A third-party SoC developer has introduced an octo-core smartphone SoC using 14nm x86 Airmont (Atom) cores from Intel:
To date, the Spreadtrum SC9861G-IA is the most powerful (and presumably energy-efficient) x86-based SoC for smartphones. It has more cores, better graphics, and a faster modem than Intel's own code-named Moorefield SoCs introduced in 2014, made using its 22 nm fabrication process, or the SoFIA chips (designed by Rockchip) launched in 2015 made using TSMC's 28 nm technologies. Using Intel's 14 nm manufacturing technology for this new SoC helps to reduce minimum power requirements and die size (which still remain unknown).
The SC9861G-IA is the first x86-based SoC by Spreadtrum, and the development was enabled by an agreement signed in late 2014 after Intel acquired a $1.5-billion worth stake in Tsinghua Unigroup, the owner of Spreadtrum. The chip will not carry the Intel Atom brand, and thus Intel will not help makers of devices to integrate it or make any other incentives to popularize the platform. It will also not invest in its advertising. What is interesting is that the SC9861G-IA will not be Spreadtrum's last x86-based SoC, according to the CEO of Intel.
(Score: 0) by Anonymous Coward on Monday March 20 2017, @04:27PM (3 children)
If x86 couldn't succeed in phones with Intel's marketing and incentives behind it, how do you think it will fare without that support?
X86 only has a distinct advantage when you need compatibility with existing x86 binaries or assembly code. How many people want to run Skype, or Windows apps with wine, or any such x86-only software on a phone?
The only place I can see this changing is gaming, where we've recently seen a historically unprecedented (though still small) fraction of closed-source games released for Linux on x86, but not for other architectures. This could be big, but I'm not sure most of those games are even playable on a phone, unless you turn it into an ersatz laptop by adding a bluetooth keyboard and mouse. And given any of the usual phone OSes, there's going to be annoying hoops to jump just to run them. (At minimum, you'll need some sort of X server, whether alongside or replacing your phone OSes native display system.) So unless you're making a "steam-ready" phone specifically for gaming, and doing X server integration yourself, I can't imagine why you'd use this SoC instead of one of the gazillion ARM SoCs.
If this were a Chromebook SoC, rather than smartphone, I could see it -- I have a RK3288 (quad-core ARM) Chromebook, and am pretty happy with crouton (a chroot-based way of installing full Linux distros alongide ChromeOS), but I recently bought a Braswell NUC rather than an ARM-based mini PC specifically because I can't seem to get Android Studio running on the Chromebook. (AFAIK, it's all open-source, so I should be able to build it for ARM, but in practice, I haven't been able to make that work.) The gaming support is, of course, a nice extra; goodbye, browser-crashing asm.js version of FTL! So I can see why many people prefer x86 Chromebooks, and if this were a Chromebook-targeted SOC, it could make sense, though I'm not sure how it would compete with Intel's own Chromebook SOCs.
(Score: 1) by petecox on Monday March 20 2017, @09:54PM (2 children)
The article says phones but it's more likely to be in $200 "Wifi + Cellular" Windows 10 tablets.
(Unless MS is still clinging to the Surface Phone vaporware - they are emulating x86 on Snapdragon 835 for lack of a suitable Atom)
(Score: 2) by takyon on Tuesday March 21 2017, @12:50AM (1 child)
Have 8 cores ever been put in a tablet or Chromebook?
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 0) by Anonymous Coward on Wednesday March 22 2017, @08:33AM
Is it using Intel signed microcode, or Spreadtrum microcode, and does it include Intel's Management Engine technology, or signed GPU firmware?
If not, the followup question would be: Does the memory controller contain the circuitry necessary for driving RAM DIMMs instead of RAM chips directly integrated on the circuit board?
The final question: Does it contain ECC support?
No, Yes, Either would be a pretty nice x86 chip. With ECC support it would open up a whole lot of robust low power device opportunities on the bottom end, especially with VT/IOMMU support.
(Score: 2) by iWantToKeepAnon on Monday March 20 2017, @11:53PM (1 child)
"Happy families are all alike; every unhappy family is unhappy in its own way." -- Anna Karenina by Leo Tolstoy
(Score: 0) by Anonymous Coward on Tuesday March 21 2017, @12:42AM
I can only assume they mean to compare it in a performance/W basis. AFAIK all prior x86 SoCs targeted anywhere in the phone/tablet/netbook range have had no more than 4 cores, so I'm quite sure total power consumption will increase despite the gains from 14nm.