Etern4l

Yep, the HX1500i uses the C19 power cord as well: https://www.eteknix.com/corsair-hx1500i-platinum-power-supply/3/ Will be interesting to see how this works out for you. First time I've seen this, probably mostly used in America. Unpleasant surprise really, as totally unnecessary for 240V. Bro, 500km in 2 weeks? How does that delivery work? Do they have to use sleigh and reindeer? Edit: Bro @Papusan, looks like your DarkPower Pro 1500 also uses this C19 power cord. Any play, or does the cord sit snugly in the PSU? https://www.kitguru.net/components/power-supplies/zardon/be-quiet-dark-power-pro-12-1500w-titanium-review/3/

So you didn't see this on the HXi? I'd assume not, since that's clearly ridiculous. In contrast, the power cord in my Thermaltake is snug, zero play - plus there is a LED to indicate when the power is flipped on. Edit: thanks for the link. Great to see a complete fail with a non-standard C19 power cord. Obviously that's just needed for the 110V markets. The supplied power cord is 18A or 4.3kW at 240V lol

It's a pathetic joke, but Jensen is just a funny guy, which why he decided to pull another one on people, this time with a 3040 or 3060 Li pushed as "3060". I mean clearly those Nvidia people don't care about the consumer market anymore.

Just imagine the plug moving very easily side to side. Not falling out but really loose. Amazon talked me into trying a free replacement, have a bit of a bad feeling about this but OK. Will give them a chance, but one more strike and Corsair is out. I can't imagine it's working like this by design, but at the same time there is clearly either no QC or it really sucks, which is a worry.

Right off the bat, I'm not too impressed with one aspect of this AX1600i PSU. The power cord is quite loose. Not to the point where it would fall out, but just doesn't fit snugly, the plug very easily moves up and down and sideways when all the way in. It takes very little muscle power to pull it out. The unit seems fine otherwise. I'm wondering if this is a normal feature of Corsair PSUs or whether Amazon sent me a flawed unit/cable. On the other hand, the inner dimensions of the plug are noticeably smaller than the outer dimensions of the socket, so perhaps this is as per Corsair' s design? I saw this mentioned in an unboxing video. Unbelievable. Given all the recent stories of issues caused by loose 12VHPWR cables, I should probably send this back. The plug is non-standard, so don't have another one or try instead. @Papusan

Actually, Razer Core X also has a 650W PSU, and supports 3-slot wide cards, as opposed to the 2.5 slots supported by the Aktio Node Titan. https://www.razer.com/gaming-egpus/razer-core-x/RC21-01310200-R3U1#specs It does not come with an aesthetic carrying handle though. The Node Titan looks a bit like the G.E.C.K from the Fallout series, pretty cool in my book ;)

I had watched "Air Crash Investigation" / "Mayday" series a lot. I've found the production value, commentary, and the narrator to be quite good in comparison to other offerings in this grim genre. Turned out this didn't improve air travel comfort lol. My favourite episode has to be "Flying on Empty":

I do need a new PSU bro - the old one is 14 years old. Amazing reliability but I had to replace the fan already, now it started developing some weird vibrations, and I'm wondering whether it's causing those rare resets I'm trying to troubleshoot. This is used in mostly 24/7 operation, so I have been really pushing it with that old buddy, its retirement will be well deserved. The efficiency of the old PSU is somewhere in the 60-70% ballpark - that's how I'm losing money every day I continue using it. Of course, especially with moderate use. However, as mentioned earlier I had other reasons for purchasing this PSU, and my usage pattern is anything but moderate. About the same prices here, and we rely heavily on Norwegian electricity imports. So much so that should this source fail, there might be blackouts this winter. That's a well-functioning and resilient private energy market right there. Why invest in power plants and energy sufficiency, when you can just import from Norway or France? At least half of the world is suffering, including the Russian people. Hopefully the madness ends sooner rather than later. I think you misunderstood - I will be saving money thanks to buying a new PSU to replace a 60-70% efficient one. So much so, I'm not that price sensitive. I would have gone with the HX1500i in this scenario too. I paid 1.5x which is a bit more acceptable :) It's just that it will be used in a heavily-loaded system, so I felt things like amazing performance in hot environment were worth it. If getting the best super-critical component means paying $150 extra I'm OK with that in this crazy PC building business. It may be a little overkill but if it works reliably as advertised and improves system stability, I will be very happy. Well, one could imagine workstations being used to handle heavy workloads. Not a "normal"/average consumer use, not a normal PSU :) BTW I just wanted to say I appreciate your advice here and elsewhere Bro, it has been very helpful. Hope my extravagant purchasing decision is now at least little better justified in your eyes :) Now, can you please chastise bro @jaybee83 a little? He bought a Seasonic TX-PRIME 1600, that is even more expensive, and unfortunately no proper explanation has been provided yet ;)

For sure, I would get most benefits with the model you suggested, or something like Thermaltake TF-1. My point is that with the current electicity costs, an upgrade from the very inefficient old PSU will pay for itself in a couple of months anyway. The premium will cover the improved electrical and acoustic performance, including in extreme heat - which wouldn't matter to you, thanks to the use of GaN components. Fair bet electricity is cheaper in Norway, which affects the calculation. Again, the PSU was discounted, I paid just a little more than a BeQuiet DarkPower Pro 1500W would set me back right now :) @Mr. Fox Absolutely, too many times the promise of energy savings is basically a scam, often made easier by difficulty or impossibility of accurately calculating the expected savings. Luckily, in this case the overall calculation was trivial. Yes, I would have recovered the cost even faster with a cheaper sweet spot PSU, but with my heavy use and the ever more punitive electricity costs it doesn't make much of a difference in terms of the time-frames. BTW I've also started hearing political talk of penalising heavy electricity users in this country.... Ugh.

To be fair, MSI also has named profiles albeit something like 6 if memory serves. They can also be stored on a USB stick with arbitrary names, which effectively provides unlimited storage of profiles.

Nice. I guess Assus is not ideal for Linux. What's that application you used to adjust clocks?

Corsair managed to upsell me on a discounted AX1600i. I expect great things, the efficiency of my veteran Thermaltake 1200 looks to be around 60% lol. This will pay for itself through the electricity cost savings. @Papusan

Alright, we have another line! We're learning here! Elon Musk suspends Kanye West from Twitter for inciting violence 'Twitter's new boss Elon Musk was asked by one user to "fix Kanye".' All you need it pray to Elon himself and fix he shall. This is awesome, finally a greater power that listens to the people.

Thanks. I'm sure the main benefit comes from the OC/WC. So my windows score in Bedroom was basically dead on 3000 on stock, so your +10% due to OC sounds about right. My 12900KS CL indigo scores were bedroom 4142, supercar 9061: basically, correcting for OC: Bedroom +30%, supercar +10% on CL vs Windows. The 13900 bedroom score with +200Mhz on E-Cores was kind of insane: 5980 (+45% vs 12900KS on CL). The snappiness of this distro is palpable just during normal interaction with the GUI/window manager, and doing the same work on Windows 11 just feels very noticeably laggy. But yeah, the Nvidia drivers support situation is not pretty. The initial setup was actually bearable, but I've been unable to normally update the OS for several versions now (actually only tried like twice). Just keeping it as is on the "don't fix it if it ain't broken basis", but eventually I will need to update and that will probably involve some tinkering unless they fix whatever broke either X or the the Nvidia dynamic kernel modules. A number of updates prior worked fine, so there is hope. I think they are shooting themselves in the foot with this Nvidia thing. Nobody likes Nvidia, they are def the enemy of the people now, but they are the market leader in GPUs so they should just support the drivers for the sake of their users until Arc or something else takes over. CL adoption would go through the roof that way.

Wow, great work. Is this for a particular CPU type or CPU - agnostic?

Yeah, kind of as expected. I did look into SuSE, as this is one of the first distros I've ever used, but didn't find any benchmark evidence of outstanding performance. I presume those CPU scores are just a sum of the 3 component benches: junkshop, monster and classroom, if so then the total CPU score I was getting with stock 12900KS running something like 4800Mhz CL38 RAM was 493. I presume your scores were obtained on an extensively OC-ed system, if so the real gain to be realised is somewhere between 10 and 30% vs Win11 in blender, depending on how much RAM speed weighs into this, otherwise it's "just" ca. +10% here. In Indigobench the first scene score was over 4k. I remember this clearly, was a jaw-dropping result - over 30% faster than on Win11. The second scene was closer. GtG, will check tomorrow. (GPU scores are virtually identical BTW).

I've always said: 1GB of VRAM should be enough for everyone, but people wouldn't listen! But seriously, how did this even work???

That is some mighty achievement, outstanding! I would suggest 3DMark add a "Total frames rendered during the test" metric, as in this case it would be very helpful to know the exact number. ;)

For those who don't mind watching a good anti-ngreedia rant, it doesn't get much better on YT than this lol

To the point summary from a legal perspective: Probably one of the main points: taking funds from U.S. investors automatically placed them under the jurisdiction of applicable U.S. laws. Although evidently that legal framework wasn't robust enough to prevent the scam. Then again, Madoff operated right in the US.

Yeah, I'm not sure there was any other way that would not involve CZ having to write off his sizeable stake in FTX.

To be fair, 30% is daylight robbery.

Could an old PSU be a source of instability, under fairly constant load, running at about 60-70% capacity? I'm taking about once in 2 days crash events. Put another way, have people experienced improvement of stability after upgrading their PSUs? Seems like it: https://www.techpowerup.com/forums/threads/power-supply-causing-system-instability-what-are-the-cause-of-system-instability-aside-from-software-issue.250684/ https://forums.anandtech.com/threads/system-instability-psu.1270077/ Well, I guess it's time to upgrade, even though the particular timing is not ideal (the 12VHPWR mess, ATX 3.0 around the corner).

On this C-state thing: some Linux distros expose the familiar "high-performance, balanced,.." power settings that might help. I don't know how to set this via kernel parans etc. because I didn't have to look - the Clear Linux philisophy is io run everything at full speed. I'm sure you would love this distro from the performance perspective, but getting Nvidia to work properly and keep it that way across updates takes extra work. Other distros woth looking at from this perspective are RHEL variants: Alma, CentOS Stream, and possibly Rocky. Installing Nvidia drivers on those is trivial in comparison, it's supported. Nvidia explicitly supports Rocky. Windows obviously has vastly better CPU tweaking and monitoring tools, but in the end all you are doing is tinkering around the margins while certain Linux distros give you an outright 30% performance boost. I've seen a YT review (the one posted in the CL thread) where the guy claimed games run noticeably faster too. I wonder if CB would see a gain, but a first quick attempt to get that to work failed. BTW if you are interested in giving Clear Linux a try (it supports Plasma), I will walk you through the Nvidia setup, it's actually quite easy because someone wrote a script for this, although one or two extra config tweaks are needed to enable coolbits for instance. The first question is: is there an actual performance delta between SuSE and CL? Could you post blender and indigobench CPU benchmark results from your SuSE?

Correct core count showing in neofetch, but clock is 4200 no matter what lol I'm sure there is an explanation that likely involves the hybrid core architecture. Oh well. Edit: There is also this GUI tool called System Monitoring Center. Shows correct core frequencies, with per core graphs etc. the 4200 is the "max frequency" reported by something (not sure what), but this particular tool, like htop, ignores it. Would use it instead of htop if it displayed temps on the same screen as well.