As the incremental progress of Moore's Law continues to slow, mobile chip design is changing to reflect that situation, this analysis notes. One factor is the requirement to add artificial intelligence capability and neural network accelerators to mobile ICs, says Ron Lowman of Synopsys.
Analog content in chip designs is often causing test failures and field returns, Brian Bailey writes. Steve Lewis of Cadence Design Systems says, "Analog is a dinosaur waiting for a meteor."
High-density fan-out IC packages continue to advance in complexity, creating new challenges for the lithography equipment and related tools involved, Mark LaPedus writes. "Mobile continues to be one main growth driver for both low-density and high-density fan-out," said John Hunt of ASE, along with automotive and server applications.
MediaTek's mediocre financial performance in the first quarter is expected to turn around in the second quarter thanks to volume shipments of its Helio P90 and P35 mobile system-on-a-chip devices, market observers say. Its Q2 revenue may surpass $1.95 billion, those observers predict.
Taiwan-based foundries don't have much visibility on second-quarter orders as IC design firms are still clearing out their inventories, industry sources say. While 8-inch wafer fabrication lines remain busy, Taiwan Semiconductor Manufacturing Co. and United Microelectronics are seeing Q2 order opaqueness for their 12-inch wafer foundry services, those sources note.
An international research team used artificial intelligence and machine learning technology to predict the most energy-efficient strain pathways that could turn diamond materials into high-performance semiconductor devices. The discovery has potential applications in energy, communications and information processing.
STMicroelectronics brought out the LSM6DSOX iNEMO inertial sensor, which has a machine-learning core to identify motion data from known patterns. ST designed the new sensors to offer improved activity-tracking performance and battery life in wireless and wearable gadgets.
Next week's International Solid-State Circuits Conference in San Francisco won't have any papers on new general-purpose microprocessors, yet the conference will cover quite a bit about the latest developments in artificial intelligence, automotive electronics, 5G cellular communications and health care. IBM, Samsung Electronics, Arm, Intel and Nvidia will give talks on their work in AI and machine learning.
Ericsson and Intel have teamed up for the development of a platform for distributed cloud, 5G and network function virtualization. Ericsson will provide its software-defined infrastructure management software in the project, while Intel will leverage its Rack Scale Design.