Qualcomm Announces Samples of New 7nm SoCs with 5G Modems
Qualcomm doesn’t normally announce when it’s shipping new SoCs to customers, but the company has done so this time. According to QC, its unnamed next-generation SoC is sampling to specific customers with an included X50 5G modem. That’s a significant achievement for the new product, since pushing 5G out to handsets is a vital component of ramping up 5G support.
“[We’re] on track to help launch the first 5G mobile hotspots by the end of 2018, and smartphones using our next-generation mobile platform in the first half of 2019,” said Qualcomm president Cristiano Amon.
The new chip will be built on TSMC’s 7FF process node, which is the foundry’s first 7nm node deployment. TSMC’s 7nm is generally expected to perform roughly equivalent to Intel’s 10nm node, except 10nm, of course, has been delayed in volume back to Q4 2019. These upcoming chips will therefore represent a major new shift in which companies are effectively leading the foundry industry in overall lithography.
Qualcomm isn’t leading the charge for 7nm, though, at least not from this announcement. Huawei has claimed that they’ll be the first company to ship 7nm silicon (beating even Apple). Full details on the new platform are expected in the fourth quarter.
As for the overall gains from moving to 7nm in general, expect them to be fairly modest and to come with the typical “either/or” phrasing — which is to say, you can increase performance by X percent or you can improve power consumption by Y percent, but where you choose to sit on that curve will change a great deal about how a device performs and improves on its predecessor. Gains in the smartphone space have slowed in recent years as the platform hit a “good enough” plateau for most functions and it became harder to extract higher overall performance from the system without simultaneously stacking more problems together (higher heat dissipations, higher levels of battery drain, and so on).
One reason we now see Qualcomm and other companies putting much more focus on the performance of their various accelerators or AI/ML-devoted hardware is that marketing a new phone as “the same as what you used to have but with faster clocks” has become a weak point. The general-purpose CPU cores in phones still have room to improve, but it often makes more sense to spend time optimizing a workload to run on a specialized core, like Qualcomm’s DSPs, than it does to try and improve it by boosting an ARM CPU’s clock rate. Mobile computing’s narrow TDP range and steep clock throttling over time make it vital to always try and execute CPU instructions in the most advantageous way possible, which very much includes moving that data into the execution units best equipped to calculate it.
As these chips continue to shrink, expect to see more accelerator cores folded into the mix and brought online. Improving smartphone and IoT device performance these days is as much or more about boosting the performance of associated accelerators as it is with boosting the CPU itself.
Continue reading
NASA Created a Collection of Spooky Space Sounds for Halloween
NASA's latest data release turns signals from beyond Earth into spooky sounds that are sure to send a chill up your spine.
Astronomers Might Finally Know the Source of Fast Radio Bursts
A trio of new studies report on an FRB within our own galaxy. Because this one was so much closer than past signals, scientists were able to track it to a particular type of neutron star known as a magnetar.
Sony’s PlayStation 5 Debuts to Strong Reviews
Reviews have come in for Sony's PlayStation 5, and while they're a bit preliminary for the same reason as the Xbox Series X, they're broadly positive about Sony's latest gaming effort.
How to Build a Face Mask Detector With a Jetson Nano 2GB and AlwaysAI
Nvidia continues to make AI at the edge more affordable and easier to deploy. So instead of simply running through the benchmarks to review the new Jetson Nano 2GB, I decided to tackle the DIY project of building my own face mask detector.