The Case for a Unified Qualcomm SoC: Bridging Mobile and PC
The strategy of repurposing mobile silicon for larger form factors—as seen with MediaTek’s Kompanio line or Apple's Macbook Neo—presents a compelling roadmap for Qualcomm. By utilizing a common die across both smartphones and PCs, Qualcomm could fundamentally shift its competitive position in the PC landscape.
Potential Benefits
Lower Production Costs: By using the same dies for mobile and PC chips, Qualcomm can leverage massive economies of scale, minimize R&D costs, and improve time to market.
Annual Upgrade Cadence: Smartphone chips receive yearly refreshes. Adopting a unified die strategy would enable Qualcomm to deliver yearly chip upgrades for the PC market, allowing them to remain competitive with the annual release cycles of Apple Silicon and Intel. This would be less economical if Qualcomm taped out seperate dies solely for PCs, since their PC volume is small.
Potential Disadvantages
The primary hurdle is "wasted die area". There are differences between the requirements of PCs and smartphones.
Smartphone Chips: Require an integrated modem for power efficiency and a large ISP (Image Signal Processor) to handle complex mobile photography.
PC Chips: Demand expanded I/O (more USB and PCIe lanes) and additional display controllers to support multiple high-resolution external monitors.
Including PC-specific I/O on a phone chip, or a massive ISP on a laptop chip, creates "dead weight" on the silicon that increases costs without adding value to the specific device.
One way to mitigate the "wasted area" is to move the cellular modem off-chip. While integrated modems are traditionally superior for power management, multiple generations of iPhones have proven that Qualcomm’s external modems are exceptionally high caliber. Even as Apple transitions to in-house silicon, their custom modems (such as the C1) remain external. Reports suggest Apple may eventually merge their N-series (Wi-Fi/Bluetooth) and C-series (5G) into a single connectivity chip.
Qualcomm could adopt a similar modular approach. By removing the modem from the primary SoC and integrating it into their FastConnect (Wi-Fi/Bluetooth) chip, they could create a universal SoC die for both PCs and smartphones. Hardware configurations would then be determined by the secondary connectivity chip:
PC Configuration: SoC + FastConnect (Wi-Fi + BT)
Phone Configuration: SoC + FastConnect (Wi-Fi + BT + Cellular)
How would this SoC setup work in practice?
| SoC die | Example config | Mobile | PC |
| Die 1 | 2L+6M CPU
3 slice GPU | 8 series
(for phones/
tablets) | X# Plus
(for budget laptops) |
| Die 2 | 6L+6M CPU
5 slice GPU | 9 series
(for tablets) | X# Elite
(for premium ultrabooks) |
It doesn't make sense for high end PC chips like X# Elite Extreme to share common dies with phones/tablets. So those will have to use seperate dies for solely that purpose, and might have to stick a 2 year release cycle.
This whole post has been written assuming the SoCs are monolithic. I didn't mention chiplets, which opens a whole new sea of possibilities.
