The tech seems to have yassified Grace in the main promotional image that’s being used by every publication.
The rest of the images look improved, for the most part, and I do think it does make sense to offer this functionality as an option in the upscaling pipeline.
I’m sure that those repeating artist’s intention have never dared use a reshade or texture swap.
Hopefully, that yassification will be a side effect of the current state of the tech and not something that ships to end users.
This seems like it’s intended as a texture and lighting improver, not an “AI Slopificator”
Among the other screenshots, a lot of them seem to have a marked improvement.
Aspects of that Grace image comparison definitely look bad, but this is a work in progress that we’re getting a glimpse into. I really hope that bimbofication doesn’t make it into the final product.
The use of two GPUs is required right now as DLSS 5 still has a long way to go in terms of optimisation - both in terms of performance and its VRAM footprint. However, DLSS 5 is designed for use on a single GPU and that's how it will ship later this year. Quite how scalable it is also remains to be seen, but in common with other DLSS technologies, Nvidia tells us that the computational cost scales with resolution.
It seems like the problematic variable is likely to be external to the PCs themselves. My inclination would be to have the end users disconnect everything but the monitor to see if the systems POST, and then reintroduce peripherals one by one.
Picked up all the Splinter Cell games on a sale and started with the first using the EnhancedSC Mod, on Steam Deck.
That’s not technically a “Patience” thing as I’m revisiting games from my childhood, but it sure is fun!
My partner and I have also been playing Harold Halibut. The story and visual style are an absolute delight, simpler adventure games are awesome when you’re tired or burnt out!
I had my eye on it since it was first announced, it sat on my wishlist waiting for a < $20 sale, and then the GOG version got included “free” via Amazon Prime Gaming.
Always nice to have things pan out that way, there have been too many occasions where I bought a game, didn’t get around to playing it, and then it turns up free or “free”.
Full filesystem validation is probably a good place to start:
chkdsk /r
dism /online /cleanup-image /restorehealth
sfc /scannow
I generally run Memtest86 and a disk diagnostic before running this - an adverse condition in either part can make these commands cause problems instead of fixing them.
Whether or not the account gets locked should be based in your Windows settings, if you have it set to require password after display timeout / “screensaver” then it will also require it in this context
21002Mhz
Assuming you mean 2133Mhz or 2666Mhz here
Which Ryzen 3000 Chip, how many memory modules, and how many ranks on each module (single-rank has chips on one side, dual-rank has chips on both sides)?
Memory support is going to vary on those three factors.
Search for your CPU and open AMD’s “Drivers and Downloads” page for that CPU, then expand the “Connectivity” section and scroll to see what your CPU officially supports.
Most CPU samples can run faster than what’s listed there, but it’s not a given.
In any case, I recommend running 4-8 passes of Memtest86 to make sure a memory configuration is stable. That becomes especially important when running “out of spec”.
Memtest86 can also be used to validate whether or not your memory (or CPU memory controller) is defective. Statistically speaking, it’s usually the memory.
If needed, you can enable XMP and then go under advanced settings and manually set a a specific memory frequency.
Please note that the lack of BSODs does not necessarily indicate actual stability. I’ve diagnosed plenty of cases where a client’s installation is silently corrupting over the span of months because of an erroneous memory configuration. The only way to know a config is for sure stable is to set a configuration, run Memtest86, and then see 4-8 passes complete without reporting any errors.
You’re not entirely wrong, in that the Apple Tax is real.
Nonetheless, the quality of the Magic Keyboard is substantially higher than that of a keyboard you can get for “few dollars”
Ultimately, your assertion was:
An iPad Air costs the same but comes with a much better M4 processor. The main difference is a less crap operating system in macOS.
An iPad Air with a keyboard that matches the form factor and build quality of a MacBook Neo does not actually cost the same, it costs an additional $270.
I’ve been meaning to go back into this, played at launch and spent most of my time in the modding metagame that I don’t have time for anymore.
Did you mod this yourself, use a Wabbajack list, or do a combination thereof?
If you’re mostly using Wabbjack, how was the learning curve for the various VR systems? Fairly intuitive, or necessary to spend time reading outside of VR?
The laptop form factor is engineered with lid and palmrest assemblies, if you’re going to compare the two then you’ll want to add a nice keyboard to that iPad. Apple’s is $270.
Memory utilization is relative to the user though. For someone who wants to do nothing more than check their email and manage online banking, no specs matter (well, within reason, but people do use Chromebooks with 4GB RAM)
Just because such a system would not be suitable for your use-case does not mean that it is not suitable for any use case.
The iPhone 16 Pro is a very capable device, yet it “only” has 8GB of RAM. We don’t have the full picture for these new devices, but it’s possible that Apple will be handling memory in a similar manner to iOS, making it possible to do more with less.
Repairing broken hinges on such a cheap laptop practically has to be a DIY repair. I get this exact repair inquiry every now and then, the owner often balks when the repair cost is more than 50% what they paid for the device.
For these low-end laptops, I also find that parts are usually less available than those for most Apple devices. Apple tends to use certain part designs / assemblies for multiple generations. Apple stuff is consistent enough that there are plenty of used parts available aftermarket.
Far as your repair scenario is concerned, I can only think of 2-3 times where a Mac came in with hinge related failure and those cases all stemmed from abuse like opening the lid too far / egregious mishandling.
Meanwhile, I’ve bread lots of butter with HP laptops whose hinges break through regular operation.
If something costs more to fix but only breaks 1% as often, are you really saving money by purchasing the cheaper solution with the higher fail rate?
There’s more to a computer than RAM (or even ither specs), comparing what’s shown in the article to the low-cost option you linked the two systems are leagues apart in terms of build quality.
Wouldn’t be surprised if the battery life was miles apart too.
That cheap plastic HP laptop is destined to have its hinge mounts snap away from the upper palmrest through normal day-to-day use.
Back when LLM/AI was fairly nascent in the public eye and largely relegated to image generation, I had a real off character come into my computer repair shop wanting to ask questions about it but not knowing what to call it.
He wound up describing it as “you know, when they make the computers dream”
The tech seems to have yassified Grace in the main promotional image that’s being used by every publication.
The rest of the images look improved, for the most part, and I do think it does make sense to offer this functionality as an option in the upscaling pipeline.
I’m sure that those repeating artist’s intention have never dared use a reshade or texture swap.
Hopefully, that yassification will be a side effect of the current state of the tech and not something that ships to end users.