"Unlike the Snapdragon 8c connectivity, the entry-level Snapdragon 7c comes integrated with Qualcomm's Snapdragon X15 LTE modem with support for 800Mbps 3x20 DL and 150Mbps 2x20MHz UL LTE.
You also get Qualcomm Wi-Fi Solutions with standard (11ax) ready 2.4GHz/5GHz dual-band Wi-Fi, Bluetooth 5.1, and GPS."
Fudzilla says 802.11ad and Bluetooth 5.0 for the 8c, and 802.11ax and Bluetooth 5.1 for the 7c. So not a typo.
Any normal binary would have a regular x64 codepath anyway - otherwise it would break compatibility with the 90+% of x64 processors that don't implement AVX-512.
AVX2 is not commonly available either (which Atom supports it?), the fallback would be an SSE(N) codepath, which the simulator does handle.
In any case if you believe that vector performance matters here then you're quite wrong. The vast majority of x86 applications don't ever use vector instructions (being compiled with plain -O2 or even -Os settings).
No currently shipping consumer ARM has SVE. The overwhelming majority of currently shipping x86 lacks AVX512. Any sane x86 program has fallbacks to SSE4 or older, and Win/ARM64 presently offers binary translation for x86 but not x64 (which is rumored to come next year.)
Performance of emulated x86 binaries is quick enough for programs that don't have a lot of heavy compute to be very usable. For those that are more processor intensive, the story is obviously different. Adobe is going to be doing a native ARM64 port of Creative Cloud, which may be the single largest pain point.
If you were coding for AVX-512 on anything outside of a benchmark, you don’t have a fallback. You coded AVX-512 for a specific scenario that it enhances greatly it’s in so few CPUs you did it for a reason. If you are just looking for an uptick in vectorized performance you code AVX2 with a fallback to AVX.
These really aren't the devices you'd be running AVX specific applications on. If that's a significant use case for you, there are much better options.
Any sane developer would create a fallback, if anything to validate the AVX code (because its far easier to write/validate non-SIMD code, and then compare the SIMD against that). And most will also have a AVX2 version of their AVX512 code, just to broaden the usability.
The vast majority of developers don't write in assembly and are at the mercy of the compiler author (often Microsoft, Intel or the GNU compiler authors). AVX512 instructions are very limited in applicable usage and as far as I know, you need to actively choose to use it.
Yes the Microsoft compiler won't emit any AVX instructions by default, otherwise x86 applications won't run on Atom.
This whole discussion is insane, nobody is using their laptop as a render server. The use case for high performance SIMD in consumer products used mostly for browsing is simply non-existent.
Is the 7c really just 8x “little” cores like this chart implies? If so that’s going to be absolutely miserable to use on Windows, especially with only half the memory bandwidth of a single-channel x86 chip. They’re going to have to do a lot better than just “sub-$800” to make these worth it; you can get Ryzen 5 U or Core i5 U machines easily in that range.
I believe you are correct. However, this 2BIG+6little core design should be reflected in the Table, because it does make a difference. Not Anton's doing, either, it's QC's screwy way of naming their cores. And yes, 8 little cores only would be badly underpowered for Windows on Arm; even Intel's slowest current-gen Atoms run circles around A55 cores.
I'm a bit surprised (or more likely naive to the issues related with the conversion of a Snapdragon to an "Xc" SoC process) to find out that the 7c is not ready. I imagine it's an issue with software / driver compatibility as the hardware changes from a 7xx series must be fairly minimal; hobbyists on twitter have even booted Windows on Arm onto phones.
It's probably cheaper to re-purpose existing cellular parts than to validate a new design with those portions explicitly cut out. That said, having a modem is a killer feature as 4G access is generally better than WiFi at a lot of public venues. Having access to a DSP for image processing purposes is useful for popular apps like Snapchat or video conferencing.
I wonder if Qualcomm uses these internally instead of x86 laptops. Otherwise no way I can see other companies using them. The pricing here is terrible, off by a factor of 3X
Smart move diversifying from the $1000 Windows on ARM segment that moves effectively zero units down to the $800 segment that moves effectively zero units.
According to their interview on this site, they move ~1,000 units of WARM parts a day; that's not much, but a good start. Given that the Hololens 2 runs WARM because no x86 part has the features at the power envelope and that they're doing an Android phone as well (which I suspect will have some kind of capability to Virtualize an instance Windows), Microsoft seems serious about WARM in the long run.
Microsoft would love to get rid of Legacy x86 applications, but the real problem is business has applications written on x86 and it hard to replace machines when you dealing with 1000's of them.
I think it long run they want to be more like Apple and that is why they doing, cloning of PC has done wonders for making Windows popular - but it hurt the bottom line at Microsoft.
My big concern is that people will be mislead by these machines thinking that it can run applications and when it does not they will be upset.
What is odd I would think this type of processor would be better for Android or Chrome market than Windows. a
Yep very true. It is pretty clear MS wants everyone to switch to the Windows store because they make money from every App sold that has an up front cost. People not conforming to their ways and want to use programs bought or downloaded from any where else does not make them any money. My best guess is MS thinks if they get the ball rolling with these Arm setups that at some point they can try to force more and more people over to their MS store because lets face it they know as long as x86 is the king of the play ground most people will not switch over to their MS store because x86 has such a huge software base and there is really no need for people having to rely on MS and their store for most if not all software.
MS knows this and they are going to try their best to make x86 a thing of the past as fast as they can. Right now it is these small steps in that direction but as time moves forward they will of coarse get more bold and make bigger moves in that direction. This has always been the way MS does things to herd people in the direction they want us all to go. If they could get away with things like Apple can they would have just done it already and forced the change over I also hear rumors of Apple maybe dropping Intel in favor of them making their own desktop CPU which of coarse would not be based off of x86 at all but more inline with what they use in their mobile hardware. I am not sure how true that is but we all know how Apple likes to screw over their hardware partners and go their own way.
Would be way better than the chip in there currently. It's technically a "core" architecture, but without Turbo boost it performs worse than some of the *Monts now.
Even at 10 times less that that they would not sell in any significant quantities (especially after returns when people realize that almost nothing works).
Personally I am not finding myself very excited about these Arm CPU's getting to the desktop and Microsoft going out of their way to try to make windows work with them. At the end of the day if you use your PC with programs not from the Microsoft store which rely on x86 the performance is going to be so low that most of us at least me would want to through the PC out the window and go get something else that has a non emulated x86 CPU inside of it.
I know they have to start some where and probably about 5-10 years down the road after these PC's have gone through a few versions of hardware and most software has moved to non x86 they will become more useful to more people. Until then to me at least these are nothing more than little play toys that are good for grandma's daily internet surfing on Facebook and other minor tasks. Just my opinion on the subject nothing more nothing less.
Given the Cortex-A76 in the 8cx has similar IPC as Skylake and is about twice as fast as the best Atom, emulated performance is more than fast enough - better than what you'd get from a cheap x86 laptop.
"emulated performance is more than fast enough" Uh huh. Keep telling yourself that. The rest of us look at the laughable real world performance of WinArm device and just laugh.
It'll be a LONG time before ARM is the main CPU for the windows world.
A LONG LONG time. Think, they'll have to emulate a 5.0GHz 9900KS or similar! Have you ever thought what that would be like? They can't require companies to release 2 binaries, 1 for x86 and 1 for ARM. Even if they could, all the SW that used to work will be broken.
Apple already has a phone which outperforms your precious 9900KS... Like it or not, but Arm is rapidly moving up in performance, so software will be recompiled when it runs faster on Arm.
Actually, the 9900KS is Intel's. I'm not bragging about it. More to the point. I've never seen a non-synthetic benchmark that apple's iphone has won and Intel's 9900KS has lost. But if you've found such a benchmark please don't hesitate to reply. Yes, I do like that Arm's processors are moving up in performance. I'm not criticizing a performance increase, Arm, AMD, or Intel. I'm criticizing the notion that emulation works, let alone works at a decent speed. I speak from experience of windows on Linux non-direct-x game emulation. I can't out perform a Pentium 2!
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.
53 Comments
Back to Article
fred666 - Friday, December 6, 2019 - link
only Wifi 5 on the higher ends units? Or typo?nandnandnand - Friday, December 6, 2019 - link
https://www.fudzilla.com/news/mobile/49931-qualcom..."Unlike the Snapdragon 8c connectivity, the entry-level Snapdragon 7c comes integrated with Qualcomm's Snapdragon X15 LTE modem with support for 800Mbps 3x20 DL and 150Mbps 2x20MHz UL LTE.
You also get Qualcomm Wi-Fi Solutions with standard (11ax) ready 2.4GHz/5GHz dual-band Wi-Fi, Bluetooth 5.1, and GPS."
Fudzilla says 802.11ad and Bluetooth 5.0 for the 8c, and 802.11ax and Bluetooth 5.1 for the 7c. So not a typo.
nandnandnand - Friday, December 6, 2019 - link
Same info can also be found here: https://en.wikipedia.org/wiki/List_of_Qualcomm_Sna...Meteor2 - Sunday, December 8, 2019 - link
Which probably came from here 😂nandnandnand - Sunday, December 8, 2019 - link
Nope. 159. "Snapdragon 7c Compute Platform". Qualcomm. = https://www.qualcomm.com/products/snapdragon-7c-co...nandnandnand - Friday, December 6, 2019 - link
7c gets 802.11ax and 8c doesn't? Ok, then.Also, all of these SoC names suck.
brucethemoose - Saturday, December 7, 2019 - link
As do the CPU names. I think Qualcomm is going for maximum consumer confusion.geo774 - Saturday, December 7, 2019 - link
Its because when 8cx was announced..wifi 6 wasn't a thing..and 8c is its modified version while 7c is completely new designMeteor2 - Sunday, December 8, 2019 - link
Thispeevee - Friday, December 6, 2019 - link
How good is their x64+AVX512 to ARM64 v8.latest translator?x64+AVX is so complex now I doubt it is going to be very good. Probably no AVX512 at all, I would be surprised if AVX2->SVE is even there.
thetrashcanisfull - Friday, December 6, 2019 - link
Any normal binary would have a regular x64 codepath anyway - otherwise it would break compatibility with the 90+% of x64 processors that don't implement AVX-512.peevee - Monday, December 9, 2019 - link
But then the performance of this code on all computational tasks will be abysmal.Wilco1 - Monday, December 9, 2019 - link
Well that means 99.99% of the PC market then since practically PC even supports AVX512...peevee - Tuesday, December 10, 2019 - link
I mean regular x64 codepath vs at least AVX2.Wilco1 - Tuesday, December 10, 2019 - link
AVX2 is not commonly available either (which Atom supports it?), the fallback would be an SSE(N) codepath, which the simulator does handle.In any case if you believe that vector performance matters here then you're quite wrong. The vast majority of x86 applications don't ever use vector instructions (being compiled with plain -O2 or even -Os settings).
SarahKerrigan - Friday, December 6, 2019 - link
No currently shipping consumer ARM has SVE. The overwhelming majority of currently shipping x86 lacks AVX512. Any sane x86 program has fallbacks to SSE4 or older, and Win/ARM64 presently offers binary translation for x86 but not x64 (which is rumored to come next year.)Performance of emulated x86 binaries is quick enough for programs that don't have a lot of heavy compute to be very usable. For those that are more processor intensive, the story is obviously different. Adobe is going to be doing a native ARM64 port of Creative Cloud, which may be the single largest pain point.
AlyxSharkBite - Saturday, December 7, 2019 - link
If you were coding for AVX-512 on anything outside of a benchmark, you don’t have a fallback. You coded AVX-512 for a specific scenario that it enhances greatly it’s in so few CPUs you did it for a reason. If you are just looking for an uptick in vectorized performance you code AVX2 with a fallback to AVX.Reflex - Saturday, December 7, 2019 - link
These really aren't the devices you'd be running AVX specific applications on. If that's a significant use case for you, there are much better options.nevcairiel - Sunday, December 8, 2019 - link
Any sane developer would create a fallback, if anything to validate the AVX code (because its far easier to write/validate non-SIMD code, and then compare the SIMD against that). And most will also have a AVX2 version of their AVX512 code, just to broaden the usability.Flunk - Sunday, December 8, 2019 - link
The vast majority of developers don't write in assembly and are at the mercy of the compiler author (often Microsoft, Intel or the GNU compiler authors). AVX512 instructions are very limited in applicable usage and as far as I know, you need to actively choose to use it.Wilco1 - Tuesday, December 10, 2019 - link
Yes the Microsoft compiler won't emit any AVX instructions by default, otherwise x86 applications won't run on Atom.This whole discussion is insane, nobody is using their laptop as a render server. The use case for high performance SIMD in consumer products used mostly for browsing is simply non-existent.
lowlymarine - Friday, December 6, 2019 - link
Is the 7c really just 8x “little” cores like this chart implies? If so that’s going to be absolutely miserable to use on Windows, especially with only half the memory bandwidth of a single-channel x86 chip. They’re going to have to do a lot better than just “sub-$800” to make these worth it; you can get Ryzen 5 U or Core i5 U machines easily in that range.nandnandnand - Friday, December 6, 2019 - link
Look again. 8c is for sub-$800, while 7c is for sub-$400.I envision the 7c debuting at $200-$300, eventually dropping down to $100-$200 in sales. As for performance, that's what we have benchmarks for.
d8e8fca - Saturday, December 7, 2019 - link
2x Cortex-A76 @ 2.4 GHz and 6x Cortex-A55. So essentially a faster Snapdragon 675.eastcoast_pete - Saturday, December 7, 2019 - link
I believe you are correct. However, this 2BIG+6little core design should be reflected in the Table, because it does make a difference. Not Anton's doing, either, it's QC's screwy way of naming their cores. And yes, 8 little cores only would be badly underpowered for Windows on Arm; even Intel's slowest current-gen Atoms run circles around A55 cores.Raqia - Friday, December 6, 2019 - link
I'm a bit surprised (or more likely naive to the issues related with the conversion of a Snapdragon to an "Xc" SoC process) to find out that the 7c is not ready. I imagine it's an issue with software / driver compatibility as the hardware changes from a 7xx series must be fairly minimal; hobbyists on twitter have even booted Windows on Arm onto phones.Rudde - Saturday, December 7, 2019 - link
Microsoft made an official closed beta for snapdragon 820 equipped smartphones.eastcoast_pete - Saturday, December 7, 2019 - link
If you have a link for download, I'd be interested in it. I have an 820 device that might serve as my guinea pig for me to play with itzamroni - Friday, December 6, 2019 - link
the dsp and modem can be removed to resume price because most laptop use cases don't need themRaqia - Saturday, December 7, 2019 - link
It's probably cheaper to re-purpose existing cellular parts than to validate a new design with those portions explicitly cut out. That said, having a modem is a killer feature as 4G access is generally better than WiFi at a lot of public venues. Having access to a DSP for image processing purposes is useful for popular apps like Snapchat or video conferencing.eastcoast_pete - Saturday, December 7, 2019 - link
Agree; the key attraction of these for me is the "always connected" bit.nandnandnand - Saturday, December 7, 2019 - link
Sounds awful. I'll stick to Wi-Fi.geo774 - Saturday, December 7, 2019 - link
I wanna see 8cx in a gaming oriented tablet to compete with ipad..preferably with high refresh rate oledyeeeeman - Saturday, December 7, 2019 - link
These look more like sub 400$ chips, not sub 800$. Qualcomm is dreaming...webdoctors - Saturday, December 7, 2019 - link
I wonder if Qualcomm uses these internally instead of x86 laptops. Otherwise no way I can see other companies using them. The pricing here is terrible, off by a factor of 3Xnandnandnand - Saturday, December 7, 2019 - link
7c is for sub-$400. They say 8c is for sub-$800.For some values of sub.
Gunbuster - Saturday, December 7, 2019 - link
Smart move diversifying from the $1000 Windows on ARM segment that moves effectively zero units down to the $800 segment that moves effectively zero units.Raqia - Saturday, December 7, 2019 - link
According to their interview on this site, they move ~1,000 units of WARM parts a day; that's not much, but a good start. Given that the Hololens 2 runs WARM because no x86 part has the features at the power envelope and that they're doing an Android phone as well (which I suspect will have some kind of capability to Virtualize an instance Windows), Microsoft seems serious about WARM in the long run.HStewart - Sunday, December 8, 2019 - link
Microsoft would love to get rid of Legacy x86 applications, but the real problem is business has applications written on x86 and it hard to replace machines when you dealing with 1000's of them.I think it long run they want to be more like Apple and that is why they doing, cloning of PC has done wonders for making Windows popular - but it hurt the bottom line at Microsoft.
My big concern is that people will be mislead by these machines thinking that it can run applications and when it does not they will be upset.
What is odd I would think this type of processor would be better for Android or Chrome market than Windows. a
rocky12345 - Monday, December 9, 2019 - link
Yep very true. It is pretty clear MS wants everyone to switch to the Windows store because they make money from every App sold that has an up front cost. People not conforming to their ways and want to use programs bought or downloaded from any where else does not make them any money. My best guess is MS thinks if they get the ball rolling with these Arm setups that at some point they can try to force more and more people over to their MS store because lets face it they know as long as x86 is the king of the play ground most people will not switch over to their MS store because x86 has such a huge software base and there is really no need for people having to rely on MS and their store for most if not all software.MS knows this and they are going to try their best to make x86 a thing of the past as fast as they can. Right now it is these small steps in that direction but as time moves forward they will of coarse get more bold and make bigger moves in that direction. This has always been the way MS does things to herd people in the direction they want us all to go. If they could get away with things like Apple can they would have just done it already and forced the change over I also hear rumors of Apple maybe dropping Intel in favor of them making their own desktop CPU which of coarse would not be based off of x86 at all but more inline with what they use in their mobile hardware. I am not sure how true that is but we all know how Apple likes to screw over their hardware partners and go their own way.
tipoo - Saturday, December 7, 2019 - link
Surface Go 2, get a move on Microsoft.Would be way better than the chip in there currently. It's technically a "core" architecture, but without Turbo boost it performs worse than some of the *Monts now.
yannigr2 - Sunday, December 8, 2019 - link
I guess Windows on ARM was so fast on the 8cx, that these new SOCs are completely logical steps.nandnandnand - Sunday, December 8, 2019 - link
8cx is the best chip and should be used in $5,000 systems.peevee - Monday, December 9, 2019 - link
Even at 10 times less that that they would not sell in any significant quantities (especially after returns when people realize that almost nothing works).rocky12345 - Monday, December 9, 2019 - link
Personally I am not finding myself very excited about these Arm CPU's getting to the desktop and Microsoft going out of their way to try to make windows work with them. At the end of the day if you use your PC with programs not from the Microsoft store which rely on x86 the performance is going to be so low that most of us at least me would want to through the PC out the window and go get something else that has a non emulated x86 CPU inside of it.I know they have to start some where and probably about 5-10 years down the road after these PC's have gone through a few versions of hardware and most software has moved to non x86 they will become more useful to more people. Until then to me at least these are nothing more than little play toys that are good for grandma's daily internet surfing on Facebook and other minor tasks. Just my opinion on the subject nothing more nothing less.
Wilco1 - Monday, December 9, 2019 - link
Given the Cortex-A76 in the 8cx has similar IPC as Skylake and is about twice as fast as the best Atom, emulated performance is more than fast enough - better than what you'd get from a cheap x86 laptop.TheinsanegamerN - Monday, December 9, 2019 - link
"emulated performance is more than fast enough"Uh huh. Keep telling yourself that. The rest of us look at the laughable real world performance of WinArm device and just laugh.
It'll be a LONG time before ARM is the main CPU for the windows world.
Wilco1 - Tuesday, December 10, 2019 - link
How would you know? You don't sound like you've ever used a WOA device...If you accept that Atom devices can give reasonable performance, then so does this.
ballsystemlord - Monday, December 9, 2019 - link
A LONG LONG time. Think, they'll have to emulate a 5.0GHz 9900KS or similar! Have you ever thought what that would be like? They can't require companies to release 2 binaries, 1 for x86 and 1 for ARM.Even if they could, all the SW that used to work will be broken.
Wilco1 - Tuesday, December 10, 2019 - link
Apple already has a phone which outperforms your precious 9900KS... Like it or not, but Arm is rapidly moving up in performance, so software will be recompiled when it runs faster on Arm.ballsystemlord - Wednesday, December 11, 2019 - link
Actually, the 9900KS is Intel's. I'm not bragging about it.More to the point. I've never seen a non-synthetic benchmark that apple's iphone has won and Intel's 9900KS has lost. But if you've found such a benchmark please don't hesitate to reply.
Yes, I do like that Arm's processors are moving up in performance. I'm not criticizing a performance increase, Arm, AMD, or Intel. I'm criticizing the notion that emulation works, let alone works at a decent speed. I speak from experience of windows on Linux non-direct-x game emulation. I can't out perform a Pentium 2!