ASUS ROG Flow X13 | A Proprietary eGPU Set-Up Worth Your Attention

by in General | Apr, 1st 2021

The Asus ROG Flow X13 has garnered a lot of attention, in no small part because of its proprietary (and obscenely expensive) eGPU. On paper, it’s quite a nifty package, but is it really worth it? Well, the answer would almost surely be a resounding “no” but it also depends on your individual workflow and use-case scenario. For some, splurging nearly $3000 on such a portable set-up does make sense — although they’re surely in the minority.

Still, there are some benefits, and even though this whole combo leaves a bit to be desired (or a lot, depending on how deep your wallet goes), there’s still a reason to talk about it. It is, at the very least, a short glimpse at the future, and we like what we’re seeing.

Before we delve any deeper into the topic, we need to first go over some technical stuff. We’ll try to keep it as short and sweet (not to mention simple) as possible. This isn’t supposed to be an exam for MIT nor anyone’s PhD thesis — we’ll go easy for the time being!

What’s an eGPU?

An eGPU is an acronym that stands for external graphics card. Why mankind has decided to create such a peculiar thing, however, is anyone’s guess. All jokes aside, it is an exciting concept; the sheer thought of separating a graphics card from its intended enclosure makes a lot of sense for a slew of different reasons.

Simply put, most folks don’t really need a beastly GPU — integrated ones will do just fine for everyday work and some web browsing. The only instances in which that’s not the case is either if you’re a gamer or someone who has a job that is a bit more hardware-intensive (with GPU-heavy workloads like 3D/video rendering or something else of that nature).

To game, you need a graphics card. To game comfortably, you need a really solid graphics card (depending on your preferred graphical settings and resolution of choice). Now, if you’re someone who’s stationary and doesn’t really move all that much (by which we mean daily commutes, or work-related travel), then there’s no reason whatsoever for you to ever purchase an eGPU. Just build a full-fledged gaming PC and call it a day.

If, however, you’re often commuting — and really want to slay some beasts in Doom or traverse the American frontier in Red Dead Redemption 2 during your downtime — then an eGPU might be of interest. You could also just buy a gaming laptop, but they’re often chunky, loud, and can burn a hole through the ground if left unattended. The bad ones run so hot they’ll keep you warm even in the harshest of winter nights, when the wind’s blowing and you feel like the world’s getting ripped to shreds. They’re a veritable furnace, albeit with a screen that refreshes at upwards of 144 times per second. A contemporary furnace, if you will. And if you’re especially “needy” and want as many frames as your machine can muster, then it’ll draw about as many watts from the wall as a nuclear power plant.

Gaming laptops often aren’t that good of an option, unless you’re willing to invest in one of the finer ones on the market. Still, even then, an enjoyable gaming experience is far from guaranteed.

An eGPU set-up, on the other hand, is a lot more flexible in nature. You can have a thin-and-light ultrabook on your person for those “simpler” tasks, and then when you return “back to base” all that’s standing in the way between you and a spectacular skirmish on Verdansk is a Thunderbolt 3 cable. 

The times we live in!

That’s partly how eGPUs came into existence. It wasn’t just to provide creative professionals and gamers with more horsepower but also to give us all a broader set of options. But, as is so often the case with such avant-garde endeavors, it came at a cost — often too great a cost, in fact. A standard Thunderbolt 3 connection ensured a bandwidth of no more than 40Gb/s. That’s the maximum. Now, 40Gb/s is by no means a meagre amount, but it’s still not enough to allow your GPU to run at full steam. Laptops equipped with Thunderbolt 3 can support only (up to) four PCIe lanes which bottlenecks your graphics card of choice, no matter how big or beastly it might be. There’s always a performance drop-off no matter what you do and, given the fact that you’re already paying a premium — a laptop, a graphics card, and an obscenely priced external GPU enclosure — it’s easy to understand why so many folks have decided to just build a standard gaming PC and be done with it.

This problem is all the more exacerbated if you’re looking to game on the laptop’s display — that’ll incur an even bigger performance penalty as your laptop has to send data to the eGPU through the TB3 cable and then receive the image back for it to be displayed. A good workaround would be to connect your eGPU to an external display, but that’s not always an option.

For a direct head-to-head comparison, make sure to check out this phenomenal video, courtesy of Jarrod’s Tech. As showcased throughout the video, you’re losing a metric ton of power no matter which graphics card you go with. The better the GPU the better the results, of course, but you’ll still be leaving a lot of performance on the table. In other words, a 2080 Ti, for instance, won’t perform nearly as well as it should. That, in short, has always been the biggest bane of eGPU set-ups. Well, that and the price.

Enter the ASUS ROG Flow X13 and Its Proprietary eGPU

Instead of going the “traditional” Thunderbolt 3 route, Asus developed their own proprietary port in order to circumvent the aforementioned bottleneck. The result? Noticeably higher bandwidth: 63Gb/s instead of 40Gb/s

Right now, at the time of this writing, their XG Mobile eGPU comes equipped with up to a mobile RTX 3080. Now, that “mobile” prefix is of the utmost importance here. It’s just not as powerful as its full-fledged desktop variant, and yet it’ll still set you back for around $1200, depending on the retailer. Whether that’s a dealbreaker or not heavily depends on your budget, but it certainly doesn’t come cheap.

We’re talking about a 20-30% performance delta (sometimes even higher) compared to the desktop RTX 3080. And sure, the mobile variant is still mighty powerful, but it’s not like you’re getting this proprietary eGPU set-up for a meagre amount of money. You’re supposed to splurge a sizable wad of cash for a noticeably weaker version of the real thing. Hardly anything worth writing home about. The fact that they both carry the same name as well only adds insult to injury.

So let’s compare the mobile RTX 3080 to a full-blown desktop RTX 3080, just so that we can better illustrate the point. It’s packing 6,144 CUDA cores (instead of 8,704), has a boost block of 1,645Mhz (instead of 1,710Mhz), a 150W TGP (instead of 320W), 256-bit memory interface (instead of 320-bit), 448 Gb/s memory bandwidth (instead of 760 Gb/s), and slower memory speed at 14 Gbps instead of 19 Gbps. The only upgrade is in the memory department (16GB vs 10GB) but it’s of the slower GDDR6 variety (instead of GDDR6X).

Still, even with these drawbacks, it’s impossible not to stand in awe at what Asus managed to pull off. This incredibly small eGPU packs a ton of power inside a bafflingly small footprint. The XG Mobile comes equipped with a built-in 280W power supply (enough to drive both the mobile RTX 3080 and the Flow X13), it’s just over an inch thick, weighs in at just 2.2 pounds, and comes with a slew of ports including a full-size HDMI, DisplayPort, LAN, SD card reader, and four USB 3.0 ports as well for any peripherals or external hard drives/SSDs. Not too shabby!

If you’ve got the money to burn, then this sort of set-up might be the way to go. Then again, there are a few things you need to keep in mind. 

The Drawbacks

Let’s say the XG Mobile paired with the ROG Flow X13 tickles your fancy. Before you run our and buy the whole thing — given that you can find it in stock, of course — you need to understand that this is a brand-new kind of set-up. It was built by Asus for Asus products only. You can’t just connect this eGPU to any laptop of your choice and are wholly dependent on Asus providing support and releasing updated models further down the line.

Right now, at the start of its life cycle, Asus are promising long-term support, but what else are they supposed to say? They’re surely not going to call this an expensive experiment and hope for the best. Google did a similar thing with Stadia and we all know how that whole thing turned out.

Simply put, for Asus to continue supporting the whole XG Mobile ecosystem, it’ll have to move the needle and sell in large enough quantities for Asus to justify further development. If they can only sell five of these then you can be darn sure there won’t be another XG Mobile for a long time, if ever. So, in that sense, there are many factors outside of your control.

A $1200 eGPU that works with one laptop and one laptop only probably won’t sell volumes. Asus are supposedly looking to implement this proprietary solution into their upcoming laptops as well but there’s still no information as to when they might be released or how much they’ll retail for. There are too many intangibles right now, and there’s a very real chance they might abandon this whole project altogether in a year or two’s time.

Alienware (a subsidiary of Dell) also tried a similar thing with their proprietary Graphics Amplifier port and it never caught on. Turns out that pairing an expensive niche set-up with an equally expensive gaming laptop isn’t a recipe for success. Who would’ve thought?

What Are Your Options?

The answer to this question isn’t clear-cut — it really depends on your personal preference, the kind of set-up you need, how much you’re willing to invest, and how much power you seek. That’s a long list of questions, but they’re all equally important.

If a mobile RTX 3080 can provide you with the power you need, then the XG Mobile is definitely the way to go as it will provide you with noticeably better performance (in a smaller and more portable package) than any eGPU set-up in the past.

Even though the Asus ROG Flow X13 isn’t exactly a great deal (at least not when paired with its proprietary eGPU), it has nonetheless given us a glimpse at the future!


Leave a Reply