The Extreme model is cheaper than I thought it would be. In fact, why would you even get the vanilla version when the price difference is only $100 between them?
From a Marketing perspective, you sometimes position one product to make the other look like a better deal and push people towards that. I suspect that is happening here. Though given that their margins are likely pretty small, it may just be how things landed (at least to some extent).
Yeah, I understand that concept of Good-Better-Best, or whatever it is called. Apple does it on all their products to try and get you to not buy their lowest end models.
What I don't understand is why the "premium" Z1 Extreme Ally is not priced to match. I would have figured at least $200 price difference between the two models.
Even the Steam deck goes from $399 for the base model, to $529 for the better model, and has a $649 option for the best model.
Arguably, the low end Ally is priced a bit high compared to the steam deck.
The Z1 Extreme is the most extremely bandwidth/power/thermally limited device I have ever seen - so the value proposition of 3x (12 vs 4 CUs) isn't real in terms of performance ... I really wonder why Asus went with the 12CU version except for spec sheet bragging rights.
There's come comparisons out there now with Steam Deck and this thing is looking dissapointing imo (eg https://arstechnica.com/gadgets/2023/05/the-asus-r... ) to be honest (except when RTX on when the newer hardware leaps ahead)
Clearly 15W/30W "performance"/"turbo" is not enough for the Z1 Extreme despite it being 4nm .. looks like it needs more like 45W+ to actually perform well.
[more] in 15W performance mode it's getting outperformed by the Steam Deck (also 15W) in every game they tested (same ArsTechnica source) - that takes some doing considering they are on 4nm vs 7nm
I can only assume the 120Hz and bigger LCD screen is using a lot more power, but it looks bad, will wait for more reviews for now (maybe some launch issues ?)
Kind reminder that RDNA3 drivers are still on a very early stage where sometimes you have better fps on the RDNA2 equivalent. AMD does have a bad rep (deserved) of launching with very non-optimized drivers, but fixing them later on, so... let's see what happens.
ASUS reputation is currently at the bottom where even hardware reviewers (who rely one sponsorship) are kicking the brand out. They need to fix their reputation first after the shady way they are trying to get out of Warrenty for their faulty hardware with the AMD fiasco.
If AMD was truly designing a chip for portable consoles, they would do something similar to the steam deck. 8 core CPU with lots of cache is a waste of die space and power for negligible gain in this scenario. And possibly that's why steam deck is so good at low TDP configs. But sadly AMD is doing what makes the most profit - rebranding/repackaging laptop chips.
Assuming that you're serious, it's not a module, driver, or syscall. It's the whole architecture and design philosophy.
Linux has historically been more efficient when it comes to storing data on your HDD/SSD (no idea if that's true with win11). Linux FS-es don't need defragmenting. Linux doesn't try and run any background tasks without asking you first. Linux doesn't need to download huge service packs to update the system. Linux doesn't auto-update (unless you ask it to), so you can game or whatever for as long as you like. Linux allows you to compile code for your arch, instead of a generic one, thus saving your time and power while running apps.
I could go on, but I think the point has been made.
I'm afraid I have to disagree. I'm an ASIC designer hence I have a lot of experience with Linux since all of my tools only run on Linux. And sometimes I had to run minimum Linux on some RISC cores that's completely compiled from source.
EXT4 does not solve the fragmentation issue, windows does trim and so is Linux. Linux also runs a bunch of background tasks without asking you first, as it assume you agree them to run when installing the system. One example would be gnome's tracker-miner file indexing process. The update of Linux system is much more complicated than windows. You need kernel update, the individual library update, and the application update. Depends on which distro you choose there's also possible huge system upgrade to download. The auto-update situation is indeed a lot better on Linux, but I'd argue it doesn't affect efficiency. As for the last bit, assuming you are talking about the kernel itself, there's no such thing as "your arch" in x86 world. Both windows and Linux will produce a specific kernel for a specific ISA. RISC is problematic since everyone does their things differently, but that's irrelevant here.
There's no inherent efficiency difference between Windows and Linux. You may achieve higher efficiency on Linux if you know what you are doing. The design philosophy of Linux isn't efficiency, it's "do you know what you are doing?". But, do you?
I disagree with you for the same specific reasons that @erinadreno does. But you are correct more broadly, because it is the whole architecture and design philosophy. On Linux, you have a lot more flexibility to customise it than you have on Windows. If you want to use ext4, you use it; if you want to use btrfs, you use it; and if you want to roll your own custom FS for fun, you can do that too. If you want to replace the scheduler, or rip out systemd, and so on, you can do all that too. This flexibility from lowest level (kernel) to highest level makes it possible to do things that are simply unthinkable on a Windows system. One of those things is targeting extreme efficiency.
There are things that are still very difficult to do, of course, such as removing sysfs or otherwise rearchitecting the driver architecture. This is because a lot of the ecosystem relies on these things. But that does not make them impossible, if you are willing to throw enough time and money at the "problem" to make what will effectively be a fork of Linux. The point is that these things would simply be impossible on Windows, and are possible-but-difficult on Linux. And as SteamOS demonstrates, you don't really need to go to that level to get most of the benefits anyway.
"EXT4 does not solve the fragmentation issue..." Well, I've filled my EXT4 FS to +99% and had only 5 fragmented files on my system. I've done this more than once with near equal results. It was 14 files the previous time, IIRC.
"...windows does trim and so is Linux." That's not related to fragmentation. Maybe it was a side comment?
"Linux also runs a bunch of background tasks without asking you first,"... Well, sure, the 1-click install options make many assumptions about how you want the system to run. But that's for maximum simplicity (newbies), not for efficiency or performance, or anything else.
Regarding upgrades, with most distros, you just use synaptic/apt or yast/zypper/yum. It might be more complicated, at the lower level, but for regular users, it's pretty simple. If you want something which you have more control over, Gentoo gives you USE flags and stuff to customize what packages you want and what they're capable of. I was saying that it's more efficient to download a few MB of files to unzip and upgrade, than to download many GB for a service pack.
"Your arch" was referring to the specific extensions that the CPU you use supports. For example, compiling for X86 means that you don't use AVX. Likewise with some SSE instructions IIRC, but I'm going back a few years. These advanced instructions allow you to achieve significant speed-ups, thus "racing to halt" at a higher rate.
"There's no inherent efficiency difference between Windows and Linux. You may achieve higher efficiency on Linux if you know what you are doing. The design philosophy of Linux isn't efficiency, it's "do you know what you are doing?". But, do you?"
Maybe you've not read about this, but time and time again while I was reading about devs who wrote code for linux I noticed that they would write the code to be as small and as fast as they could because (it appears) that they had a poor PC (or were just performance and/or efficiency freaks). Off the top of my head I can think of ctwm, and mpeg123. Here's a quote from mpg123's Readme.txt file:
Now mpg123 catched up with MPlayer's mp3lib concerning decoding speed on my Pentium M (which supports SSE): Decoding a certain album (Queensryche's Rage for Order) to /dev/null took 22.4s user time with mpg123-0.66 compared to 24.7 with MPlayer-1.0rc1 . Also, beginning with mpg123 1.8.0, there are fresh x86-64 SSE optimizations (provided by Taihei Monma) which make mpg123 the fastest MPEG audio decoder in my knowledge also on current 64bit x86 systems.
There are findings of a smaller "Phoenix2" SoC in Linux drivers. It reportedly has a 2x Zen4 + 4x Zen4c CPU, 4x RDNA3 WGPs (8CUs) and I read somewhere that is supports the faster LPDDR5X 8533MT/s.
The Phoenix on the Z1 / 7x40U was designed for laptops so it isn't all that interesting for handhelds, but this Phoenix2 should be the actual successor of Van Gogh.
It could also be that a bunch of laptop brands (Dell, MSI, Gigabyte, Lenovo) are waiting for Phoenix2 to release their handheld.
AMD hasn't had anything below 6c since Zen 2 so I wouldn't expect to see lower core count until they move to hybrid architecture. I'd have been happy with a high end IGP and a 6 core chip for this, but the Z1 appears to be a power optimized mobile chip rather than being fully new silicon and the 12CU IGP are limited to 8c processors.
Mendocino is based on a quad-core die, so it's actually a step in the right direction (5125C uses an 8-core die with 6 cores disabled). But it took a 75% cut in RDNA2 CUs instead of being a carbon copy of Steam Deck's 8 CU Van Gogh, and it's overpriced IMO. In the long run, they could spam out Mendocino as a 6nm budget chip for years to come, and it could end up in the $80-150 laptops. It has decent performance for the low-end, and AV1 decode.
Its missing a magnetic or clamp latch keyboard to put on the bottom of it that possibly could include a touchpad. If it had something like that and could be left connected and maybe closed up via hinge that could help give if flexibility and utility. It's already running Windows 11 so its not far away from being useful for things like Word and Excel on the go. I'd imagine it could either use a kickstand to support the rear of the machine or, even better, an additional battery in the keyboard to add portability.
The battery will die after 1-2 years, I bet it is glued to the chassis like Steam Deck. And even more bonus if it's like that L shape which makes it even harder to remove without damaging the Lithium Ion nonsense.
Heatsink Repaste DIY won't be easy, because it's so small the tolerances are too tight and if it has poor garbage design like Steam Deck or similar it will be same.
The SoC is soldered. A PC with BGA hardware with high usage = high heat = no longevity.
Software is a mess. It's ASUS, they have worthless bloatware in it, from that ArmouryCrate UWP Malware which you cannot remove unless you do a full OS wipe. Then the Windows 11 which is a crippling OS that has a ton of Microsoft Telemetry in it. It needs Linux and with Linux half of those ASUS specific features won't work, plus it needs a proper Kernel for the Linux, I doubt ASUS does it, their ASUS phones have BL unlock long time back not sure how ASUS handles it.
Overpriced.
All these issues make all portables which are Currently Produced make them poor products, I'd rather take a Nintendo 3DS XL and mod it and play it over that Switch with garbage SoC / Joycon drift / horrid 3 battery (2 joycon and the unit) using sealed packs of Li Ion unlike 3DS of modular design.
But since masses will consume it like a smartphone, the market will be riddled with the disposable garbage. Rather spend that money on an AM4 budget box and enjoy longevity, why would you want to play AAA in a pathetic small screen with console controls when you have a nice Desktop PC at your home with Keyboard and Mouse without any battery baggage BS.
I would like a keyboard and mouse to be integrated and I agree that a removable battery would also be quite nice. It is also priced a bit too high, but I don't think, even though I use Linux as my primary OS, that Windows 11 itself is that bad in terms of product capability. Its a good operating system for the masses though I would prefer if it didn't have an online activation requirement. That is mainly an unfortunate response to software piracy and to Google and Apple's means of monetization through data mining for which I mainly blame Google for leading the charge on...not that Microsoft wouldn't have eventually pulled that cat out of the bag on their own sooner or later.
Don't forget that SilverSurfer said all that which may or may not persuade the reader only to conclude with wonder at why someone wouldn't just use a stationary PC at a desk or at best while sitting on a couch instead of a Steam Deck or ROG Ally.
The conclusion renders all the points relevant. People want the portable gaming PC because it is PORTABLE.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
35 Comments
Back to Article
meacupla - Friday, May 12, 2023 - link
The Extreme model is cheaper than I thought it would be. In fact, why would you even get the vanilla version when the price difference is only $100 between them?ingwe - Friday, May 12, 2023 - link
From a Marketing perspective, you sometimes position one product to make the other look like a better deal and push people towards that. I suspect that is happening here. Though given that their margins are likely pretty small, it may just be how things landed (at least to some extent).meacupla - Friday, May 12, 2023 - link
Yeah, I understand that concept of Good-Better-Best, or whatever it is called. Apple does it on all their products to try and get you to not buy their lowest end models.What I don't understand is why the "premium" Z1 Extreme Ally is not priced to match. I would have figured at least $200 price difference between the two models.
Even the Steam deck goes from $399 for the base model, to $529 for the better model, and has a $649 option for the best model.
Arguably, the low end Ally is priced a bit high compared to the steam deck.
xol - Friday, May 12, 2023 - link
The Z1 Extreme is the most extremely bandwidth/power/thermally limited device I have ever seen - so the value proposition of 3x (12 vs 4 CUs) isn't real in terms of performance ... I really wonder why Asus went with the 12CU version except for spec sheet bragging rights.Source - AMD's own publicity material ie https://community.amd.com/t5/gaming/the-all-new-am... specifically 2nd image down : it never even gets 2x performance, barely manages +50% in most cases
There's come comparisons out there now with Steam Deck and this thing is looking dissapointing imo (eg https://arstechnica.com/gadgets/2023/05/the-asus-r... ) to be honest (except when RTX on when the newer hardware leaps ahead)
Clearly 15W/30W "performance"/"turbo" is not enough for the Z1 Extreme despite it being 4nm .. looks like it needs more like 45W+ to actually perform well.
skavi - Sunday, May 14, 2023 - link
going wider and slower generally gets you better efficiency (ISO performance)xol - Friday, May 12, 2023 - link
[more] in 15W performance mode it's getting outperformed by the Steam Deck (also 15W) in every game they tested (same ArsTechnica source) - that takes some doing considering they are on 4nm vs 7nmI can only assume the 120Hz and bigger LCD screen is using a lot more power, but it looks bad, will wait for more reviews for now (maybe some launch issues ?)
heffeque - Friday, May 12, 2023 - link
Kind reminder that RDNA3 drivers are still on a very early stage where sometimes you have better fps on the RDNA2 equivalent.AMD does have a bad rep (deserved) of launching with very non-optimized drivers, but fixing them later on, so... let's see what happens.
s.yu - Monday, May 15, 2023 - link
Well, they already got the bad press. They probably lost half the target segment with the first round of reviews(like Notebookcheck).sharath.naik - Saturday, May 13, 2023 - link
ASUS reputation is currently at the bottom where even hardware reviewers (who rely one sponsorship) are kicking the brand out. They need to fix their reputation first after the shady way they are trying to get out of Warrenty for their faulty hardware with the AMD fiasco.Samus - Sunday, May 14, 2023 - link
Asus reputation is current at the bottom? Seriously?s.yu - Monday, May 15, 2023 - link
Me neither.erinadreno - Friday, May 12, 2023 - link
If AMD was truly designing a chip for portable consoles, they would do something similar to the steam deck. 8 core CPU with lots of cache is a waste of die space and power for negligible gain in this scenario. And possibly that's why steam deck is so good at low TDP configs. But sadly AMD is doing what makes the most profit - rebranding/repackaging laptop chips.NeuralNexus - Saturday, May 13, 2023 - link
Steam Deck is power efficient because it is running on Linux plain and simple.erinadreno - Saturday, May 13, 2023 - link
Could you point out which kernel module/driver/syscall etc made Linux more efficient?Threska - Saturday, May 13, 2023 - link
The LRDF (Linux Reality Distortion Field).ballsystemlord - Monday, May 15, 2023 - link
Assuming that you're serious, it's not a module, driver, or syscall. It's the whole architecture and design philosophy.Linux has historically been more efficient when it comes to storing data on your HDD/SSD (no idea if that's true with win11).
Linux FS-es don't need defragmenting.
Linux doesn't try and run any background tasks without asking you first.
Linux doesn't need to download huge service packs to update the system.
Linux doesn't auto-update (unless you ask it to), so you can game or whatever for as long as you like.
Linux allows you to compile code for your arch, instead of a generic one, thus saving your time and power while running apps.
I could go on, but I think the point has been made.
erinadreno - Monday, May 15, 2023 - link
I'm afraid I have to disagree. I'm an ASIC designer hence I have a lot of experience with Linux since all of my tools only run on Linux. And sometimes I had to run minimum Linux on some RISC cores that's completely compiled from source.EXT4 does not solve the fragmentation issue, windows does trim and so is Linux.
Linux also runs a bunch of background tasks without asking you first, as it assume you agree them to run when installing the system. One example would be gnome's tracker-miner file indexing process.
The update of Linux system is much more complicated than windows. You need kernel update, the individual library update, and the application update. Depends on which distro you choose there's also possible huge system upgrade to download.
The auto-update situation is indeed a lot better on Linux, but I'd argue it doesn't affect efficiency.
As for the last bit, assuming you are talking about the kernel itself, there's no such thing as "your arch" in x86 world. Both windows and Linux will produce a specific kernel for a specific ISA. RISC is problematic since everyone does their things differently, but that's irrelevant here.
There's no inherent efficiency difference between Windows and Linux. You may achieve higher efficiency on Linux if you know what you are doing. The design philosophy of Linux isn't efficiency, it's "do you know what you are doing?". But, do you?
Carmen00 - Tuesday, May 16, 2023 - link
I disagree with you for the same specific reasons that @erinadreno does. But you are correct more broadly, because it is the whole architecture and design philosophy. On Linux, you have a lot more flexibility to customise it than you have on Windows. If you want to use ext4, you use it; if you want to use btrfs, you use it; and if you want to roll your own custom FS for fun, you can do that too. If you want to replace the scheduler, or rip out systemd, and so on, you can do all that too. This flexibility from lowest level (kernel) to highest level makes it possible to do things that are simply unthinkable on a Windows system. One of those things is targeting extreme efficiency.There are things that are still very difficult to do, of course, such as removing sysfs or otherwise rearchitecting the driver architecture. This is because a lot of the ecosystem relies on these things. But that does not make them impossible, if you are willing to throw enough time and money at the "problem" to make what will effectively be a fork of Linux. The point is that these things would simply be impossible on Windows, and are possible-but-difficult on Linux. And as SteamOS demonstrates, you don't really need to go to that level to get most of the benefits anyway.
ballsystemlord - Tuesday, May 16, 2023 - link
EDIT: I should have originally written, "your ISA", not "your arch".ballsystemlord - Tuesday, May 16, 2023 - link
"EXT4 does not solve the fragmentation issue..." Well, I've filled my EXT4 FS to +99% and had only 5 fragmented files on my system. I've done this more than once with near equal results. It was 14 files the previous time, IIRC."...windows does trim and so is Linux." That's not related to fragmentation. Maybe it was a side comment?
"Linux also runs a bunch of background tasks without asking you first,"... Well, sure, the 1-click install options make many assumptions about how you want the system to run. But that's for maximum simplicity (newbies), not for efficiency or performance, or anything else.
Regarding upgrades, with most distros, you just use synaptic/apt or yast/zypper/yum. It might be more complicated, at the lower level, but for regular users, it's pretty simple. If you want something which you have more control over, Gentoo gives you USE flags and stuff to customize what packages you want and what they're capable of.
I was saying that it's more efficient to download a few MB of files to unzip and upgrade, than to download many GB for a service pack.
"Your arch" was referring to the specific extensions that the CPU you use supports. For example, compiling for X86 means that you don't use AVX. Likewise with some SSE instructions IIRC, but I'm going back a few years. These advanced instructions allow you to achieve significant speed-ups, thus "racing to halt" at a higher rate.
"There's no inherent efficiency difference between Windows and Linux. You may achieve higher efficiency on Linux if you know what you are doing. The design philosophy of Linux isn't efficiency, it's "do you know what you are doing?". But, do you?"
Maybe you've not read about this, but time and time again while I was reading about devs who wrote code for linux I noticed that they would write the code to be as small and as fast as they could because (it appears) that they had a poor PC (or were just performance and/or efficiency freaks). Off the top of my head I can think of ctwm, and mpeg123. Here's a quote from mpg123's Readme.txt file:
Now mpg123 catched up with MPlayer's mp3lib concerning decoding speed on my Pentium M (which supports SSE): Decoding a certain album (Queensryche's Rage for Order) to /dev/null took 22.4s user time with mpg123-0.66 compared to 24.7 with MPlayer-1.0rc1 .
Also, beginning with mpg123 1.8.0, there are fresh x86-64 SSE optimizations (provided by Taihei Monma) which make mpg123 the fastest MPEG audio decoder in my knowledge also on current 64bit x86 systems.
xol - Saturday, May 13, 2023 - link
If it's going to play Series S/X / PS5 targetted games at 720p then I would expect it would need the 8 cores...That's my assumption anyway
StevoLincolnite - Tuesday, May 16, 2023 - link
It's an incorrect assumption as no game on Xbox Series S/X and the Playstation 5 uses 8 cores.And they will never use all 8 cores.
ToTTenTranz - Saturday, May 13, 2023 - link
There are findings of a smaller "Phoenix2" SoC in Linux drivers. It reportedly has a 2x Zen4 + 4x Zen4c CPU, 4x RDNA3 WGPs (8CUs) and I read somewhere that is supports the faster LPDDR5X 8533MT/s.The Phoenix on the Z1 / 7x40U was designed for laptops so it isn't all that interesting for handhelds, but this Phoenix2 should be the actual successor of Van Gogh.
It could also be that a bunch of laptop brands (Dell, MSI, Gigabyte, Lenovo) are waiting for Phoenix2 to release their handheld.
StevoLincolnite - Tuesday, May 16, 2023 - link
Considering these parts are mostly bandwidth limited anyway, that would be a massive boon.I would rather a smaller more efficient SOC that can boost to high clocks when needed that is backed by oodles of bandwidth than the opposite.
nandnandnand - Saturday, May 13, 2023 - link
8-cores are tiny, getting even smaller with each node, and are what you need to keep up with the big consoles.thestryker - Saturday, May 13, 2023 - link
AMD hasn't had anything below 6c since Zen 2 so I wouldn't expect to see lower core count until they move to hybrid architecture. I'd have been happy with a high end IGP and a 6 core chip for this, but the Z1 appears to be a power optimized mobile chip rather than being fully new silicon and the 12CU IGP are limited to 8c processors.xol - Sunday, May 14, 2023 - link
5300G says hinandnandnand - Sunday, May 14, 2023 - link
There's a reason it's OEM-only. 8-core dies and chiplets should almost never be turned into quad-core products.BTW, if you want an even more absurd example, look no further than the Ryzen 3 5125C.
meacupla - Sunday, May 14, 2023 - link
Ryzen 3 7320U is also anemic, despite having LPDDR5nandnandnand - Sunday, May 14, 2023 - link
Mendocino is based on a quad-core die, so it's actually a step in the right direction (5125C uses an 8-core die with 6 cores disabled). But it took a 75% cut in RDNA2 CUs instead of being a carbon copy of Steam Deck's 8 CU Van Gogh, and it's overpriced IMO. In the long run, they could spam out Mendocino as a 6nm budget chip for years to come, and it could end up in the $80-150 laptops. It has decent performance for the low-end, and AV1 decode.PeachNCream - Sunday, May 14, 2023 - link
Its missing a magnetic or clamp latch keyboard to put on the bottom of it that possibly could include a touchpad. If it had something like that and could be left connected and maybe closed up via hinge that could help give if flexibility and utility. It's already running Windows 11 so its not far away from being useful for things like Word and Excel on the go. I'd imagine it could either use a kickstand to support the rear of the machine or, even better, an additional battery in the keyboard to add portability.Silver5urfer - Monday, May 15, 2023 - link
It's a garbage device for longevity like a PC.The battery will die after 1-2 years, I bet it is glued to the chassis like Steam Deck. And even more bonus if it's like that L shape which makes it even harder to remove without damaging the Lithium Ion nonsense.
Heatsink Repaste DIY won't be easy, because it's so small the tolerances are too tight and if it has poor garbage design like Steam Deck or similar it will be same.
The SoC is soldered. A PC with BGA hardware with high usage = high heat = no longevity.
Software is a mess. It's ASUS, they have worthless bloatware in it, from that ArmouryCrate UWP Malware which you cannot remove unless you do a full OS wipe. Then the Windows 11 which is a crippling OS that has a ton of Microsoft Telemetry in it. It needs Linux and with Linux half of those ASUS specific features won't work, plus it needs a proper Kernel for the Linux, I doubt ASUS does it, their ASUS phones have BL unlock long time back not sure how ASUS handles it.
Overpriced.
All these issues make all portables which are Currently Produced make them poor products, I'd rather take a Nintendo 3DS XL and mod it and play it over that Switch with garbage SoC / Joycon drift / horrid 3 battery (2 joycon and the unit) using sealed packs of Li Ion unlike 3DS of modular design.
But since masses will consume it like a smartphone, the market will be riddled with the disposable garbage. Rather spend that money on an AM4 budget box and enjoy longevity, why would you want to play AAA in a pathetic small screen with console controls when you have a nice Desktop PC at your home with Keyboard and Mouse without any battery baggage BS.
PeachNCream - Tuesday, May 16, 2023 - link
I would like a keyboard and mouse to be integrated and I agree that a removable battery would also be quite nice. It is also priced a bit too high, but I don't think, even though I use Linux as my primary OS, that Windows 11 itself is that bad in terms of product capability. Its a good operating system for the masses though I would prefer if it didn't have an online activation requirement. That is mainly an unfortunate response to software piracy and to Google and Apple's means of monetization through data mining for which I mainly blame Google for leading the charge on...not that Microsoft wouldn't have eventually pulled that cat out of the bag on their own sooner or later.foodprocessingunit - Thursday, May 18, 2023 - link
Don't forget that SilverSurfer said all that which may or may not persuade the reader only to conclude with wonder at why someone wouldn't just use a stationary PC at a desk or at best while sitting on a couch instead of a Steam Deck or ROG Ally.The conclusion renders all the points relevant. People want the portable gaming PC because it is PORTABLE.
pbollwerk - Monday, May 22, 2023 - link
Lack of Hall Effect sticks is a deal breaker for me.