Week In Review: Design, Low Power

Infineon reported fourth quarter 2021 financial results with revenue of €3.0 billion (~$3.4 billion), up 21% compared to the same quarter last year. For the full year, revenue was €11.1 billion (~$12.7 billion), an increase of 29% from the previous year. “In light of the continued high demand for semiconductors needed for the energy-efficient, connected world, we expect the 2022 fiscal year to be a strong one,” said Infineon CEO Reinhard Ploss. “We are continuing to expand our manufacturing capacities – for silicon as well as for the compound semiconductors silicon carbide and gallium nitride.”

Panasonic Automotive utilized Ansys‘ Medini Analyze modeling and analysis solutions in reaching ISO 26262 ASIL-D process certification and establish a systematic, model-based approach to the safety lifecycle for functional safety analysis throughout its eCockpit development process.

Corigine uncorked the MimicTurbo GT card for FPGA-based prototyping at the desktop. Based on the Xilinx UltraScale+ VU19P FPGA, it offers automated FPGA partitioning and interconnect, supports up to 48 million ASIC gates each and can be configured to operate with additional connected cards.

IAR Systems added support for the Arm Cortex-M55 processor to its Embedded Workbench for Arm as well as support for latest MCU devices from several semiconductor vendors in the 9.20 version of its toolchain.

Memory & storage
Samsung Electronics revealed 16 Gb LPDDR5X DRAM based on a 14nm process. While LPDDR is commonly associated with smartphone and mobile applications, Samsung said that it is also targeting applications like servers and automotive.

Market research firm IC Insights expects PC and server DRAM prices to slip up to 5% in the fourth quarter due to fewer orders from PC and server manufacturers that are still using the inventory purchased earlier in the year. Graphics DRAM could fall as well, as shortages of other components limit graphics card production. It expects mobile DRAM to be flat through the rest of the year. Overall, DRAM saw a big boost this year, rising 41% through the first eight months of 2021.

Kioxia announced a number of data center focused SSDs. Its CD7 Series of EDSFF E3.S SSDs uses PCIe 5.0 and targets high-performance, highly efficient servers and storage. Based on fourth generation BiCS FLASH 3D flash memory technology, the CD7 Series supports x4 PCIe lanes, but is optimized for PCIe Gen5x2 performance.  Another line, the XD6 EDSFF E1.S data center class SSDs, targets hyperscale applications, including the performance, power and thermal requirements of the Open Compute Platform NVMe Cloud SSD Specification. The EM6 Series Enterprise NVMe-oF SSDs  were released for Ethernet Bunch of Flash (EBOF) systems and targets AI, HPC, and storage expansion.

FADU Technology debuted a family of SSDs and SSD controller that support PCIe 5.0×4, NVMe 1.4a, and OCP Cloud Spec 2.0. Available in E1.S, U.2, E3, and E1.L form factors, it targets OEM and hyperscale data centers.

Quantum computing
Nvidia debuted a software development kit to speed up simulations of quantum computers on classical systems. The first library in the cuQuantum SDK is cuStateVec, an accelerator for the state vector simulation method, an approach that tracks the full state of the system in memory and can scale to tens of qubits. Planned for December is cuTensorNet, an accelerator using the tensor network method that can handle up to thousands of qubits on some near-term algorithms.

The government of New South Wales, Australia, plans to build a quantum technology hub near Sydney’s Central station to run the city’s transport network, reports The Age. “Quantum computing can actually help us to deploy resources far more accurately, and we genuinely don’t know what the long-term impacts of the pandemic are going to be on travel patterns and on travel preferences,” said NSW Transport Minister Rob Stokes. “While this might sound like the stuff of science fiction, Transport for NSW is making quantum computing a reality. It has the potential to solve problems on the network in real time by instantly recalculating timetables and routes.” Australia’s universities and research institutes have put a big focus on quantum computing, making the country one of the early leaders in the space. It is also home to a significant number of startups across the entire quantum ecosystem, from hardware to programming to security.

NSW is not the only one that hopes quantum computing can help solve transportation problems this week. Cambridge Quantum and German railway service provider Deutsche Bahn Netz are working together to demonstrate how current quantum processors and algorithms can improve the rescheduling of rail traffic after delays.

The UK Research and Innovation’s Commercialising Quantum Technologies challenge is putting £50 million (~$66.3 million) into twelve projects that aim to bring quantum computing to practical applications. One of the projects will focus on using a quantum computer alongside a classical supercomputer to model the properties of drugs, and another will develop quantum technology to detect gas leaks for hydrogen energy projects. Others will work on developing resilient quantum clocks for financial trading, energy distribution, and navigation.