Qualcomm Continues to Disrupt the PC Industry with the Addition of Snapdragon X Plus Platform

[theouto]

That's a lot of smoke being sold, but who's buying?

 

[Dr. Dro]

I think their mobile objectives are at odds with whatever this PC push is supposed to be. Trying to impress compared to the PC incumbents would take some massive investment.

Adreno was built by ATI guys. They reorganized the letters in Radeon. I wonder if any of them are still over there.

I don't think that modern Adreno has anything to do with the original or the involvement of the people who created it... but interestingly enough, Qualcomm denies vehemently that Adreno was named after Radeon. We believe them. Right?

Reminiscing about the early days of Adreno somehow always reminds me of Imageon (which can be considered the precursor tech to Adreno, AMD would eventually sell this IP to Qualcomm) and Xilleon, another former ATI IP that AMD sold to Broadcom. While Adreno was always intended to be more of an embedded graphics IP, Xilleon was an SoC for broadcast-grade video decoding and decompression.

Relics of the past indeed.

Why and how will this benefit anyone?

It's not faster.
It's not more efficient.
It's not compatible with all software.
Windows Arm is a joke, and you have still have to pay for it.
It can't do much more than flash games, or games at their lowest settings at 720p.
It's not cheaper.
It's not upgradable.

Why does it exist?

The primary benefit of ARM is reduced cost and far superior battery life figures against x86 processors. Especially under Windows and its horrible power management.

Software compatibility is not a massive issue this time around as anything developed for Windows x64 should run unmodified thanks to a transparent emulation layer in Windows 11, works on the same principle as Apple's Rosetta. Performance might be a concern but it looks like it performs decently enough according to the preliminary testing on the internet.

Apple somehow had x86 with great battery life (custom system and power management controllers and a superior OS written for the hardware it ran on), but their ARM chips are something else in this department. 20h+ battery life is something achievable in their Apple Silicon MacBooks.

 

[stimpy88]

The primary benefit of ARM is reduced cost and far superior battery life figures against x86 processors. Especially under Windows and its horrible power management.

Software compatibility is not a massive issue this time around as anything developed for Windows x64 should run unmodified thanks to a transparent emulation layer in Windows 11, works on the same principle as Apple's Rosetta. Performance might be a concern but it looks like it performs decently enough according to the preliminary testing on the internet.

Apple somehow had x86 with great battery life (custom system and power management controllers and a superior OS written for the hardware it ran on), but their ARM chips are something else in this department. 20h+ battery life is something achievable in their Apple Silicon MacBooks.

So we have the only argument that "may" hold up... Battery life... Maybe... But the rest is very much in the "proof is in the pudding" camp, isn't it. And again, is this going to enable performant sub $100 tablets and notebooks running a copy of Windows and Office? Or are the prices going to be the same as everyone else in the low end? Also, Qualcomm will not guarantee software compatibility, but admit to doing "some" testing.

Again, what is this going to "actually" achieve in the market that is different to the current offerings? What is "disruptive" about this in the PC market?

 

[wNotyarD]

"Disrupting the PC industry" is such a bold claim it's really laughable when the product falls short of the hype by a long shot.

I mean, we're used to labels "Industry Leader" from no-name startups, but Qualcomm's PR should know better.

It will disrupt the industry alright, everybody will stop for a moment and have a laugh at it.

 

[Bwaze]

Not the disruption we need but a disruption we deserve.

 

[Chrispy_]

Better wait for third party reviews:
Qualcomm Is Cheating On Their Snapdragon X Elite/Pro Benchmarks - SemiAccurate

Aww maaaannn....

I would love ARM to gain traction in the x86 monopoly but this isn't the news I wanted to hear. I'll wait for actual tests on real-world hardware and software, but this sure doesn't look very good at all.

 

[Noyand]

So we have the only argument that "may" hold up... Battery life... Maybe... But the rest is very much in the "proof is in the pudding" camp, isn't it. And again, is this going to enable performant sub $100 tablets and notebooks running a copy of Windows and Office? Or are the prices going to be the same as everyone else in the low end? Also, Qualcomm will not guarantee software compatibility, but admit to doing "some" testing.

Again, what is this going to "actually" achieve in the market that is different to the current offerings? What is "disruptive" about this in the PC market?

Software compatibility isn't in the hands of QUALCOMM, but windows and mostly the devs.

Sub 100$ computers is not really a market where anyone is interested to make waves in... there's always a catch when something is that cheap, and not just about performance. Just look at the cheapest chromebook, even if they manage to make a SOC that's somehow fast and absurdly cheap to manufacture, there's other stuff that's going to make this a pain to use in the long term. It's a lot of R&D to spend on something that will have razor-thin margin and might not even sell enough to compensate.

The "disruption" is really all about the claimed performance per watt. The potential to get a full day of work done without having to plug your laptop once. Performance on battery is something that windows laptops are struggling with. It would be fair to say that they are rubbish when they are unplugged. But that's assuming that Qualcomm didn't lie about how the chip actually perform.

 

[Darmok N Jalad]

Software compatibility isn't in the hands of QUALCOMM, but windows and mostly the devs.

Sub 100$ computers is not really a market where anyone is interested to make waves in... there's always a catch when something is that cheap, and not just about performance. Just look at the cheapest chromebook, even if they manage to make a SOC that's somehow fast and absurdly cheap to manufacture, there's other stuff that's going to make this a pain to use in the long term. It's a lot of R&D to spend on something that will have razor-thin margin and might not even sell enough to compensate.

The "disruption" is really all about the claimed performance per watt. The potential to get a full day of work done without having to plug your laptop once. Performance on battery is something that windows laptops are struggling with. It would be fair to say that they are rubbish when they are unplugged. But that's assuming that Qualcomm didn't lie about how the chip actually perform.
View attachment 345065View attachment 345067View attachment 345069

Then slap a battery life graph on there while running those benches. I have an i7 1265U in my work laptop, and the two lousy P cores means a lot of waiting in Excel. No way that something with 2 performance cores should be called an i7. The E cores simply don’t make up the difference.

 

[stimpy88]

Software compatibility isn't in the hands of QUALCOMM, but windows and mostly the devs.

Sub 100$ computers is not really a market where anyone is interested to make waves in... there's always a catch when something is that cheap, and not just about performance. Just look at the cheapest chromebook, even if they manage to make a SOC that's somehow fast and absurdly cheap to manufacture, there's other stuff that's going to make this a pain to use in the long term. It's a lot of R&D to spend on something that will have razor-thin margin and might not even sell enough to compensate.

The "disruption" is really all about the claimed performance per watt. The potential to get a full day of work done without having to plug your laptop once. Performance on battery is something that windows laptops are struggling with. It would be fair to say that they are rubbish when they are unplugged. But that's assuming that Qualcomm didn't lie about how the chip actually perform.
View attachment 345065View attachment 345067View attachment 345069

Your post is as useful as this product is, and reads like corporate PR.

Read your answers and then go back to my original questions... How is this going to disrupt the PC market? What does this product achieve/bring to the table that others do not?

 

[Noyand]

Your post is as useful as this product is, and reads like corporate PR.

Read your answers and then go back to my original questions... How is this going to disrupt the PC market? What does this product achieve/bring to the table that others do not?

I already said why it could potentially do it...but you choose to be a troll instead. Current X86 PC laptops don't perform that well unplugged; the loss of performance is massive, they are not even close to a MacBook in that regard, they must lose a lot of performance to not drain the battery too fast. That's simply the most realistic disruption that you can expect. I'm sorry if you find it disappointing, but it is what it is. And again, that's assuming that Qualcomm didn't feed the press a bunch of manipulated data.

Also, Qualcomm marketing has been clear that they're going for high performance, past news also revealed that those Qualcomm laptops are going to be quite expensive. It was already established that they aren't interested at all in the low-end market. It's meant to be used in premium laptops, and they've been clear about that since the beginning. (But they might spectacularly fail if the rumours are true)

And again, is this going to enable performant sub $100 tablets and notebooks running a copy of Windows and Office?

So, if you followed the news, you would have already known the answer to that question. (Yes, that question is the one that caught my attention the most. The 100$ segment is the segment that's always been the most neglected, if you weren't making a joke and expect a billionaire corporation to care and make waves in that segment, then you are in for a world of pain )


My"corporate PR" is just how the market currently works. Haven't you noticed that the low-end CPU/GPU market has never been as bad as it is today? The low end is either a massively cut down version of old silicon, or a CPU full of e-cores. And according to Darmok, that's an awful experience to have on Windows. The low-end on Windows is basically a dead market. Even the Surface go who was supposed to be a cheap sub-400$ PC ended up being a rather mediocre 699$ computer.

Even beyond the PC, there's a tendency right now in tech to focus on the high-end and midrange, the low-end is generally unexciting beyond a few niches Chinese stuff. Big corporation tends to do massive price cut on old products and they're done. Here's your low end.

The fact that you didn't see my post as useful, is ultimately down to the reality that there's really nothing mind-blowing to expect beside better performance for watt. Unlike QC I'm trying to be real here, and don't oversell the AI stuff. TSMC and ASML are not really trying to enable cheaper computers, regular windows will never be a lean OS, and people absolutely hate any cut down version of windows, which could run on weak silicon. Expecting a 100$ Windows device to run well is akin to expecting Nintendo to release their games on PC. It won't happen, and they aren't interested to make it happen either.

Then slap a battery life graph on there while running those benches. I have an i7 1265U in my work laptop, and the two lousy P cores means a lot of waiting in Excel. No way that something with 2 performance cores should be called an i7. The E cores simply don’t make up the difference.

notebookcheck load test is a full GPU + CPU benchmark running in a loop until the battery dies with the screen at max brightness. They don't test performance on battery, sadly. It's pretty much in the same ballpark as many windows laptop without losing any performance. The biggest outlier is that zenbook, but we don't know how it performs. So in the best case, a PC laptop can last twice as long, but according to puget system, it is also often thrice as slow at executing a heavy task. So I'm not sure with poison is the worst here, when the MacBook is otherwise beating anything of the same performance class when websurfing or during video playback.

I don't own a MacBook btw, my laptop is a PC. And beyond some light photo editing/2D graphics, it just can't keep up unplugged. If I want to render something, I have to go home or find a plug, otherwise it'll be so slow, the battery will take a hit anyway. I, for one, don't think that the PC makers need to be pampered on that aspect. They must figure out a way to keep more performance unplugged.

On another note, how come unplugged performance benchmark of laptops are surprisingly hard to find?

 

[Darmok N Jalad]

I already said why it could potentially do it...but you choose to be a troll instead. Current X86 PC laptops don't perform that well unplugged; the loss of performance is massive, they are not even close to a MacBook in that regard, they must lose a lot of performance to not drain the battery too fast. That's simply the most realistic disruption that you can expect. I'm sorry if you find it disappointing, but it is what it is. And again, that's assuming that Qualcomm didn't feed the press a bunch of manipulated data.

Also, Qualcomm marketing has been clear that they're going for high performance, past news also revealed that those Qualcomm laptops are going to be quite expensive. It was already established that they aren't interested at all in the low-end market. It's meant to be used in premium laptops, and they've been clear about that since the beginning. (But they might spectacularly fail if the rumours are true)

So, if you followed the news, you would have already known the answer to that question. (Yes, that question is the one that caught my attention the most. The 100$ segment is the segment that's always been the most neglected, if you weren't making a joke and expect a billionaire corporation to care and make waves in that segment, then you are in for a world of pain )


My"corporate PR" is just how the market currently works. Haven't you noticed that the low-end CPU/GPU market has never been as bad as it is today? The low end is either a massively cut down version of old silicon, or a CPU full of e-cores. And according to Darmok, that's an awful experience to have on Windows. The low-end on Windows is basically a dead market. Even the Surface go who was supposed to be a cheap sub-400$ PC ended up being a rather mediocre 699$ computer.

Even beyond the PC, there's a tendency right now in tech to focus on the high-end and midrange, the low-end is generally unexciting beyond a few niches Chinese stuff. Big corporation tends to do massive price cut on old products and they're done. Here's your low end.

The fact that you didn't see my post as useful, is ultimately down to the reality that there's really nothing mind-blowing to expect beside better performance for watt. Unlike QC I'm trying to be real here, and don't oversell the AI stuff. TSMC and ASML are not really trying to enable cheaper computers, regular windows will never be a lean OS, and people absolutely hate any cut down version of windows, which could run on weak silicon. Expecting a 100$ Windows device to run well is akin to expecting Nintendo to release their games on PC. It won't happen, and they aren't interested to make it happen either.


notebookcheck load test is a full GPU + CPU benchmark running in a loop until the battery dies with the screen at max brightness. They don't test performance on battery, sadly. It's pretty much in the same ballpark as many windows laptop without losing any performance. The biggest outlier is that zenbook, but we don't know how it performs. So in the best case, a PC laptop can last twice as long, but according to puget system, it is also often thrice as slow at executing a heavy task. So I'm not sure with poison is the worst here, when the MacBook is otherwise beating anything of the same performance class when websurfing or during video playback.

I don't own a MacBook btw, my laptop is a PC. And beyond some light photo editing/2D graphics, it just can't keep up unplugged. If I want to render something, I have to go home or find a plug, otherwise it'll be so slow, the battery will take a hit anyway. I, for one, don't think that the PC makers need to be pampered on that aspect. They must figure out a way to keep more performance unplugged.
View attachment 345259

On another note, how come unplugged performance benchmark of laptops are surprisingly hard to find?

I suspect they don't test unplugged performance on x86 laptops because they lose so much performance, and OEMs probably don't want reviewers advertising that.

According to my battery stats, I last charged my 15" MBA to 100% 6 days ago. I use it about an hour a day for surfing, emails, texts, and some photo editing, though last Sunday I streamed about 4 hours of video. Currently sitting at 35% battery remaining, so I can usually go about a week under average use with needing to plug in. I've actually tested this laptop plugged and unplugged, and yeah, performance is the same. One run of the latest Cinebench will knock about 5% off the battery. My work laptop (HP Z book) by comparison is a joke. I had a 3 hour meeting today showing BPI through Teams, and that sucked my battery down to 30%, and I wasn't even pushing brightness. And performance on the Z book is just not all that great, IMO.