Mr. Fox

Here's another one. Sad days and getting sadder by the hour.

Galax and Zotac and PNY have overclocking tools that are standalone products and not UWP feces, and they work with other brands. There is also value left in NVIDIA Inspector. So, there will still be alternatives. It is heart-wrenching that EVGA is no longer in the mix. Aside from their superior hardware, I actually used PX1 more than Afterburner. It is a better product. My use of Afterburner has been limited for a good while. Even when I was still benching laptops, I thought EVGA Precision was more effective. But, it is still very sad that MSI Afterburner is probably going away. It seems like NVIDIA doesn't care or doesn't want end users to have the ability to overclock. It probably makes it harder for them to distiguish new products as superior when there is not a big improvement. If you can overclock a X070 to make it look like a X080 or a X080 to make it run like a X090 it is harder to demonstrate value in the more expensive part. Putting things in a box maximizes profit potential for them. By limiting the current generation flagship they create extra space for the next generation flagship to look better than last generation's flagship product than it would if last generation's flagship could run wild and free. Smoke and mirrors, and masterful execution on the art of deception. They do cripple things. Lack of voltage control is the primary way they do it. They develop robust products with great potential and then limit the potential so they can offer different tiers of performance (for more money) without having to change much in terms of hardware. Just manipulate results using firmware. Easy money by screwing the end user. And, still better than anything AMD has to offer. NVIDIA Control Panel in Windows used to be like Linux with overclock controls. Now it is very limited and includes nothing for overclocking. Overclocking without voltage control and uncapped power limits offers very little opportunity for superior results. If your GPU silicon is poor quality, you have no voltage headroom. If it is better than average you can use up the excess voltage available in the firmware, but you are still severely limited in terms of how far you can go without more voltage. The root problem is a lack of integrity, absence of principle and generally unacceptable business practices in the computing space, and those that suffer most from the unethical behavior are the enthusiasts that are getting screwed harder than any other consumer demographic. 100%. The only way to effect change is by deliberately and maliciously inflicting severe financial harm, and that is hard to do when you're outnumbered by brain-dead sheeple that are as dumb as a box of rocks. These are the same morons that view themselves as being informed and enlightened because they read "news" headlines and believe they have all the facts. You know you're scraping the bottom of the barrel when the only reason NVIDIA is awesome is because AMD sucks. We know this is true because that's what the clever marketing materials tell us, and our "friends" on Facebook say so. That's no way to live.

People that do things like this, and worse, make me wish I had chosen being a vigilante as my career path. They don't deserve the air that they breathe and it would be great if they never got any more of it. I'd take this one pro bono, LOL.

Here's one to avoid...

That is an extreme example of an overpriced cheap product. There will never be a shortage on stupidity. The decimal point is off one digit too far to the right on that jokebox.

Yes I'm not into playing cheap stuff either. The cliche that you get what you pay for doesn't always apply to expensive items but it always applies to cheap items. And I don't mind spending extra for something awesome if it's actually going to be awesome. I've never really liked Gigabutt or their product offerings and I would absolutely approach the idea of buying something from them with some level of fear and trepidation. I've always been happy with Asus products until they failed and put me in a position where I had to deal with their product support nitwits and their crappy warranty service. I'm always quick to forgive mistakes, but I don't have any respect for companies that are dishonest and look for opportunities to shirk their contractual obligations. I think Asus is that kind of sorry outfit. I have no respect for them just as I have none for the Redmond Reprobates. The contempt I have for them was created by their unacceptable behavior. It probably makes no sense in the grand scheme of things to believe that any company from China, the enemy of the free world, is worthy of trust or respect. It is unfortunate that any economic dependency on them exists. Cheap slave labor is profitable for greed-driven companies.

I'm never owned an Asus GPU. I have had zero problems with their cheaper mobos like Strix and Prime. I just don't like them as a company or their sucky warranty service. If I'm given the option between their brand and something equivalent in a competing brand I'm just more inclined to choose the alternative. I'd consider buying a Strix 4090 if it were cheaper, but I don't like the idea of paying a lot extra for something that's not going to give me a lot extra, or the idea having to fart around arguing with crooked bastards that don't want to honor their warranty if something goes wrong with their overpriced piece of crap.

I might do an Elmore voltage mod at some point and would want something overbuilt.

Yes, that was the video I referenced in reply to Brother @Rage Set but there are not any good views of the PCB. As best I can tell trying to judge by bumps in the thermal pads from far away it looks like it's probably 28 phases. But, I wouldn't swear to it. It sucks that he could not get the correct socket and soldered one on that will not work. I'm also not understanding how the owner of that card managed to break it off the motherboard unless they were like a bull in a china cabinet.

I am not sure. Nothing published is clear on 40-series, and looks are often deceiving.

I am looking for more technical details on the 4090 Aorus Master. That sucker has a HUGE air cooler on it, probably the best air cooler available anywhere. I have no idea what it has under the hood, or if it even has anything worth writing home about.

I agree with some qualifying caveats like living within your means. I think it is not a matter of if, but when, I decide to waste money on a 4090. The temptation is constant, but I am attempting to temper that with good sense and think about what, if any, actual real benefit I would realize from the expense. I need to convince myself that the degree of pleasure in having it will be enough to justify the spend. There is no reason (other than temptation) to feel like I "need" to hurry up and spend money. I am somewhat apprehensive about spending a crapload of money and ending up being disappointed and running out of things to do with it in 60-90 days. I see a fair bit of information suggesting that the overclocking capacity isn't remarkably different, even stunted, among the variety of models and price options due to the idotic lack of voltage adjustment that all of them are crippled by. I don't care about having "normal" 4090 benchmark scores. For example, I'd rather have impressive A770 benchmarks score that distiguish me from other A770 owners rather than achieving status quo mediocrity with a 4090. But, I am not going to settle on a crummy sample. What I am finding difficult is to understand the differences in build quality between them because the information is not being presented clearly and documentation available is inadequate for making an informed buying decision. If the idea behind going with a 4090 is gaming I think it doesn't matter... just get the cheapest one you can and move on. Because my sole intent is overclocking and benching, it actually does matter. I keep narrowing it down after watching YouTube teardown videos and manually counting phases, etc. I can always just cross-flash a vBIOS for a higher power limit or a tad extra default voltage so the firmware power limit or the stock boost clock isn't tremendously important to me. I've got it narrowed down to a few acceptable options and basically rejecting everything else as a possiblity at this point. Unless my perspective changes for some reason, the only acceptable choices right now are (in order of preference): MSI 4090 Suprim X (liquid or air - cheapest of) - 30 phases / memory current 280; max current 1820 NVIDIA 4090 FE - 23 phases / memory current 210; max current 1400 Zotac 4090 AIRO - 28 phases / memory current 220; max current 1320 Colorful 4090 Vulcan OC-V (same specs as Zotac AIRO but probably will never be available in the US) ROG 4090 Strix OC - 28 phases / memory current 280; max current 1680 (excellent specs on this one, but last on my list due to not liking ASUS or their crappy warranty support and ludicrous overpricing) TBD - Gigabutt 4090 AORUS Master (need more detail on this one. Could potentially move up to spot 2 or 3) It is ridiculous that this information has to be harvested manually and I can't help but think there is some kind of nefarious intent driving the lack of disclosure among all AIBs and NVIDIA. I am glad this person already went to the effort. RTX 4090 VRM meta-analysis and FE/AIB comparison

I watch his videos all the time. He is talented even though his presentation is a bit rough. I enjoy watching what he does. Did you see the one where he repaired the broken power connector on the MSI 4090 and installed a connector that can't be used due to the wrong orientation? I am not buying this, but I thought it was interesting to see how many carts it is already loaded into. Good price in the context of scalping prices, but still a very poor value for $1800.

The driving force behind turdbook design is deliberate production of poorly engineered throw-away trash that can't be practically repaired so they can sell a replacement for it when it dies outside of the standard warranty period. The OEMs never cared about customers consumers having the ability to upgrade, but they kill both birds with a single stone this way. The zombie sheeple herd think it's OK and gleefully throw their money away on compromised garbage, so this will never get fixed. Everyone will have what the lowest common denominators are willing to put up with.

Maybe look into one of these. Run 4 NVMe SSDs on an x4 slot with no need for bifurcation support in the BIOS. Pretty slick. Like a ROG Hyper M.2 adapter except better because it does not require an x16 slot and handles 4 instead of only 2 NVMe. I think I might get one of these soon. https://sabrent.com/blogs/news/announcing-the-sabrent-4-drive-nvme-m-2-ssd-to-pcie-3-0-x4-adapter-card https://www.amazon.com/SABRENT-4-Drive-NVMe-Adapter-PC-P3X4/dp/B0BRYQH443 The Sabrent 4-Drive NVMe SSD to PCIe 3.0 x4 card (EC-P3X4) is designed to add up to four NVMe SSDs to your computer with minimal fuss. Connect 4x SSDs into the PC-P3X4 and start using. No bifurcation setting in the BIOS is required. The adapter fits M-Key 2280 SSDs and can be placed in x4/x8/x16 PCIe slots with forward and backward compatibility. This makes it ideal for adding storage to legacy machines. Your drives will stay cool, reliable, and safe thanks to the aluminum housing with built-in thermal padding. Enjoy NVMe levels of performance with support for a PCIe payload size of up to 512 bytes. Running software RAID/JBOD has never been easier. The EC-P3X4 has also been designed to be compliant with a number of power states and other features for optimal efficiency. This includes Active State Power Management (ASPM) with the L0s/L1/ L23/L3 power states and L1 sub-states. It’s also compliant with the S3 and S4 sleep states for further power control. The adapter also supports Latency Tolerance Reporting (LTR), Advanced Error Reporting (AER), and Separate Reference Clock with Independent Spread (SRIS) for maximum compatibility and state management. This makes it plug and play for most systems. Windows 8.1+ or macOS 10.3.3+ is recommended. Expansive | 4 NVMe SSDs with one slot: Add up to four NVMe SSDs to your computer with a single x4 PCIe 3.0 adapter card. The package includes everything necessary including M.2 screws and a screwdriver. Installation is simple and easy! Resilient | Cool and Dependable: Cool your drives efficiently with the included aluminum heatsink. Integrated silicone thermal padding helps maximize heat dissipation so that your drives never throttle. The housing also adds structural support to the adapter card for more protection. Compatible | Fast with legacy support: Fits in x4, x8, and x16 PCIe slots. The adapter card is backward compatible with PCIe 2.0/1.0 at respective maximum bandwidths.This adapter card is an effective way to add additional NVMe storage to older systems and can take advantage of higher speeds and lower latency. Supports a PCIe maximum payload size of 512 bytes for maximum performance. Efficient | Wide power protocol support: Supports multiple power protocols including Active State Power Management (ASPM) with the L0s/L1//L23/L3 power states and L1 sub-states. The adapter is also compliant with the S3 and S4 sleep states. It supports Latency Tolerance Reporting (LTR), Advanced Error Reporting (AER), and Separate Reference Clock with Independent Spread (SRIS) for maximum efficiency.

I've actually thought about just not having any laptop at all and finding a cheap secondhand Chromebook to access my desktop remotely when traveling. I normally turn my computers off when I am not using them, but I could leave one of them turned on for remote desktop access. Then if it gets lost, stolen or stops working it wouldn't matter much. I could just find another cheap used Chromebook. My interest in mobile computing is almost non-existent at this point. The products available are universally disappointing and unappealing. I don't think it is even possible that I could be truly satisfied by any notebook at this point, even if it had a desktop CPU and MXM graphics.

The desktop 4090 is only a TGP of ~450W. That's not a big deal for a desktop. I get a real chuckle out of the people that pretend it is. It takes about twice that much to get me excited. That shouldn't even be a problem for a laptop that's worth having. The hard part is actually identifying a laptop that's worth having.

If we are all honest there are things that we have to put up with that suck real bad. Putting up with it doesn't mean that you like it, agree with it, approve of it or condone it. Choosing the lesser of two evils doesn't make the lesser evil something good or wonderful. It can still suck real bad. Windows 11 sucks real bad. I continue using it in spite of its extreme suckiness because doing so is required to complete a task or activity that I am required to complete or want to complete. That doesn't make Windows 11 any less sucky. The same was true, and continues to be true, of Windows 10. I still use Windows 7 for the things that I can use it for. I haven't stopped liking it, but there are some things that I can't do with it anymore because the Redmond Retards have fabricated a scenario that prevents me from doing so. That's something worth complaining about. I love Linux and use it for the things that I can. For the things that I can't, I have to do what I have to do and sometimes that means sucking it up and using Windows 10 or 11. If it didn't stink then I wouldn't have anything to complain about. But, it still does. Complaining about it is fine. So is complaining about the people that complain about it. Complaining is useful and good. If I hear people complaining and I think it has no merit I ignore it. I don't expect them to stop complaining if I don't agree with the basis for their complaint. In fact, I want them to continue complaining about it because there are probably others that want to complain and if none of them do that simply guarantees that the things that they are complaining about will continue. Complaining is especially useful and good when there's nothing you can do about it. I would much rather have people complain about things that they hate than have them do something irrational or regrettable that could put them in jail or something worse. It is a very healthy thing to do and I highly recommend doing it. We are blessed to live in a country where complaining is still legal and a freedom that we are all entitled to. There are some places out there that complaining can get you killed.

........um.......uh........

Be sure to enable ECC on your GPU or they'll delete your CBR23 scores. *sarcasm*

Not sure how the SP ratings compare to 13900K. The SP rating looks like an impressive number on both P and E cores, but the voltage values seem crazy high don't they? 1.600V for 6.0GHz?

Here is a live example of the impact even quiet fans can have on video conferencing and video recording. As you can see in the decibel meter, the noise level is inconsequential in a live environment, but magnified in a recorded or audio streaming environment.

Most of my life I never cared about loud fans. I kind of still do not care very much, but I have to care to some degree only because I spend so many hours daily in teleconferences and video meetings now that having loud fans howling 3 feet away from my mic is just not going to fly. Even with the quiet/weak F12 fans the mic is very sensitive and the fan sound is grossly exaggerated. Sometimes people complain and I am oblivious because they are not loud from where I am sitting.

That's correct. They are not difficult at all to beat on performance. They are very inexpensive and only a good option when you don't need powerful fans to get the job done. Their only noteworthy virtue and admirable attribute is the low cost. If you only need a basic fan and they are strong enough to get the job done, then they are a very sensible option. I hate spending money on fans, but I also hate wasting money on junk. They are not junk, just weak fans. Mine are very old and have seen lots of use. They are holding up well. I do like how thin the wires are. When you are packing an array of 9 fans (3x3) onto a square radiator, fat wires with fancy sleeves quickly become more detrimental and inconvenient than useful. The thin wires are very easy to manage and take up very little space.

I am not ready to buy a 4090, but based on how many of them are being under-built, if and when I land on one it seems like the good options are more or less limited to Suprim X, FE and Zotac Amp Extreme. And, those are the hardest to find. The cheaper gamer versions that are adequate for gaming leave a lot to be desired and they are not overly difficult to source. I could have purchased 30 of the mediocre gamerboy options in the past week or two. The other options are light on power phases and more likely to not hold up to the kind of overclocking I would expect. If I were buying one for gaming, the least expensive option would be the best approach because any of them should function flawlessly at stock clocks and power limits. All they would need is a good cooler and a good warranty to function as a run-of-the-mill gamerboy 4090. It really is crazy how small they are now and how much space is wasted on gigantic air coolers. When I look at how my 2080 Ti FTW3 and 3090 KPE, even my 3060 Ti FTW3, were built, it is not very confidence-inspiring. It is difficult to comprehend how much difference there is in size and to know to what degree it matters. Either the old GPUs were engineered in a very antiquated, wasteful and thoughtless manner in terms of PCB size and complexity, or the 40-series GPUs are cutting lots of corners and setting us all up for failure with lackluster engineering. Not sure which view is more accurate.