I'm loving this eGPU so far. Details and benchmarks posted at

eGPU.io

And if curious, the recent firmware update has enabled Resizable BAR for the 150W RTX 3080.

Quote from: Mothertrucker19 on March 04, 2021, 09:06:26

Quote from: Aegis on March 03, 2021, 14:27:31
While the CPU performance will be exactly the same.

Except if the CPU is limited by the laptop's GPU, in terms of power or cooling. (in a gaming sceneraio)

The gpu being measured in this review is an external gpu with its own power supply so it can not limit the CPU by added heat (unless you blew the egpu exhaust directly into the laptop chassis) or power budget.

Quote from: Mothertrucker19 on March 04, 2021, 09:06:26

Quote from: Aegis on March 03, 2021, 14:27:31
While the CPU performance will be exactly the same.

Except if the CPU is limited by the laptop's GPU, in terms of power or cooling. (in a gaming sceneraio)

From what I've read, the CPU runs in a 15W mode while undocked, at least in the default performance mode. Though it also seems that there are higher performance modes (meant to run in "tent" or "stand" modes where the fans have better access to cool air than the standard laptop mode) available also without the dock. When docked one would assume the internal dGPU is disabled and the entire internal cooling system is dedicated to the CPU.

Whether this has major effects on performance or not depends largely on how heavily threaded the games in question are, as a couple of heavily loaded threads should be able to run at relatively high clocks even at 15W.

Quote from: Aegis on March 03, 2021, 14:27:31
While the CPU performance will be exactly the same.

Except if the CPU is limited by the laptop's GPU, in terms of power or cooling. (in a gaming sceneraio)

Very interesting and about a tier better performing than I expected, at least for whoever the laptop is suitable this is a really good solution. Personally I'd need good stylus support at this size so I'll do with a separate 15" and an integrated 3080.

Quote from: Aegis on March 03, 2021, 08:30:14
TL;DR This set up more or less performs equally to a regular RTX 3080 gaming laptop while being more portable and expensive.

Pretty much, yes. More portable, more flexible, but no more quiet or powerful (except for thin-and-light versions, though those come closer on the first two points). Should be a fantastic option for frequent travellers or those who want to be able to easily bring their eGPU with them for other reasons.

Quote from: Compare on March 03, 2021, 13:28:55
Why not add the XG Flow _without_ XG mobile results to the charts? Is this whole setup faster? I assume so, but maybe not since it's not in the comparison.

How is the "lapability" with this setup? Does it impact battery life?

Lapability is zero. Battery life is meaningless. Why? Because the XG Mobile requires an AC plug to work. It's meant to use on some sort of surface. I mean, you could balance this on your lap somehow, but given the short (30cm?) cable of the XG mobile, that would leave you with a big, hot, heavy brick to balance alongside your laptop. Then, to be fair, no RTX 3080 laptop is really lapable - either it's too hot or too big.

Performance comparisons are also rather meaningless, as the eGPU is in another class entirely from the built-in dGPU. You can also manually add those results through the search box below the graphs. Tl;dr (all comparisons vs. 3080 on integrated display): Time Spy seems to not have been tested, Fire Strike Ultra 1635 vs. 7441 points, FS Extreme 3934 vs. 13397 points, while the XG Mobile just barely beats the 1650 in Sky Diver and Cloud Gate, where the integrated display seems to significantly bottleneck performance (likely due to the extremely light loads and insane frame rates stressing the PCIe link). For game benchmarks, just go plug in "Flow X13" into the search field and check the results yourself. Mostly the 3080 is 2-4x faster than the 1650, but it depends heavily on the game.

Quote from: Compare on March 03, 2021, 13:28:55
Why not add the XG Flow _without_ XG mobile results to the charts? Is this whole setup faster? I assume so, but maybe not since it's not in the comparison.

How is the "lapability" with this setup? Does it impact battery life?

Because it is meaningless to compare results with and without XG mobile. Surely, it is obvious that the graphical performance will be a lot better with the XG mobile. While the CPU performance will be exactly the same.

Why not add the XG Flow _without_ XG mobile results to the charts? Is this whole setup faster? I assume so, but maybe not since it's not in the comparison.

How is the "lapability" with this setup? Does it impact battery life?

TL;DR This set up more or less performs equally to a regular RTX 3080 gaming laptop while being more portable and expensive.