SomeEdward

Unfortunately, the lowest I can set is 71 for PPT (which I used already) and 47 for TDC - that doesn't change anything because running CB single core, watching youtube and listening to music, it's only at 60%.

I recently got myself one of these (Apex 15 laptop with a 3950X) and while trying to tweak it to my liking, I found that multiple single core workloads are a problem for thermals (at reasonable fan speeds). Because when using PBO with 71W (can't set anything lower in Ryzen Master), the CPU is like "Oh, I'm allowed to boost to 4.7 Ghz and I don't mind exceeding 90 degrees.". With a manual profile, I could tame that by limiting it to something lower but then multi-core work thinks it's fine to use e.g. 3.8 Ghz on all cores at the same time... And so I would love to have something like PBO but limit it to 3.8 Ghz or whatever. But at least in Ryzen Master, that doesn't seem possible? Am I missing something? Would it maybe be possible with some other tool, like ryzenadj? (Which doesn't work on this laptop currently but if it would be possible to get what I want there, then I would dig a bit more...)

Is there an argument here? It may not be the most professional company but hey, if the product works well (and so far, all benchmarks seem to indicate that it does) and Schenker/XMG provides good customer support (notebookcheck recently did a survey of ~1000 people and they won by a landslide), why not?

Actually, it's a bit smaller than my HP Probook 450 G5. Although 50% thicker and 25% heavier. But neither of that is a big deal IMO. I can fit the Probook into my backpack just fine and so the Apex will work as well. As for battery life - the battery does seem a bit on the weak side (wish they would've chosen ~90 Wh over those ~60 - but I don't know how much that would've impacted the weight) but you can switch to a power saving profile (~28W) while on battery. The Asus TUF A15 for instance throttles a LOT more when on battery. Which... is of course a blessing and a curse, especially since you can't even choose to use a more powerful profile. I guess XMG could choose to spread out their profiles even further to offer something similar. Although the problem then may be the dGPU. I haven't read anything about profiles for that and it seems to actually gobble up the most juice.

Yes, it does. And good to know, thanks!

I've read reports of many people by now and none of them have mentioned that, so it might just be your setup? Did you actually check that you e.g. have a stable ping of max. 1 ms?

Wireless SteamVR (with Oculus Quest and Virtual Desktop) Currently, this is the probably best video with the most views on this topic. As you can see, probably not as many people have seen it as they should've. Now, I'm still waiting for my Ryzen 4000 laptop and can't try it out myself until then but after looking a bit more research into this topic, it seems perfectly feasible physically. After all - if you get a stable <1ms to 1ms ping from your PC to your Quest, the difference compared to using Oculus Link should be negligible. So aside from making more people aware of this option, you could possibly come up with more accurate techniques to compare quality and latency (both video stream and input) compared to the video I linked to. The primarily interesting comparison of course being the Link but potentially also e.g. the Index when it comes to latency? (Personally, when I realized this, I immediately ordered a Quest. Being able to spin without having to worry about wrapping a cable around yourself seems so awesome!) (And as another aside - you could also mention that Oculus tried to get rid of Virtual Desktop. I guess they prefer people to stay tethered with their Link.)

That polish review is awesome, thanks! Quite the little beast, it seems. Odd that it never got reviewed by more prominent sites.

Speaking about limited data - Eisbaer 120 is in the list but I only found this review: https://www.hardwareasylum.com/reviews/cooling/alphacool_eisbaer-120/page4.aspx Do you recall any other sources? Because that seems like a very interesting product (copper radiator and would theoretically get rid of as much heat with just one relatively low throughput fan where others in the same tier usually require 2 or higher throughput ones) but I'm not all that confident because of the lack of data.

So... Linus Tech Tips' own benchmark here: Is a case of an outlier? Because here, a Tier 4 cooler is declared superior to a Tier 1 cooler. (I'm not just being snarky but am looking into buying a new cooler and all the conflicting info is driving me nuts )