LAwLz

Those colors are not actually part of the specification. Some of them are suggestions and most of those colors are not even official guidelines. They are just something some brands might do on their own because they feel like it makes somewhat sense. But I question if this is even a big issue to begin with. Both the "USB-A can have different speeds" and "USB-C can have different capabilities". I understand that it could be an issue in theory, but how big of an issue is it really in real life? If you ask me, it's not a big issue at all. USB-C has a lot of new things to offer compared to USB-A. Reversible plug, power delivery and video output. Doing all three inside a single cable is also not something that was possible with USB-A or any other single cable before (at least not an open standard). Some of these things could be done using multiple cables before, but that made it limiting because you were tied to a specific port (rather than being able to use any one of the available ports), a lot of them were proprietary (I don't miss the days when a HP charger wouldn't work on a Lenovo laptop) and you were stuck with whatever port you had on the laptop, because adapters were not that good. Some adapters worked fine, but USB to video for example (if your laptop didn't have any video out port) were awful because they did not actually support video out natively. So they had to do something like DisplayLink. As someone who uses docks with my laptops every single day I really appreciate being able to plug a single cable in and then get data, video and power through the same cable. Docks are not niche either. Maybe for home setups but not in the corporate world. They are very, very standard there. If you want into any fairly modern workplace today you will probably find a ton of docks or monitors with USB-C out. But as I said earlier, it's not like this is an either or situation. If we just look at USB-A vs USB-C then I do not understand how someone can genuinely say that USB-A is better. As I said before, USB-C can do everything USB-A can do, but also do it better. There is a long list of things USB-C can do that USB-A can't, and there is nothing USB-A does that USB-C don't. The whole thing about using USB-C being like playing "Russian roulette" is in my opinion a non-issue. How often do you actually run into an issue with an expectation from a port not matching what it does? I would argue that the non-reversible connector of USB-A is a far, FAR bigger issue that genuinely impacts people pretty much every time they are going to use USB-A. Unlike USB-C where there is a theoretical issue that might happen, but that pretty much never happens. I think you are being completely unreasonable and it feels like you have just decided to dislike USB-C for no reason.

I think we are still talking past each other a bit. I am not arguing for USB-C-only laptops with two ports and a pile of dongles. I don't like that either. I said earlier that I actually like that this laptop has one USB-A port, and I would rather have two USB-C + one USB-A than three USB-C in many cases. My point is more that I don't understand the argument that USB-C itself is worse, especially not for laptops. I can kind of agree that the USB-C ecosystem is messy, although I feel like the mess is greatly overstate. The issue is optional features, unclear markings, and bad spec sheets. I think we can both agree with that, right? If a laptop has three USB-C ports and only one of them supports display out or charging, that should be obvious from the markings or the product page. I don't disagree with that at all. I do however think that those things were common in the beginning of USB-C but no longer are. These days it seems like most laptops and other host devices are clearly labeled and usually all ports support all features. But I still see that as a problem with implementation and labelling, not really a problem with USB-C as a connector. A USB-A port is predictable because it basically only does one thing. It transfers data. That is useful, sure, but also very limited. USB-C always does data (on host devices) just like USB-A, but on top of that it can ALSO do charging, display out, docks, etc. Not every port supports every feature, which is annoying, but the alternative with USB-A is not that those features become more predictable. The alternative is that they just don't exist on that port. On laptops especially, charging, docking and display out are very real use cases, all of which USB-A can't do but USB-C can. The problem is when manufacturers provide too few ports, remove useful dedicated ports, or fail to clearly label what each USB-C port can actually do. That is not an issue with USB-C but rather with the implementation, and it is only an issue sometimes.

It seems like Frank Azor now confirmed that FSR 4.1 isn't coming to RDNA 3.5 at this time. He says no decision like the one reported has been made, but also says that he won't comment on "future product plans". So to me that reads like FSR 4.1 on RDNA 3.5 is not a "current plan", unlike FSR 4.1 on desktop RDNA 3 and 2. FSR 4.1 on desktop RDNA 3 and 2 are their current plans which has been confirmed and they can talk about it. If they can't talk about FSR 4.1 on RDNA 3.5 then that can't be in the same release plan as the former.

I think this is such a weird stance to have. You would rather not have something (USB-PD, or video out) than have the potential to have it? If you plug something into an USB-A port you're guaranteed to only get data transfer. With USB-C you are also guaranteed to get data transfer, but you could potentially also get power, or video. To me, USB-C is clearly superior. Not being able to support something shouldn't be seen as a benefit. Also, I don't even think the whole "you never know what you get" fear is very overstated. Since the last 5 or so years I've never ran into a situation where I plug in a USB-C port and I don't get exactly what I expect. All laptops I've used have always supported power, video and data on their ports. In some very rare cases it's been a specific port but then it has been clearly labeled, and this hasn't been an issue in a long time for me. And again, alternative is that it never works at all. There are no USB-A peripherals that would lose functionality by switching to USB-C. The same can not be said for USB-A.

I know, but David McAfee (corporate vice president and general manager of AMD's client business unit) apparently said in an interview that they do not have any plans to bring FSR 4.1 to RDNA 3.5. Maybe they are worried those GPUs won't be able to run the INT8 version fast enough to offer a decent experience? I don't know why they wouldn't do it and things might change in the future, but I wouldn't bet on the mobile chips getting it anytime soon.

Apparently FSR 4.1 won't work on RDNA 3.5 APUs, like all the current high end mobile chips from AMD. Which is a big shame because I would argue that's where FSR 4.1 would have the biggest impact. If this is true then it feels like yet another case of Nvidia showing far superior support for their stuff. It honestly boggles my mind how AMD can be so inconsistent and far behind on this. https://www.hardwareluxx.de/index.php/news/hardware/prozessoren/69300-comeback-des-ryzen-7-5800x3d-mit-herausforderungen-womöglich-kein-fsr-für-rdna-3-5.html This means not even Strix Halo will get the new upscaling. Hell, since Gorgon Halo is rumored to be the same GPU, it also means that not even the next generation of AMD's laptop products will support FSR 4 or FSR 4.1.

There existed clearly labeled cables before LTT made their cables, and I don't think the cables have been a major issue. The issue, if you ask me, has always been the ports themselves. How do you know if one of your 3 USB-C ports supports video out, or fast charging, or faster speeds etc? Personally I don't think this is a big issue, but I imagine that the people who are confused by the various things potentially supported by USB-C are mostly confused by the port side of things, not the cable side.

I have to admit, I find it a bit sad that so much of the discussion around this machine has turned into another USB-A vs USB-C debate. To me, the actually interesting part is the laptop itself. Nvidia entering the Windows laptop space in this way is interesting. The chip is interesting. The unified memory setup is interesting. Windows on ARM with a proper Nvidia GPU is interesting. The display, trackpad, form factor and target market are all more interesting than whether one of the ports is rectangular or oval. But since everyone apparently wants to talk about USB ports, I might as well add my two cents. USB-C is clearly the better connector. Not necessarily because every USB-C port supports every possible feature, because obviously they don't. Some support display out, some don't. Some support Thunderbolt/USB4, some don't. That can be annoying. But that does not make it worse than USB-A. USB-A also does not support video out, Thunderbolt, high wattage charging or any of the other things people complain about USB-C being inconsistent about. At its minimum, USB-C can do the same basic USB device stuff as USB-A, while being smaller and reversible. That alone makes it the better physical connector. In an ideal world, I would love all my peripherals to be USB-C. I don't think USB-A is better. I just think it still exists everywhere because of legacy devices. Not everyone has thrown out all their USB-A devices to bought new USB-C ones. I still have a lot of USB-A things. Mostly USB memory sticks but also some external hard drives, mouse, keyboard and so on. That is why I still like having one USB-A port on a laptop like this. Not because USB-A is the future, but because it is useful today. I very rarely need more than three USB-C ports at once, but I do still run into situations where I need USB-A. So for me this is not really about USB-A being better than USB-C. It is just a fact that for a lot of people (me included), USB-C is better but one USB-A port adds a lot of flexibility. If the choice is four USB-C ports or three USB-C ports plus one USB-A, I would take the latter pretty much every single time. I don't want USB-A to live forever, but it makes the machine more useful without really taking much away today.

Yes, I notice the difference. I have actually acknowledged that difference, several times. It almost feels like you are not reading my posts at this point. iMessage has the better user experience here as long as everyone is using Apple devices. Apple runs it, Apple controls it, and it only has to work inside Apple’s own ecosystem. RCS is not trying to be "iMessage, but from Google". RCS is a carrier messaging standard meant to replace SMS/MMS across different phones, manufacturers, apps and carriers. So saying "it needs carrier support" is not some shocking contradiction. That is literally what a carrier messaging standard is. This is where I think we keep talking past each other. You are arguing "iMessage works more smoothly for the user when everyone is on Apple devices". I agree. I have never said anything contradicting this. This isn't some shocking revelation you just revelation to me. What I am disputing is the leap from that to "therefore RCS is a scam", "therefore carrier support makes no sense", or "therefore Google Messages always requires carrier RCS support". Those are very different claims. You also keep replying as if I said RCS is entirely independent and works everywhere with no conditions. I did not say that. I said Google Messages can use Google-hosted RCS/Jibe without carrier RCS support, but apparently not all regions anymore. That is why it works for me on a Samsung phone in Sweden even though my carrier does not provide its own RCS implementation, while it apparently does not work for you in Slovenia. That's too bad for you. I just want to remind you that you're not the center of the universe. Your experience is not the universal truth or the only one that matters. Yes, iMessage is better in this specific regard. Nobody is denying that. But the reason it is better is also its biggest limitation. It is a single-vendor Apple service, locked to Apple devices. It is smooth because Apple controls the whole thing, not because it is some open universal messaging standard. But the moment you mix in anything that is not an Apple device, iMessage does not gracefully remain this magical universal system. It falls back to SMS/MMS, loses the iMessage features, and you are right back in the same old messaging mess. So the "it just works" argument is only true inside Apple’s walled garden. RCS is messier because it is trying to be interoperable across carriers. Your situation is a good example of that mess. Your carrier does not support RCS, and Google's fallback apparently does not work in your region. That sucks, and Google should communicate those limitations much better. But "this rollout is fragmented and badly communicated" is not the same thing as "RCS is a scam". It just means a carrier standard has a worse and messier rollout than a proprietary service controlled end-to-end by one company.

"All this bullshit" meaning "your carrier has to support it". So the same requirements as for example text messages, MMS, phone calls, 5G, 4G and so on. I would also argue that iMessage is inferior since it only works on Apple devices. I think that is a rather big drawback.

It highly depends on the workload. Moreso than usual. We have quite a bit of info about the DGX Spark, which RTX Spark seems to be a rebranded version of. Someone ran Cyberpunk 2077 on the DGX Spark and got about 50 FPS (please note that it was running x86 to arm translation as well). Strix Halo on the same benchmark got about 90 FPS. On the other hand, for compute the DGX Spark can be 10% to 100% faster than Strix Halo. CPU wise Strix Halo should pull ahead. It's has 16 Zen 5 cores. The X925 cores in the RTX Spark can match a Zen 5 core fairly easily, but half the cores are A725 cores which are something like 30% behind Zen 5 in performance. So I would say that Strix Halo will probably be better for most users. But for some users, that really know what they want to do (mostly AI stuff) then the RTX Spark will pull ahead.

I didn't know that. I also thought good hardware on paper + Nvidia = good. I quickly read through the thread and it seems like it might be a mix of things. Someone reported that there is a power delivery bug which causes performance to tank. Then there is the special GPU core flavor they use as well, like you mentioned. Apparently the GPU has double the amount of Tensor cores and TMUs per SM compared to desktop Blackwell, but about half the ROPs. I just have to assume that the power stuff will be fixed before it rolls out in the RTX Spark, and the Tensor stuff will probably be fixed as time goes on.

I feel like people are very negative towards this but I am not sure why. I think Nvidia getting involved in the Windows laptop space is really interesting. I like that we are starting to get more unified memory on Windows machines. I like that large haptic trackpads are getting more common. That screen, despite not being OLED, sounds like it will be really good. I like that they include one USB-A port. I would rather have two USB-C + one USB-A than three USB-C ports. I do still end up needing USB-A from time to time. USB memory sticks, some odd adapter, maybe a mouse or keyboard and so on. I rarely if ever need more than two USB-C ports though. The RTX Spark is sounding really good too. I assume it's the same chip as the DGX Spark. It has 6144 CUDA cores built on the blackwell architecture. So that puts it somewhere between the 5070 mobile and 5080 mobile. Although those chips might be clocked higher. But with access to less RAM. We also know it has 20 CPU cores. If it is a rebranded DGX Spark then it has 10 Cortex-X925 cores and 10 Cortex-A725 cores. The X925 cores are in general on-par or faster than Zen 5 and Lion Cove cores even when it runs at a lower clock speed. The A725 cores are roughly on par with Skymont, which I believe delivers something like 70-80% of the performance of Lion Cove. So still pretty damn good. This chip should deliver really good performance on both the CPU and GPU. The only issue I see is probably Windows on ARM. It will probably be fine for general use, but people buying these premium laptops probably does more than just general use. I guess that's why they are going with the AI angle. Because that will work well.

That is not what I said. You seem to be treating those as 2 conditions that both have to be met, but that is not how it works. It is not: 1) Your carrier must support RCS AND 2) Google Guest RCS must be available It is: 1) Your carrier supports RCS OR 2) Google Messages can use Google Guest RCS as a fallback If either of those works, you get RCS. These conditions are not "arbitrary" or "nonsensical". I really don't understand how you can think "your carrier needs to support this protocol for it to work" is "nonsensical". Do you think the conditions for text messages and MMS are "nonsensical" as well? Because those are also things your carrier has to support in order for them to work. The Google guest program seems to be arbitrary where it works, but you have to remember that it was a temporary thing that Google did because carriers were bad. My carrier does not support RCS. I am still able to use RCS on a Samsung phone through Google Messages, because Google's guest program works here. In your case it sounds like neither route is available. Your carrier does not support RCS, and Google's fallback apparently does not work in Slovenia. That sucks, and I agree that it makes RCS a confusing for some in practice. But that is not the same as saying there are "2 conditions" for RCS to work. There are 2 possible ways for it to work, and unfortunately neither seems to work for you. iMessage has a better user experience in this specific regard because Apple controls the whole thing. But that still does not make my original point wrong. RCS through Google Messages is not Pixel-only, and it does not always require carrier support. It depends on whether Google's fallback is available where you live. This is also why I object to calling it a scam, a word that is brutally misused these days because it gives Youtubers a bunch of clicks. "Scam" does not mean "it is bad", "it does not work for me", or "I do not like it". It means there is some kind of deception or fraud involved, and I do not see how that applies here. I think your frustration is reasonable, but I still think you are blaming the wrong thing. If your carrier does not support RCS and Google's fallback is not available in your country, then the main problem is still your carrier not supporting the standard.

I understand that it sucks. I would be annoyed too if RCS just randomly didn't work in my country. But you are mixing up two different things here. RCS being a universal standard does not mean "Google personally guarantees that every user in every country can use Google-hosted RCS forever, regardless of carrier support". It means it is meant to be a cross-device, cross-carrier standard, unlike iMessage which is Apple's own proprietary service. The problems in your case seems to be that: Your carrier does not support RCS. Google's Guest RCS / Jibe fallback is apparently not available in your region. That is bad. I agree with you there. Google should be much more clear about where this works, where it doesn't work, and why. It is also completely fair to say that this makes RCS a mess in practice for users in affected countries. But that still does not make it a "scam". I am not sure what you think the word "scam" means, but you are using it incorrectly. It also does not change my original point. It is not true that RCS through Google Messages is Pixel-only, and it is not true that it always requires carrier support. I am literally using it on a Samsung phone in Sweden with people on other Android brands, despite our carrier not having their own RCS implementation. So I think we are talking past each other a bit so I will try to clear things up. You are saying "RCS is bullshit because it does not work for me in Slovenia". I am saying "RCS can work through Google Messages without carrier support in many places, but apparently not in every region anymore". Both can be true. Your situation sucks, and I agree that the rollout is way more confusing than it should be. I just don't agree that this makes the whole thing a scam, or that Google Messages universally requires carrier RCS support. I also don't think this is Google's fault. I really don't see how anyone can think it is if they have read up on RCS.