, and the voltage drops down for a moment, if the CO is too aggressive.
But as all the cores get the same voltage that the worst currently active core asks, you don't notice that for the better cores on all-core.
When in this case the worst core asks for about 50mV more than the best core.
Getting all-core core-cycler would be so nice...
And even better would be configurable core one.
Now I have to run core-cycler on all cores.
Then add static load to best core, run core-cycler on rest of the cores.
Then add static load to two best cores, run core-cycler on rest of the cores.
And so on.
With different settings and cooling performance (because the voltage fed has temperature and frequency dependant steps).
I may do this
Tho I donr seem to have stability issues
Tho ther then Radeon drivers being fucked for Dx11 titles
Might need to disable DXNAVI for DX11.
nimez-dxswitch.pages.dev/NzDXSwitch
So you would want DX9 NAVI with Regular DX11.
And in this the best core isn't the core preference order, but the Core VID values in HWiNFO with "Snapshot CPU Polling" enabled in Main Settings.
When the cores are all-core loaded with the next test.
Which then shows the per core voltage requests, highest which is then used by the VRM.
All tests are done for this year. End me
I am confused. If the core asks more, shouldn't you increase the offset instead of decrease?
Negative offset.
Less is nearer to stock.
1.20V + 0.05 CO = 1.25V?
Ok, that is confusing lol.
So if stock was 1.00V, and CO was -10, and each CO step was 2mV, the result would be 0.980V.
The step size isn't fixed in mV.
So yeah, you giving the core less V when idle π And then when load it jumps to 1.00V?
At all points of the frequency-voltage curve.
Undervolting to make it use less power, less heat and so turbo higher.
And that makes it so hard to test, as the instability can be at any point of the 3d, per-core, curve.
As it also uses temperature as one of the inputs to the table.
When I read this, I would assume you would have added CO.
I can relate a little bit (my test was GW2 and let it run for hours).
That was just about the difference between cores.
The best core voltage wise is the core 0, with -27.
Worst core is the core 7, with -12 right now.
But the core 3 with -9, uses less voltage than the core 7 with -12, or the core 6 with -18.
The CO is offset, to whatever the AMD per core made table would otherwise set.
And that table is different per core, and per CPU.
but with this CO you got everything stable?
Don't know yet.
Still testing.
Did the AVX512+HT at 100% fan tests with Prime95 and Y-cruncher.
Now doing the same with SSE+HT.
1h per core per cores in use.
Do you think these tests are healthy for hardware? Constantly putting hardware under extreme pressure with tools does sound like a way to reduce hardware quality quicker than normal use.
No more or less as any other load.
I doubt it. More heat. Fans spinning more, so more rotations etc, less lifespan of fans π
Of the same level of usage and temperature
I am talking about: using tools daily to OC a computer vs just game with the computer (over long time periods).
Not talking occasionally testing your hardware.
Changing temperature is more damaging than constant.
No matter the load.
In my head I think: this hardware should have been QA for this.
On the other hand: it's tech companies 
This is OC stability testing.
Not normal.
Interesting.
Undervolting is OC.
Using Curve Optimizer in negative direction is undervolting.
I know.
but tools like: 3DMark, Y-cruncher, memtest86, they interact with hardware differently than normal computer use.
Just wondered if it had an impact on the hardware.
That if I run these things for days and days, the quality/lifespan of my PC would be the same as if I gamed for days.
Active silicon cares about voltage differences, temperature for the die degradation, and temperature difference/change related physical stresses.
Not the load itself, just the effects of the load to those.
Ok π
So the low fan speed "hot boxing" causes more degradation than normal, the 100% fan at all times causes somewhat more of that physical stress in some cases, but real usage would have more of those changes from the fan speed changes.
Same for the fans, running at same speed is better than changing speed, no matter the static speed when above very low ones.
As I will be running the same tests again with fixed 20% CPU fan speed after the 100% fan speed ones are done.
Any idea why Ubuntu might take about a minute running with black screen when turning off sometimes?
On the topic of fans: i have this weird deepcool fan that constantly rapidly speeds up and stops with soemthing like 3 second period
No, all amd
Is there anything of note in system logs?
journalctl -r -b -1
And does hitting esc do anything? 'silent' boot param might still be in effect at this stage, hiding the text output while shutting down.
I believe 90 seconds is the time it takes to wait for a stopped job, like a disk that can't be unmounted.
Thank god it still hasn't changed at all I love it
Sometimes. I don't have much space for linux system (my work quickly eats it, and i reserve some).
But generally i decided to separate gaming to be primarily on windows, in dual boot.
I think it might be something with didk now... Cause w8ndows recently told me it needs restart to repair driver π
But that thing on linux occurs rarely but for along time...
And yesterday i, for whatever reason, forced shutdown at that moment - and so it couldn't boot Ubuntu back. Surprisingly, windows fixed it with disk check.
I think maybe ntfs hdd that is mounted to both os is the source of rpoblem because beforenboth events i ran ntfsfix on it cause Ubuntu again, mounted it in ro mode
Ye, the kin dof thing i do works better on Linux (but on windows should be fine too, anaconda tries to be friendly). But omg, even on linux i run into "file ame too long" issue. I need some solution to browse rpoduced files i guess.
The maximum file name length will be based on the filesystem, not the OS. π
this channel really does make you feel like that as a windows user dosent it
Answering "maybe" to a "what do you think ..." question π
I'm finding that many tech spaces are being dominated with Linux discussions in the last few years.
"What do you think about the AMD?" ... "Maybe."
When given a boolean question, answer with 0.5.
yea in general linux has invaded the convos in the space more
It might be that Windows simplifications have driven people away, and Linux desktops are becoming ever better. Plus with Proton, switching becomes so simple.
just feels like the public view of it went from ha stoooopid to hey now... linux may not be that bad in what feels like barely 3 years
Anyway, i was 5rying to support windows as well with all its stupid limitations...
And then i hit limit on linux, which suprusingly includes the whole path, haha.
To me, switching from Windows 10 to Linux Mint felt like when you leave a noisy room full of chatter to a quiet empty space, and can just..relax.
Checked curves? DC fan? Checked fan manual?
AC fan? DC fan? ACDC fan?
Yeah, at the time, DRG with protin (it wasn't natuve support by the game at the time) ran better thatn just windows
Iirc Valheim does that too. Though native vs native.
You have (or at least had) to use workarounds to get it to perform on Windows due to usual Windows quirks.
Funny thing is that it doesn't care about curve. So even in flat line config it jumps 0-100-0
Idk if it even has manual... Simple fan with 3 pins
Motherboard can opertae with it in dc, pwm and xauto" modes
modern pmw are 4 pins. So it sounds like you got DC/AC fan
Btw, is opengl rip?
Ok, i need to check once i get back. But strabgely it's only 1 fan like this in a bunch of 3
all 3 the same brand?
Probably that's ehy it get limited ups
Ok, that's weird. Is it connected to a hub? Check the connections to the motherboard to see if it's in the right slot.
Unity I think?
How do you mean?
Bought 3 of same model. But mb also connects to 4th other fan and cpu, ofcourse. I think i left "pump" port alone...
What motherboard you have?
I think there were tomes when gaming on linux required opengl, and now the only thing i can remember that runs on opengl is TF2. Vulkan is the new thing that replaced it?..
Msi b450 max, or something close to that
In time we might see OpenGL drivers dropped from modern driver stacks, but we have projects like Zink which implement OpenGL on Vulkan.
And they're getting pretty good at it already.
Directly to mb. What do you "right" slot? I see fan's actual speed graph in UEFI
Oh, right. Some use opengl for ml and stuff
When you plug a 3-pin (Non-PWM) fan to a fan connector in PWM mode, the fan speed will always maintain at 100%
From the manual.
That mode is setting in the BIOS.
You can switch between PWM mode and DC mode and adjust fan speed in BIOS > HARDWARE MONITOR
Critical knowledge unlocked: i remember it was consistent before i tried to tune it up a little (winter...)
Idk why i gave up making it stable - perhaps hoped it would get better after reboot. But sometimes i still notice its noise of 0-100-0 loop
But that stopping and starting would suggest that the voltage is either too low (needs to be in DC mode with curve), or that the fan is broken.
Or the specific output is broken.
I would like that to be so... Its a pretty silent fan anyway. But it jumps π¦
You could try different connector. To make sure it's not the connector.
Because like said, 3-pin fan on 4-pin header on PWM control mode would run the fan at 100% with no control.
Ye, ye. Thatcs where i was adjusting its speed (flatline) and seeing the graph of its speed.
Test one of the working fans on the connector that the jumping fan on, and the jumping fan on connector a working fan was on.
Like Letter said.
To see if it is the MB output or the fan.
I need to check if i used pump port...
What if i screwed it up with changing the level of desired speed (voltage) in uefi?..
Then you just change it again?
Priced?
Only way the MB control can cause the behavior, is if the minimum voltage is too low.
I remeber getting anxiety last time i put effort into making it work as desired, and than giving up, levaing in just "stable whatever" mode π
Minimum for the fan, or you mean md giving too few power?
Why so many... Anyway i use 4.
Got to invest into LVM mirroring
The curve going too low.
Which lowers the voltage in DC fan mode.
And each fan has minimum voltage they can run on.
Going below 50% for DC is often bad idea.
Thatcs fun part - at max it didn't work afair. So i guess it worked at lower levels... Its weird, it needs invwstigation
Cause they can overprice off that? π
4 tb - so just 500MB? Huh
Czech is good
Its a fine country with good people
Hah, i remembered dr Zelenka from sg atlantis
My mother's got a Czech name due to her parents having good friends they visited in the country in the Soviet era. :3
Poor Zelenka.
Bleh..he never got enough praise or attention but he was just as brilliant.
Maybe not AS, but still damn good. Better than dr Lee for sure
And how much he actually cared for Rodney too, even when the guy was so abnoxious.
I never got a chance to understand anything he said in Czech, despite being familiar with couple eastern languages
Definitely an underrated character.
Yes, that's definitely where he outshines him
still got that shitty name I got from you
My partitions have such interesting names.
"root"
"home"
"games"
"virtual machines"
π
Do you talk about Matrix on Matrix?
just a .txt where he talks to himself
My only partition is HDD - its named DataHDD and mounted inside home. So i can work with it.
Everthing else is LVM
I don't much like LVM for some reason or another. I have fixed partitions and some free space.
Lvm is a decent way of mirroring data across physical dribes
I like Espanso. I have it set up to send neofetch in a codeblock if I type ::neofetch
OS: Linux Mint 21.2 x86_64
Kernel: 6.6.2-x64v4-xanmod1
Uptime: 8 hours, 47 mins
Packages: 4334 (dpkg), 34 (flatpak)
Shell: fish 3.3.1
Resolution: 2560x1440, 2560x1440
DE: Cinnamon 5.8.4
WM: Mutter (Muffin)
WM Theme: Adapta-Nokto (Mint-Y-Legacy-Dark)
Theme: Mint-L-Dark [GTK2/3]
Icons: Tela-red-dark [GTK2/3]
Terminal: espanso
CPU: AMD Ryzen 7 7800X3D (16) @ 5.050GHz
GPU: AMD ATI 03:00.0 Device 744c
GPU: AMD ATI 0d:00.0 Device 164e
Memory: 18683MiB / 63444MiB
And because lvm works with luks, i use both
Huh..it thinks Espanso is my terminal. Funny. It must get that from the parent process..
And mm yes, everyone knows exactly what AMD ATI device 164e is!
Idk why, but isaw dual monitors...
Big great topic: why dual monitor when you can go with one big?
my neofetch is boring aaah hell
I hope its not vega
I have window snapping disabled but want my leftmost monitor to be just a web-browser.
organisation
having one screen for the main stuff and a second for small stuff makes it way easier to work with it imo
RDNA2 I think? It's the integrated graphics on the Zen4 CPUs.
(and your work laptop is also open as a third screen for teams)
RDNA2 for the desktop ones so far, some of the laptop parts and the coming 8000G:s for desktop are RDNA3.
It transcodes as well as my RDNA3 card though, via ffmpeg.
For h264 that is. So I guess that part of the video engine wasn't upgraded.
Raphael has VCN 3.1.2.
The Navi 2x:s have VCN 3.0.0.
The Unified Video Decoder / Video Core Next is separate bit from the actual GPU arch.
So I've been passing my integrated graphics into ffmpeg to transcode while gaming. :3
Yeah that's roughly what I thought.
But Raphael doesn't have hardware AV1 encode, that the Navi 3x and Phoenix with RDNA3 have.
I'm hoping by next summer I can even start using AV1 encoding on my card.
I don't have the packages for it.
Like... What?
^^
^^
also one horizontal and one vertical goes hard
Or can fit everthing on one big screen like for example, open 4 files at once and be fine with it
Been there. Single 32" 4k i liked way more
so you watch yt and have discord open on the same screen?
and maybe also a game
that sounds like it sucks tbh
even at work where we use one big ass ultrawide I still have my laptop open for teams
I can do that with splitting screen in 2 or 3 even parts.
ON A 32"?????
I run two 27s and still don't have enough space sometimes
brother the one screen times are over. get real
One word doc taking up half of each screen, email client taking up the other half of the primary and a browser on the other half of the secondary
Everyone switching to 4k monitors
I can't even use my 1440p 27" without my glasses on anymore π¦
I just found the extra screen to be a fancy uthidentity thing that in reality is better replaced with single big screen that helps to focus on doing work with a much bigger space (both physically nad in terms of resolution),
While all the extra stuff is better left outside of view. When i need it - i switch to it. All the extra-urgent managers can go suck my ass. When i work i need to focus - i don't need 6 apps blipping in my face - i need few code tabs nad maybe a browser, and i can use alttab just fine.
Display scaling, mate
get real
My point is more.. why have 4k if I can't actually see that level of detail any more?
Been there, and maybe that's the only actual use case where i sometimes would rpefer second screen. But anyway, its not like you need or can interact with bith in the same time.
I agree with that... no point in having 4k on anything smaller than a 32
But 1080 would also just be dumb at that size
If I take off my glasses, my 1440p turns to 1080p and I can't read it properly π
Well, 1440p on 32" might not feel as good. But it's a viable option.
Agreed
But i put 125% scale on 4k resolution and feel like i have to guess some text sometimes.
Text readability would be awful
No desk space at this point. I'm just going to..wear my glasses.
But even with the power of glasses, 4k is probably out of my reach.
And for some it wont be a difference
It's not terrible, similar PPI to 1080p 24''
Big desk is a must just to be far enough from the monitor
And being further away means less resolution to your eyes..and we're back where we started.
I mean, i can see some discomfort when 4k is in 200%
When i lean in my eggchair it get quite small. But you can still guess pretty good
Wouldn't that be more similar to 1080 @ 27?
Ah, so somewhere in-between
Anything over 100ppi I don't mind, but under that it's noticeable
24'' 1080p is 92PPI
32'' 1440p is 93PPI
It's the same pixel density
Ah
My phone would make a good display then lol
436ppi
buried in work before the holiday weekend ...
so much to do, so little time left to do it
Mines at 425
Phone screens look nice
It's crazy high
No matter how close I put it to my face I can't make our individual pixels
Yeah
Tho random topic. I may buy a tuner ecu chip for the ranger. From Ford. Good for another 45bhp on 91 octane. (I'll still be using 87) but I'll use it to improve transmission prefomance
So it's slower to shift up to 10th
Giving more responisvity
As the truck dislikes shifting down when possible
arg FRICKEN CHROME .... im inching more and more towards firefox
i used to use firefucks .... till it went to shit, and i moved to chrome... the chrome was awesome for a long time and i never looked back at firefaux ...
Beware if your transmission
Those 10 speeds are pretty beefy, but if you're hoggin on it with the extra power that equates to more wear on the brakes and clutches, so make sure you're keeping up on mantineance
but now, its grating on me, more and more, with googles BS ...
Oh ofc. It's not a ton of extra power it's basically making it shift less. I'd only see the power boost if I used premium fuel
Yeah, moving the shift points just improves driveability at the cost of fuel economy and a little extra engine wear
I'd probably have to manually tune it myself to keep running on 87 and have the slower transmission shifts as the gears are decent when I shift them in the sport mode.
The truck preforms good as is ill wait till warranty is gone before I do that
Turns a ranger into a Honda civic with a laptop
Actually I can get the tune put in
I just have to go to a certified mechanic to do it
manifest v3 coming π
they done the classical "we can legally defend ourself killing the current implementation by showing the next one will be soon functional and documented .. some day in the dreamed future"
question is nvidia studio better for doing 3d work on a pc?
yah, poor ublock
its the ONLY extension i run ... ublock ....
maxeek macrodosing lsd rn
i was looking at my history on 3dmark, and saw this on the bottom and thought it was kind of hilarious ......
oh wait, different kind of test
was gonna say, time spy is very different than 3dmark 11 lol
So what was your 4080 score in timespy?
just 4080 or ti?
theres a 4080 ti ? π
4070ti is decent
i think ? maybe gimped vram
wasnt the 4070ti originally the 4080 12gb ?
and 2k more dollars
ye
Find someone who wants their own top 10 spot and ask for a new 4090 shipped to you for your golden 7900 XT
i could oc this 4080 because its barely heating up and hasnt hit any of its limiters ...
and considering a 4080 is about 1.2k⬠and my 7900xt was only 800 I'm good and happy:
i couldnt touch my 3080 because it was already hitting thermal limit
again it's only compared to my system. in the 7900xt general range I'm in the like top 20% since my cpu is limiting me hard
although getting a cooler gpu, i traded the heat with a boiling intel =p
SOmeone out there with a 5800 X3D might be bonkers enough
now...... if i did a custom loop and liquid cooled my cpu ..........
prolly something false to make gamers happy
It's not that average isn't false... It's just... Misleading regarding thermal throttling
need to have a look at it with hwmonitor
(keep in mind that having HW monitor open during 3dmark might "invalidate" your result)
3d mark doesn't like other software interfering with sensors
it seems to be ok with msi afterburner, as ive always had that running and its not complained
(the graphing part of it with RTSS)
now, ive not seen the "0/1" flip for thermal limit... so, least theres that
I said it before, but herg tactfull ignored me. If you look at top 1 score. They have lower temps and more OC. I think it's open bench or something.
So you don't even need all that custom stuff to get a decent 3Dmark score.
liquid metal instead of thermal paste would also do the trick
also my GPU has a better score but the CPU is way way lower
since I haven't OC'd it at all except a dumb ass -20 CO
stop smoking
bro vaping and thinks he sees a ghost
If you want SF help you gotta go to #1038092680493801533
nvm the monitor is good
