I don't think he's still around even, after TS went awry they all seemingly went their separate ways. I do know Qain is still pretty active (and doing some pretty neat tutorials and other solo stuff), same with Wendell.
What parts of their testing methodology and results don't hold up?
Ever wanted to strangle Microsoft? Now Outlook, Skype 'throttle'..
Microsoft McFucked up their Azure backend
For temperature testing, I think they have one of the best setups (aside from niche water-cooling websites), but for performance testing they haven't really even released shit in a long time that I can remember.
I do remember recently though, they had a video about safe VRM voltages potentially "killing" Ryzen CPUs on shoddy boards, and they concluded that via some probe testing, but later somebody went in and tried to replicate it using better testing points (such as not the ground pin on a totally separate rail...) And found it to not be an issue. GN could've avoided this by taking more time to research the topic and not just "do it live" like they so much prefer.
To be fair, it's been quite a while since anything was really released.
Don't know if it was posted about here, but the MEGA chrome extension got hacked and may have lifted logins for google/facebook/github/crypto accounts if it auto updated and you accepted the new permissions in the 4 hours the hacked copy was live
Just looked at specs for my Oneplus X (that i used for a few months and broke the bottom 2nd row of the keyboard touch screen) and my Samsung A3 2016, that works almost properly (used battery, camera issues)
The X has 3GB RAM and better other specs, but it has Android 6 and it is weak by now. The A3 with 1.5GB of RAM has been treating me a lot better, even though it is weaker but it does run Android 7.
Weird how software makes a big difference. It is even bloated with Touchwiz but the newer ones arent as bad as they were during S2 ages.
I'm considering building my own external 5.25" enclosure. I want to move to a new case which doesn't have 5.25" bays, but I still want to use my 8x2.5" hotswap cage for my SSD's.
My concern primarily now is powering the unit. They'd only need 5v and 12v, but I'm concerned with using cheap chinese adapters to power these quite nice SSDs. So then I'm thinking of just using a tiny SFX powersupply for it since it will have more controlled voltage and other protections. Or does anyone have any suggestions for an option?
My plan is to build it out of aluminum sheets perhaps and then use a 8088 to 8077 adapter to make it run off external miniSAS.
Keep us updated on what you do, I'm in a similar situation with one of my 5.25" drives.
Or this looks good with some protections in place:
https://www.amazon.com/MEAN-WELL-RD-65A-Output-Enclosed/dp/B005T7EZM8/ref=sr_1_4?ie=UTF8&qid=1536220106&sr=8-4
12v & 5v at the same time. Actually smaller than the SF corsair PSU I was looking at. Fanless, but I'd plan on sealing the edges well of the case and then using maybe a 120mm fan out the top.
Hey @Crest , still have your stash of Windows keys?
Yes. If this site has a DM you message me or message me on Twitter or something and I can provide an, amazingly, still unused Windows 7 Pro key.
Twitter DM it is!
Those kinds of power supplies do the job, but you get what you pay for and efficiency and build quality is less than a PC power supply
I've never had one fail out on me... yet... but for max reliability with 12V and 5V, just using a SFF power supply costs more and gets you more.
hacking a server PSU is also an option. You can get a 900w dell poweredge one for ~£25
Just successfully taught my first person how to build a computer REMOTELY from over 1000 miles away, using Snapchat.
I feel accomplished as fuck!
Just don't get to the point you're seriously considering buying a Mnpctech ruler.
Lol, saw a friend of mine on a forum. I'm like, "is that fucking vbulletin?"
it was 3.7.x
yikes
AMD's New $55 Athlon 200GE Desktop CPU Combines Zen Cores
I've decided to build a nas and this will be the basis of it.
Building is the easy part, troubleshooting is a whole new ballgame that you'll get very sick of quickly. Either way though, congrats
Yea it's the regulation and safety I'm concerned with. I have some enterprise SSD's that will be in this, want something safe. If the SF powersupply will work without issue I'll just get that.
Yea I won't be using much power. But perhaps a small server PSU is an option if it's very quiet with at most 50watts load.
Meanwell supplies are actually pretty nice for what you pay for. I've got several of them powering my radios and there's no real noticeable EMI from them. Their price to wattage ratio is way better than a similar TDK supply and they're almost as nice.
Do you know one thing I hate about working with offshore resources?
Having to quote the company's fucking coding standards so that they don't do stupid hacky shit.
Yea definitely not concerned with the PSU heating up, by quiet I mean if it has a temperature controlled PWM fan. Older server PSU's I have run a 6k RPM fan nonstop, so I just can't go too old if I were to use that. But this 5.25" dock has dual 40mm fans. I'd like to remove those. Seal the box very well and then set it up with some good 120mm fans to pull air through the unit.
Just looking for some more opinions. Meanwell and the Corsair SF PSU are similar in foot print. SF PSU would be easier to wire since I'd just connect the SATA power and be done with it.
https://i.imgur.com/L2NMMt1.mp4
Oh so its one of those days
Any Ivy Bridge overclockers in here? I figure I'm keeping my 3570k for another year or so before upgrading to a ryzen 3 or something, so I might as well squeeze everything it's got out of it till then. Here's where I'm at currently:
https://files.facepunch.com/forum/upload/209687/12a47dac-5210-4c07-8b76-4aa80457e84c/oc44.png
I'm setting voltage with a dynamic VID offset and it seems I've gotten it stable under a prime 95 load, but for some reason I'm getting the occasional crashes and WHEA errors in lighter loads like web browsing. Also noticed that hotter, more power hungry loads (p95 small FFTs) end up getting a lower vcore than slightly colder ones (like 20mV higher in large FFT test). I guess that means it's about time to turn on LLC and lower the voltage offset accordingly? Or maybe it would be simpler to see how far I can OC with a static voltage + LLC and then try to get dynamic voltage working afterwards?
What about a laptop power adapter with a pocopsu?
The only thing I've changed is set LLC from auto to medium, and it caused the voltages to drop by ~30mV by itself 🤔
https://files.facepunch.com/forum/upload/209687/acf3038b-eda2-4e43-9aaf-322feee35b69/44_wmedllc.png
It's running a touch cooler now and just as stable as before (ok in prime, occasional instabilities and errors once it clocks down). I'm starting to think that this arcane dynamic fuckery that I don't fully understand ain't worth it. I'm just gonna try a static voltage instead.
Sorry, you need to Log In to post a reply to this thread.