More Mobile: Troubleshooting Update (Premium)

One month after my last Most Mobile update, I wanted to provide an update to explain the troubleshooting I’ve done, and the results I’ve seen.

Most importantly, perhaps, nothing I’ve done has exonerated Intel’s 12th-Gen Core chipsets. The problems I’ve experienced only happen when I use PCs based on this family of chips.

I also wanted to clarify what the problem is: while Chromium-based web browser problems are the most obvious symptom I’m seeing, too many are focusing on this one area and forgetting a crucial piece of the puzzle: the problems I see only occur when the PC is docked to a USB-C hub or Thunderbolt 3/4 dock (I’ve now tested on both). And it’s not just browsers: I’ve seen other performance issues as well. In fact, let’s not forget my original supposition, that Intel’s shift to an Arm-like hybrid architecture with both Performance and Efficient cores has required it to introduce a 28-watt P-series lineup between the 7/15-watt U-series and the 56-watt H-series. In other words, it wasn’t seeing the same performance with what used to be the mainstream U-series. For the 12th-Gen, the new P-series has moved into that slot.

I should also point out what I’ve not done (yet): I’ve not spoken privately or on the record with Intel or any PC maker, and I should do so. That said, I know they’re reading what I write, and I’m surprised no one has contacted me. Just as I’m surprised that no other laptop reviewer, to my knowledge¸ has pointed out any performance problems. Instead, everyone seems to rely on benchmarks, in which the 12th-Gen chipsets seem to outclass their predecessors by a reasonable percentage. (That said, remember that any year-over-year comparison of a particular PC model is comparing a new 28-watt system to what was a 15-watt system last time (in most, but not all, cases).

So let’s step through some of the testing I’ve done. Starting with a few things I either forgot to mention before or maybe didn’t explain clearly enough.

Regarding docking, I use a USB-C dock at home in which a laptop is connected to an external display of some kind, and I most typically do not use the laptop’s built-in display. I’ve tested this setup in three different configurations—using the USB-C dock built into the HP Conferencing Monitor, the USB-C dock that I usually travel with tied to an HP Z27n display, and a Thunderbolt 3 dock again with that HP Z27n display. I experienced problems in each case with multiple 12th-Gen Intel Core-based PCs. And never with two AMD Ryzen-based laptops (which I only tested over USB-C).

In our Mexico City apartment, I used a Thunderbolt 4 dock and that same travel-based USB-C dock with 2-3 different laptops on the most recent trip, and in each case, the laptops were not connected to an external display. The 12th-Gen PC (an HP Envy 16) experienced the performance issue with the dock and the hub, a 10th-gen (I think) HP laptop I left in Mexico experienced no issues at all in either configuration. And a Qualcomm Snapdragon-based ThinkPad X13s laptop experienced no issues either, though I only used it with the USB-C dock. So once again, a 12th-Gen Intel Core-based system was the only one that had issues.

I also briefly tested an M1-based MacBook Pro with the USB-C dock here at home. No issues.

Looking over the feedback from my previous article, I see some obvious and necessary questions around such things as the cables, but I’ve only used the original cables for the most part: the USB-C cable is hardwired to the Thunderbolt 4 dock, the cables I use with the HP Conferencing Monitor and Thunderbolt 3 docks are original, and I’ve documented which cable I use with the USB-C dock, and that extender has always worked fine. Until the 12th-Gen Intel chipset arrived. As was the case with each configuration. I don’t think it’s a cable issue.

That said, for the past month I’ve used the Anker USB-C dock without that extender cable, in tandem with that HP Z27n display and over HDMI. (Previous external display connections were over USB-C.) Again, the same problems with 12th-Gen systems. No problems with AMD. I basically already knew this would be the case because this problem occurs across multiple docs and hubs. But that’s one more proof point out of the way.

Previous to this week, I spent a week or more running Firefox instead of a Chromium-based browser. And go figure, this did results in fewer issues with the 12th-Gen Envy 16 I was using in my home office. But not no problems: it never hung (with a “white out”) because of too many tabs, an issue I saw again and again with Brave, Chrome, and Edge. But it did have some weird performance issues where it was unresponsive for up to 20-30 seconds at a time. It didn’t happen as often, for sure.

For the past several days, I’ve been using an AMD-based EliteBook 865 with the USB-C dock and HP external display, and I’ve had no issues, including with Edge or Brave. I’ve recorded multiple podcast episodes, including the most recent Windows Weekly, have opened dozens of tabs in Chromium-based browsers, and have never seen even a hint of performance problems. (I should add the USB-C cable extender, come to think of it.)

Is this enough?

No. It’s not enough. And it’s possible that this problem I’ve communicated can and will be fixed via software updates from Microsoft and/or Intel. (The problems are less frequent now than they were earlier in the summer, though the browser issues are consistent.) And that the 13th-Gen chipsets, which arrive in mobile form next month, could likewise be updated to prevent the issues I see. That this hybrid architecture is just immature and needs to evolve.

Or, maybe I’m nuts. I guess we need to consider that possibility. But I’m at least working through it as logically as I can. And giving each configuration change the time it needs. This won’t happen quickly.

But it’s happening. And that’s what I’ve seen over the past month.