Aaron44126

I got some China knockoff aftermarket fans from an eBay seller when I was trying to get Delta fans for my M6700. I imagine that is what the Landalanya ones are. The brand that I got was "DJBJA" or something like that. They may well work fine and be suitable for you, not too hard to make an OK fan I would think... You should do some before-and-after performance measurements though, they might seem to be working fine but fail to cool your CPU as much and result in lower turbo boost speeds.

1080p should be referred to as "2K", not "1K". The name is derived from the horizontal width. 3840×2160 = "about" 4,000 pixels wide = 4K. (The same holds true for 8K and 16K standards popping up.)

At risk of being fired, Twitter engineers have created a "power user multiplier" for their "algorithm" and applied it to Elon Musk's account, to make sure that his tweets get high visibility, which is why he has been dominating many users' "For You" feed. https://www.theverge.com/2023/2/14/23600358/elon-musk-tweets-algorithm-changes-twitter

I saw they have official documentation on how to disable this message. https://support.microsoft.com/en-gb/topic/kb5017130-managing-windows-11-system-requirements-not-met-message-in-your-organization-f751d24b-a70c-4a18-9eec-dae0790e41f5

This is on the desktop side, but I just saw this... https://www.tomshardware.com/news/nvidia-rtx-4060-specs-leak-claims-fewer-cuda-cores-vram-than-rtx-3060 Desktop RTX 4060 has 3072 CUDA cores — RTX 3060 had 3584. Desktop RTX 4060 has 8 GB VRAM — RTX 3060 had 12GB. Desktop RTX 4060 has 115W TGP — RTX 3060 had 170W. Desktop RTX 4060 is a downgrade from RTX 3060?? NVIDIA marketing at their greatest... (These numbers come from kopite7kimi, a very reliable NVIDIA leaker.)

Most Precision laptop models have multiple fan models in the mix. I found this photo of Delta fans (Precision 7530 and 7540 fans are interchangeable). Here's an old thread about the different types of fans available for the Precision M4800 (see posts by @unnoticed). I ended up sourcing Delta CPU and GPU fans for my Precision M6700 after seeing those posts; Delta was preferred back then, but I'm not sure if that's still the case in Precision 7X40; in Precision 7770, Sunon is generally preferred over Delta. If you do buy aftermarket fans, you should check with the seller because there is no real guarantee that the fan model pictured is the same model that you will get. I personally use ANC headphones when gaming, fan noise becomes a non-issue. (I take steps like disabling turbo boost to limit the fan noise when I'm not running an intensive application.)

Second quarter 2023 ... Can't really pin it down more specifically than that.

I believe there is a setting in the BIOS setup regarding the boot speed which has options like "Auto", "Minimal", and "Thorough". Check and see if it is set to "minimal" and it will skip some POST checks. Also, I have noticed that having external USB storage drives connected can increase the POST time, because the BIOS will check them all to see if they are bootable before proceeding to boot the OS (even if you aren't planning to boot off of them, I think it wants to prep the F12 menu).

Dell Precision 3480, Precision 7680, Precision 7780 have been certified by DMTF. https://registry.dmtf.org/products/dell-precision-7780 https://registry.dmtf.org/products/dell-precision-7680 https://registry.dmtf.org/products/dell-precision-3480 [Edit] Dropping "quotes" from the thread title now that the model numbers for the 7000 series are confirmed.

NVIDIA has five or so mobile Ada GeForce chips launching, I think it is not a stretch to have professional chips a month or two later because they are basically the same…. I agree about the name, that is too much.

Are the external displays connected to the dGPU or iGPU? (They would be connected to the dGPU if you have graphics switching off, or if you have graphics switching on and also discrete graphics output on.) Any change in behavior if you take steps to get the external displays connected to the other GPU?

More "inside scoop on Twitter". Apparently, Musk fired an engineer who offered an explanation as to why his tweets are not getting as much engagement as he would like. Also, there is a looming FTC audit which could result in hefty fines. https://www.theverge.com/2023/2/9/23593099/elon-musk-twitter-fires-engineer-declining-reach-ftc-concerns

Just came to post the same thing. There's a video.

Yes, I'm aware of StartAllBack but I'm actually using Start11. I really like their implementation of the Windows 10 style menu (multiple pages, custom tile colors, custom tile images, etc. are things you cannot even do with the Windows 10 base menu). But the fact that third-party solutions are needed to fix up shortcomings like this is just another thing that puts me off. Windows should not be moving backwards in terms of the power user feature set.

Same. Aside from the various issues I have with Windows 11 (...wrote it up, link in sig...), I've tried to upgrade twice and both times rolled back within a day or so because it's just not ready. Dumb design things but me off like not allowing seconds on the taskbar clock, way less flexibility for the Start Menu than Windows 8/10, and then there are the outright bugs (I was getting crazy stutter/hangs on my laptop when I tried, and you only have to look at tech news outlets for things that are failing after every other monthly patch). They clearly aren't paying the same attention to QC that they used to. Not to mention that I am not a fan of them just dropping new "features" multiple times per year without any real flexibility to avoid this from non-Enterprise end users. When they're willing to commit to a long-term support version, I'll give it another shot. Prior to this I have always been the guy to upgrade to a new Windows release day one, or even earlier if I could snag the RTM bits from TechNet/etc. I happily used Vista even with its (for the time) high system requirements, or Windows 8 with its dumb user experience choices, because "under the hood" they offered serious improvements. Windows 11 is more like a new coat of paint on the surface but doesn't really seem to be "better" underneath, and in some cases it is just unnecessarily dumbed down.

I had a Best Buy rep insist that I would get better image/sound quality by using their expensive gold-plated cable.......

I have a Samsung 980 Pro 2TB, the "problem model", and it even has the "problem firmware version" (3B2QGXA73). I have over 18 TB written and it is at 99% health, so, no apparent problem. Regardless, I'll be patching it to the new firmware version within the next few days.

Looking at the motherboard render above, there is no indication that the GPU is off on a separate card. (Dell usually does these sorts of renders properly, with regard to how the system is actually put together.) In any case, if Dell did separate the GPU, I have no doubt that they'd use their proprietary DGFF "pseudo-standard" and that it would not be interchangeable with systems from other manufacturers (and likely not even interchangeable with other older Alienware/Precision systems that also use DGFF).

Locking the GPU clock speed to a certain value or range. The way I've always seen documented to do this is to use the nvidia-smi command with something like: nvidia-smi -i 0 -ac 877,1215 This would specify a specific memory and core clock speed respectively. However, this is not supported with GeForce GPUs, only pro GPUs, so it cannot be used with the GeForce RTX 3080 Ti in the Precision 7X70 systems. ...I was playing around and discovered that a command like this actually works on the GeForce GPU. nvidia-smi -i 0 -lgc 1110,1110 With this command, you are setting the GPU clock speed only (not the memory clock speed) and you are asking the GPU to stick within the specified range. In this case, I have used the same value for the lower bound and upper bound so the GPU clock speed is fixed at 1110 MHz. You can reset things back to normal with: nvidia-smi -i 0 -rgc You can see the list of supported GPU core clock speeds with: nvidia-smi -i 0 --query-supported-clocks=gr --format=csv My understanding is that the GPU may still clock down below the requested range if there are power/thermal constraints that prevent it from sticking in the requested range. In my case, I'm interested in setting the GPU clock to a fixed value just to get consistent performance. I play with some odd applications (err, game console emulators) that are sensitive to timing and hiccup or stutter if the GPU does not realize that it needs to ramp up the clock speed frequency quickly enough. There is also a parameter -lmc and -rmc to lock and reset the memory clocks... I have not played with this one yet.

They shot it down and are recovering debris. U.S. downs suspected Chinese spy balloon off the South Carolina coast - https://www.npr.org/2023/02/04/1154527155/faa-halts-flights-chinese-spy-balloon

China has admitted to the balloon currently over the U.S. being theirs (though their claim is that it is not a "spy balloon", but rather a "civilian airship" with a meteorological purpose and way off course).

Twitter is going to revoke free access to the API and make people pay for it. https://www.theverge.com/2023/2/2/23582615/twitter-removing-free-api-developer-apps-price-announcement https://arstechnica.com/tech-policy/2023/02/cash-strapped-twitter-to-start-charging-developers-for-api-access-next-week/ About a decade ago, I worked for a digital marketing company, and worked on a product that allowed clients to make promotional posts to their web site, an SMS list, Facebook, Twitter, and some other outlets simultaneously and from a centralized "dashboard", so I have some real first-hand experience with the Twitter API. I'm sure there are a number of businesses that offer a service like this. I also wrote a little scraper to turn a user's tweet stream into an RSS feed just so that I could consume it in my RSS reader rather than having to use the Twitter web site (that I stopped using after Feedly started supporting Twitter feeds directly). Being able to slap together a tool like that was definitely handy. Also as mentioned in the article, the Twitter API is the interface that bots use to function, plus third-party clients (...now banned...) and tools that just read tweets and whatever out of Twitter for research and analysis. Anyway, I wonder how many API users will really pony up for the paid access? And of course, Twitter has only given a one-week notice, and no indication of what the price will be.

Not a bad idea. I will note that the program supports command-line parameters so you can set yourself up a batch script shortcut to quickly set options like this. Run it with "/?" to see what the options are, or look at old documentation for DellFanCmd (the options are the same). However... You shouldn't try to use command line parameters to adjust the fan behavior while the GUI app is running, or the GUI app will likely end up having issues. Control via command line or control via GUI, not both.

Two "Yes" buttons. "No" button discretely hidden away at the bottom. Yes. Double yes. This is the way to do it.

Intel Arc A750 (supposedly in competition with GeForce RTX 3060) is getting a price cut down to $250 USD. https://www.theverge.com/2023/2/1/23580952/intel-arc-a750-price-cut-directx-9-performance-improvements