Qualcomm revealed X Elite's benchmark scores

[jz0309]

View attachment 2301539

Qualcomm Unveils Snapdragon X Elite CPU PC Benchmarks: Oryon Core Faster & Efficient Than Intel 13th Gen & Apple M2 Max, GPU Faster Than AMD RDNA 3

Qualcomm has unveiled first benchmarks of its Snapdragon X Elite CPU for upcoming Windows PCs which will tackle Apple, Intel, and AMD.

wccftech.com

View attachment 2301538

Maybe Apple Silicon is in a danger?

WHY?
first of all, this ne chip needs to be released by someone, most likely one of the existing OEMs, then it will most likely run WinARM. While Qualcomm keeps referencing Apple Silicon, it is x86 that would be in "danger" ...

 

[Darren.h]

They better start making Macs repairable and expandable otherwise he Mac is going o be slaughtered and put ou to pasture.

Why would you buy an ARM Mac that is no repairable or expandable and locked down when you can get the same FAST ARM hardware in the Windows World that is reparable by the user and upgradable ???

Who wants to invest in an expensive paper weight when it breaks? With no upgrade paths??

Here Come the ARM Hackintosh's on SnapDragon.

 

[MayaUser]

They better start making Macs repairable and expandable otherwise he Mac is going o be slaughtered and put ou to pasture.

Why would you buy an ARM Mac that is no repairable or expandable and locked down when you can get the same FAST ARM hardware in the Windows World that is reparable by the user and upgradable ???

Who wants to invest in an expensive paper weight when it breaks? With no upgrade paths??

We did have the same thing with "why would you buy an Intel Mac when you have the same fast Intel in Windows World that is reparable

 

[Love-hate 🍏 relationship]

wow these are great news !

 

[leman]

I wanted to ask you...with this Apple thing of scaling, from core design from A17 to the M family....how hard is for Apple to try to let or modify the clock freq based on the thermals that goes in it. For example it is possible that we get 4ghz or over for an M3 Max that goes strictly to the 16" Mbp or Mac Studio?

That's what I hope, yes. If you recall the thread we had on A17 power consumption and whatnot, I think there is some reason to speculate that it might be designed with higher operating frequencies in mind. That, or Apple design team has reached an efficiency barrier that are unable to cross.

Since we saw that QC can have turbo boost, can Apple besides lower the clock based on the heat...can make the cores run at higher clock speed than A17 Pro cores ?

M3 family will most certainly run clocks higher than A17 Pro. I can't really imagine them using the same clocks as on the A-series chip, they didn't do it with either M1 or M2...

Do you even think we can get M3 that is at least 30% improvement over M2?

I think 20-30% is what can be expected, yes, but I don't think it's a watertight prediction.

They better start making Macs repairable and expandable otherwise he Mac is going o be slaughtered and put ou to pasture.

Why would you buy an ARM Mac that is no repairable or expandable and locked down when you can get the same FAST ARM hardware in the Windows World that is reparable by the user and upgradable ???

Why would you think that Qualcomm's platform will be any more repairable? They also use on-package LPDDR RAM, which means no RAM upgrade or replacement. The SOC itself won't be replaceable either. The SSD will be, sure.

But frankly, if you really wanted replaceable SSD, you wouldn't buy a Mac since 2016.

Here Come the ARM Hackintosh's on SnapDragon.

That will probably not happen. MacOS relies on Apple Silicon specific features. You'd need to write a virtualisation layer and pretty much sacrifice a lot of performance to run macOS on a Snapdragon platoform.

 

[deconstruct60]

the most gaming profits are coming from arm...mobile platform.
Too bad Microsoft doesnt pull the trigger and like with the bs electric law...Microsoft to say to all OEM that from 2030 will be arm based only.

Microsoft doesn't have that kind of 'power'. Look at the backlash they got when said dump 2016-17 era processors for Windows 11 and that was just staying on x86-64 ( trying to retire old 32-bit boot/BIOS legacy stuff , more modern virtualization , and tighten up security. It isn't anywhere close to a platform change mandate. )

It has slowed Windows 11 adoption rate , but probably worth it long to clear out some 'baggage'. But windows is a " legacy baggage is largely welcomed " OS. That is a substantive part of what they sell.

What MS may have to backtrack off of slightly is the 'free forever' upgrade stuff. Likely going to need Windows to generate some more money to support the broader ecosystem over the long term. It is not a high growth market where can run a quasi ponzi scheme to have new customers pay for really old customers upgrades.

This is the only way to push and force Intel and Amd to come up with these kind of SoC. But legacy legacy and lack of "courage"

Intel and AMD are already responding without some dracoian threats. What is lined up for 2024-25 is substantively better than what Intel was offering 2-5 years ago. There is more competition now than in last 10 years with just AMD ; let alone throw in the new players sniffing around WinTel.

If Microsoft can stop the Windows market from relatively shrinking, it is a large enough ecosystem to have two instruction set options each have a big enough pool to have a stable subsystem to host them. Would have stable long term competition. In part, the act of declaring one instruction set dead and forcing everyone onto one is lack of competition.

Apple going to only Apple silicon isn't going to get better pricing for end users over the long term. Apple Silicon has enough small semantic differences that Oryon systems aren't going to lead to 'hackintoshes'. And really likely is vice versa also. 'raw iron' boot of Windows likely not coming to Macs either.

 

[camotwen]

Intriguing, but there have been so many claims of many parties beating apple silicon (in performance and efficiency combo) while in the end they were not even close, so I will wait to see independent benchmarking and testing to believe it. It would be amazing to be able to have (standard, non-experimental) linux systems on an apple silicon level chip.

 

[Analog Kid]

They better start making Macs repairable and expandable otherwise he Mac is going o be slaughtered and put ou to pasture.

Why would you buy an ARM Mac that is no repairable or expandable and locked down when you can get the same FAST ARM hardware in the Windows World that is reparable by the user and upgradable ???

Who wants to invest in an expensive paper weight when it breaks? With no upgrade paths??

Here Come the ARM Hackintosh's on SnapDragon.

Why did people buy Macs when they used identical processors to those PCs? Because they are better machines for those people’s needs. This doesn’t change that.

 

[camotwen]

Why did people buy Macs when they used identical processors to those PCs? Because they are better machines for those people’s needs. This doesn’t change that.

Actually I do not really get it. I see the older macbooks of my colleagues. They were more expensive and worse in performance than similar spec'ed professional lenovo laptops for instance. Making incredible noise and running extremely hot moreover, hanging and crawling too easily. Some spec choices were like several years behind, like 2019 iMacs having still hard drives instead of SSDs (wtf really - and we still suffer with plenty of such at work). Honestly, before apple silicon, due to my experience of all these Macs at work I just thought that apple computers were in general overly expensive crap.

I understand the OS advantage to some degree in professional setting (windows is crap and needs to be burnt to the ground), but honestly I really do not get why people were buying macbooks before apple silicon. Even now, if I had a chance to get a computer system at least on apple silicon level with less money and more repairability/upgradability/modability, I would not see any reason not to choose that over a mac.

 

[Analog Kid]

Actually I do not really get it. I see the older macbooks of my colleagues. They were more expensive and worse in performance than similar spec'ed professional lenovo laptops for instance. Making incredible noise and running extremely hot moreover, hanging and crawling too easily. Some spec choices were like several years behind, like 2019 iMacs having still hard drives instead of SSDs (wtf really - and we still suffer with plenty of such at work). Honestly, before apple silicon, due to my experience of all these Macs at work I just thought that apple computers were in general overly expensive crap.

I understand the OS advantage to some degree in professional setting (windows is crap and needs to be burnt to the ground), but honestly I really do not get why people were buying macbooks before apple silicon. Even now, if I had a chance to get a computer system at least on apple silicon level with less money and more repairability/upgradability/modability, I would not see any reason not to choose that over a mac.

Precisely the opposite of my experience, demonstrating why there is more than one viable option in the market.

 

[camotwen]

Precisely the opposite of my experience, demonstrating why there is more than one viable option in the market.

I cannot think of any viable use of an internal HDD in 2019 as main drive to load applications from. Heck, I cannot even think of one in 2016 even already.

But in any case, what was your experience exactly? What non-OS-specific aspect do you think mac machines provided that was better than an equivalent, professional level non-mac machine?

 

[Analog Kid]

I cannot think of any viable use of an internal HDD in 2019 as main drive to load applications from. Heck, I cannot even think of one in 2016 even already.

But in any case, what was your experience exactly? What non-OS-specific aspect do you think mac machines provided that was better than an equivalent, professional level non-mac machine?

Edit: nevermind, answering this is just going to take the thread further off topic so I've deleted my response. There's plenty of commentary on why people prefer Macs in these forums. You can find it there.

 

[nvmls]

What's Qualcomm's OS name again? Oh, they haven't got one yet?

Game over.

 

[camotwen]

Each to their own, though the things that you actually mention (driver installation stuff) are about the OS rather than the hardware, where the only argument is that it lasts long as it is - which I agree, it is in general of good quality.

 

[camotwen]

Dude why

What's Qualcomm's OS name again? Oh they haven't got one yet? Game over then.

Why would one care when one can install either windows or any linux variation? If they make their own and force people to use it that would basically be a negative, not positive.

 

[IG88]

have used laptops with 850, 7c, and 8cx. And all were awfully slow and buggy. GPU drivers are broken for Windows.

Do we know if the cause was hardware being meh, or Windows ARM being meh?

 

[Juraj22]

guys at wccftech are smoking something good.
"The Qualcomm Snapdragon X Elite delivers a single core score of 3227 points which is 14% faster than the Apple M2 Max (2841 points), while consuming 30% less power"

This certainly is NOT what Qualcomm was presenting.

Snapdragon X delivers 2841 points while consuming 30% less power. When it does produce 3227 it consumes more than M2 Max, but they didn't say by how much, obviously.

 

[camotwen]

guys at wccftech are smoking something good.
"The Qualcomm Snapdragon X Elite delivers a single core score of 3227 points which is 14% faster than the Apple M2 Max (2841 points), while consuming 30% less power"

This certainly is NOT what Qualcomm was presenting.

Snapdragon X delivers 2841 points while consuming 30% less power. When it does produce 3227 it consumes more than M2 Max, but they didn't say by how much, obviously.

This has always been the catch when others claim they have beaten apple silicon. Heck, it was also the catch when apple was presenting the M2s vs M1s. Until the system gets into real production and tested by others, it may all quite probably be number fabrication. Although for it to be successful it will probably be enough to just be on the 90% even of apple silicon level, as they do not necessarily compete on the same field. Making a viable laptop alternative for windows/linux is enough because right now there is none.

 

[nvmls]

Dude why

Why would one care when one can install either windows or any linux variation? If they make their own and force people to use it that would basically be a negative, not positive.

And yet here you are, pointing it out in an Apple centric forum.

 

[anshuvorty]

Is not a first product..they just changed their name for a more dramatic "looks". But im glad after 3 some kind failures they keep pushing the world of Windows to arm
View attachment 2301582

They keep trying to push WOA, but because for some reason, Microsoft can't create a competent emulator like Rosetta 2 and emulate x86_64 apps to help ease the transition from x86_64-based apps to ARM-based apps, WOA will never catch on, IMHO....you can't brute force it. You have to do what Apple did and create an emulator like Rosetta 2 to allow enterprises and users to adopt WOA, while still allowing them to use their legacy x86_64 apps while the devs work on updating their apps to ARM. Then, when most of the major apps have been updated to ARM, you can kill off the emulator.

Until then, even if the chipset is competent in performance against AS, if the software support isn't there, WOA market share won't go up.

My 2 cents....

 

[anshuvorty]

Speaking of Rosetta 2, how did Apple create an emulator like Rosetta 2 that was basically perfect in emulating all x86_64 apps, while Microsoft can't create the same for Windows? What did Apple do differently that allowed them to universally translate nearly all x86_64 apps and why can't Microsoft just copy Apple and create basically a "Rosetta 2 for Windows" emulator for WOA that emulates nearly all x86_64 Windows apps on WOA?

 

[leman]

Speaking of Rosetta 2, how did Apple create an emulator like Rosetta 2 that was basically perfect in emulating all x86_64 apps, while Microsoft can't create the same for Windows? What did Apple do differently that allowed them to universally translate nearly all x86_64 apps and why can't Microsoft just copy Apple and create basically a "Rosetta 2 for Windows" emulator for WOA that emulates nearly all x86_64 Windows apps on WOA?

Apple CPUs have hardware support for certain x86 features that would be otherwise costly to emulate with ARM. In short, there is a difference in how ARM and x86 synchronize data between multiple processor cores. To correctly emulate x86 behavior, every memory read and write on ARM would need to be wrapped in some costly additional instructions. Apple solves it by actually supporting the x86 behavior in hardware, via a CPU switch. It does cost some performance but not as much as software emulation.

 

[anshuvorty]

Apple CPUs have hardware support for certain x86 features that would be otherwise costly to emulate with ARM. In short, there is a difference in how ARM and x86 synchronize data between multiple processor cores. To correctly emulate x86 behavior, every memory read and write on ARM would need to be wrapped in some costly additional instructions. Apple solves it by actually supporting the x86 behavior in hardware, via a CPU switch. It does cost some performance but not as much as software emulation.

Wow, that's an insightful point. So, if I understand correctly, the lack of hardware support for x86_64 emulation on "experimental" WOA devices by the OEM PC makers is the main reason why the emulation is so poor on WOA. And, if the OEM PC makers put more effort into including this hardware support, then the x86_64 emulation would be almost as perfect as it is on Rosetta 2 on AS devices, right?

 

[deconstruct60]

They keep trying to push WOA, but because for some reason, Microsoft can't create a competent emulator like Rosetta 2 and emulate x86_64 apps to help ease the transition from x86_64-based apps to ARM-based apps,

Windows on Arm has a transcompiler that is largely similar to Rosetta 2. There a couple of differences but not real huge ones. MS has bent over backwards to keep more compatibility around. Apple threw a signficant number of apps out the windows and didn't try to work with them.

That major gap has been what the translated binaries actually run on. The new target processor largely has to have a speed increase over the old system to toss at translation 'overhead'. For example, if the M-series was 10% faster and the overhead is 2% then get a net 8% increase. If overhead is 4% still get a smaller net increase. In a contrasting example , if the Q-series is 10% slower and over is 2% then the get a 12% decrease. If the overhes is 4% then have a 14% decrease.

Going to get widely different end users reactions to those to even the overhead for translation is about exactly the same.

Early on the Qualcomm processor sales pitch was almost 100% about longer battery life, always-on celluar internet . Parity with the x86 models, let alone an increase, wasn't even on the table.

The over major stumbling block is Microsoft spend more than several years focusing on 32-bit only apps. Which technically is another Arm instruction set that Apple dumped as soon as they could get away with. And yet Miscrosoft plowed away for years trying to make that work better. ( somewhat because of all the 'sunk cost' legacy work they had put in on Windows Phone on Qualcomm and Window RT. )

Windows didn't come up with a preview for x86-64 compatiblity until 2020. The didn't 'clean up' windows to be 64-bit only kernel until Windows 11.




WoWA recompiles the code also.

"...
The WOW64 layer of Windows allows x86 code to run on the Arm64 version of Windows. x86 emulation works by compiling blocks of x86 instructions into Arm64 instructions with optimizations to improve performance. A service caches these translated blocks of code to reduce the overhead of instruction translation and allow for optimization when the code runs again. The caches are produced for each module so that other apps can make use of them on first launch. ..."
https://learn.microsoft.com/en-us/windows/arm/apps-on-arm-x86-emulation

Apple pushes almost all of the compile into a one time delay when first try to run the app. They stuff a whole complete copying into a not directly observable folder on the disk. So over time what you are doing is duplicating you app footprint on disk (making copies). Microsoft is taking a more circumscribed approach to just not blatantly making wholesale persistent copies.


Similarly Windows allows mixed binaries ( for better or worse and also relatively recently released ).

" ... Arm64EC (“Emulation Compatible”) enables you to build new native apps or incrementally transition existing x64 apps to take advantage of the native speed and performance possible with Arm-powered devices, including better power consumption, battery life, and accelerated AI & ML workloads. ..."

Arm64EC for Windows 11 apps on Arm

Learn how Arm64EC empowers you to build and incrementally update apps that benefit from native performance on Arm devices, without interrupting your current x64 functionality.

learn.microsoft.com

Apple is more of a rip-the-band-aid-off-quickly. Either transition the whole app including all possible plug-ins or don't get native. Also 'tough love' ... critical 32-bit app you needed.. too bad; dead. Need AVX .. 'no soup for you!!!' That isn't a question of competency , but rather one of different approach.

Pretty good chance Microsoft did this because people asked for it. It would be less work for for MS to tell developers 'tough luck ... it is entirely on your time/dime '. Their user base is so large they probably do have end users who has some odd-ball plug-in that is critical but has no development support under it at all ( a zombie piece of code). Nobody is going do an update and the whole 'tough love... go make a new one' isn't going to work in that context. They can't.

WOA will never catch on, IMHO....you can't brute force it. You have to do what Apple did and create an emulator like Rosetta 2 to allow enterprises and users to adopt WOA, while still allowing them to use their legacy x86_64 apps while the devs work on updating their apps to ARM. Then, when most of the major apps have been updated to ARM, you can kill off the emulator.

Err that whole really gradual incremental progress is exactly what Arm64EC is for; and Apple does not have at all.

And again with an extensive amoutn of legacy applicate lingering around .... there are some apps with no devs doing anything on them. 'Retire the emulator' is likely not going away in Windows over a dozen years after Apple dumps theirs.

You can't dump the emulator if not completely dump x86-64 for all supported systems. If you have a Intel Mac it is actively one the Vintage/Obsolete 5 year countdown clock or already on the dumped list. That is NOT happening on Windows.


Apple added some non standard augments to Arm instruction set to 'goose' the translated Rosetta 2 code . Windows isn't looking to do just one Arm implementor. And Windows doesn't have special quirks stuffed into x86 ... so not likely particularly happing in Arm implementations either. Windows doesn't really need a 1990's Intel to get looped into a single implementor orbit with. Windows is looking for a ecosystem.

Apple doesn't want any vendor ecosystem at all. That is two very different solution paths.

 

[deconstruct60]

And, if the OEM PC makers put more effort into including this hardware support, then the x86_64 emulation would be almost as perfect as it is on Rosetta 2 on AS devices, right?

OEM PC makers don't make processors. OEM buy processors. What proposing is that several Arm implementors ( or Arm itself) get into the business of adding these augments. Arm itself has much larger pressing issues to address. The implementors are all in very similar boats.

It would be far better to just give Windows developers very good tools and development sytsems/kits so that they can just do a straightforward deliberate port and just avoid the transcribe anyway. ( compilers will put in guards in some places for the automagically if use the right high level constructs. ).

Apple is putting in effort in part because they are actively burning the x86_64 behind their users. They are tossing mlliions of x86_64 onto the obsolete list as fast as they can. You don't have a choice to buy a new Mac with Intel and in 5 or so years probably won't get a macOS for one either. That hardware feature is in part so don't notice the roaring burning bridge behind you.