The tech I’m most psyched to see this year is anything but traditional PC silicon, and certainly isn’t x86
Dave James, once saw a laptop
This week I have been mostly: Playing around with one of the most ridiculous handhelds… which I still can’t help but like, and losing myself gazing lovingly at the Razer Blade 16’s glorious OLED screen.
I’m a PC gamer, a tech obsessive, and probably a bit of a contrarian, too. Which is maybe why I’m always keen to support the underdog story when it comes to a new player entering the PC gaming arena. I mean, it may be a billion dollar underdog, but when it comes to proper Windows PC silicon, Qualcomm is still a n00b, and I can’t wait to see how its new laptop CPUs finally shape up.
There are a few different reasons why Qualcomm’s new Snapdragon X Elite chips are the slices of silicon I’m most interested in getting my hands on this year. Mostly, it’s because we’re talking about a complete departure from the x86 domination of Intel and AMD over the PC processing space from a Windows perspective. And there’s a lot of inherent potential in that, especially having seen what Apple did with its own chips when it sloughed off its Intel chip obligations.
But there is also a creeping negativity around what the so-called ‘big three’ are up to this year when it comes to hardware releases in PC gaming land. Sure, Nvidia could start to ship its RTX 50-series Blackwell GPUs this year, but even should we end up with the most powerful graphics card ever, it will only be at the hyper-expensive end of the market. That’s only going to leave PC gamers with either empty bank accounts or a deep crushing sense of envy.
AMD could be shipping new RDNA 4 cards this year, too, but they’re reportedly going to be chasing the mid-range and are unlikely to be meaningfully changing it as a price category. Only if it can genuinely, drastically change the performance we expect from that tier of GPU will the Radeon RX 8000-series have our collective attention.
Then there’s the potential for AMD Zen 5 and Intel Arrow and Lunar Lake CPUs, but who cares about boring ol’ x86 silicon when Qualcomm’s new Snapdragon X Elite (and X Plus) chips are going to give us the low power, slimline laptops of our dreams?
This is the promise of the new ARM-based processors, using either 12 or 10 Qualcomm Oryon CPU cores, with an Adreno GPU capable of up to 4.6 TFLOPs of compute power inside them, the chips can can scale from completely fanless 12 W machines to 80 W configs. And they are supposedly super-efficient, delivering either the same performance as AMD’s 780M for way less power or higher performance in the same envelope.
| Platform | Snapdragon X Elite | Snapdragon X Elite | Snapdragon X Elite | Snapdragon X Plus |
|---|---|---|---|---|
| Part number | X1E-84-100 | X1E-80-100 | X1E-78-100 | X1P-64-100 |
| Cores | 12 | 12 | 12 | 10 |
| Total cache | 42 MB | 42 MB | 42 MB | 42 MB |
| Max multi-core frequency | 3.8 GHz | 3.4 GHz | 3.4 GHz | 3.4GHz |
| Dual Core Boost | 4.2 GHz | 4.0 GHz | N/A | N/A |
| GPU TFLOPs | 4.6 | 3.8 | 3.8 | 3.8 |
| NPU TOPS | 45 | 45 | 45 | 45 |
| Storage | NVMe PCIe 4.0 | NVMe PCIe 4.0 | NVMe PCIe 4.0 | NVMe PCIe 4.0 |
| Memory | LPDDR5x | LPDDR5x | LPDDR5x | LPDDR5x |
| Transfer rate | 8448 MT/s | 8448 MT/s | 8448 MT/s | 8448 MT/s |
| Release date | Mid-2024 | Mid-2024 | Mid-2024 | Mid-2024 |
And this isn’t just going to be for boring office apps either, because Qualcomm has—needlessly, in my opinion—been talking up its gaming performance in the Windows world. Remember, we’re talking about an ARM-based chip, running an operating system where most of the software is designed for an x86 instruction set and we’re told that games will “just work.”
I’m sure it’s not as simple as that, and I am absolutely expecting times where the performance is slammed to an unplayable point simply because the drivers and game code are absolutely not making friends with one another. But, if Valve’s Proton has taught me anything, it’s that sometimes things that intuitively shouldn’t actually will “just work.”
So yes, I am absolutely pumped about the potential for a sleek 13-inch laptop, with an all-day battery life, and the processing and gaming performance that will cater for all my reasonable laptop requests. If I can get the same frame rates as something like the iGPUs inside the latest handhelds and AMD’s top laptop chips, but with twice the battery life, then I am going to be a happy man.
But I am old enough and ugly enough to know that we can’t have nice things. I’ve been burned often enough that I know there’s a good chance the X Elite chips will be an absolute mess for anyone wanting to actually game on those machines. The silicon itself may well be fantastic, but Intel will tell you that it’s one thing to have effective hardware and another thing entirely to make sure you’re able to regularly squeeze that performance out across the rainbow spectrum of PC gaming good-times.
Just look at the patchy performance of Intel’s Arc GPUs—even after a ton of patches—across a broad cross-section of PC games. Sometimes the frame rates are fantastic, but also there are times where the performance just isn’t there… or the game just flat doesn’t work.
So, I do still have a hard time trusting that Qualcomm is going to be able to nail both the drivers for its new hardware as well as manage the translation layers which allow it to run native x86 code on its 64-bit ARM cores. Especially when we have Charlie Demerjian specifically calling out the company for cheating in its apparently unreplicable Snapdragon X Elite benchmarks.
It’s true, Qualcomm hasn’t been particularly detailed about either the tech or the benchmarks with the press so far, and Charlie is calling BS on a lot of those presented numbers having spoken to OEMs working with the silicon, which is rather concerning.
One, because it sucks the promised performance may well not be there in the final reckoning, and two, because why the hell has Qualcomm dug itself such a hole if the claims aren’t true? At no point did anyone expect these ARM PC laptop chips to be targeting real PC gaming performance, so the company had no reason to start making claims it knows not to be true which can only harm its rep post release.
Until we actually get hands on laptops sporting the Snapdragon silicon it’s not possible for us to tell whether Charlie’s mistaken, or if Qualcomm is actually pulling numbers out of the air, but it has dampened my enthusiasm a touch, that’s for sure.
The only other piece of tech I want to see asap—and most especially want to see actually being good—is Intel’s Battlemage graphics architecture. But after the Alchemist launch debacle I am more than a little reticent about getting too excited by Intel’s upcoming GPUs. More than anything it’s about whether Intel can launch on time this year, but also whether it’s learned the lessons from Alchemist and can actually release a genuinely competitive suite of discrete graphics cards to give us a proper GPU third way.
Which I guess is one of the reasons I want Qualcomm to succeed. Maybe it’s part of my contrarian tendencies, where I want fresh new upstart alternatives in different sectors of PC gaming. Intel in graphics cards to upset the Nvidia (and to a much lesser extent) AMD hegemony, and Qualcomm in processors to disrupt the x86 monopoly.
And if not Qualcomm, then… who? Because with the potential for ARM-based chips running Windows’ x86 apps the market is suddenly wide open, and we know Nvidia has some history with Arm. Surely a ShieldBook is still a possibility?
