Optimised packages of AI models and workflows with API have been packaged as NIMs (Nvidia Inference Microservices) which developers can use as building blocks to develop generative AI for healthcare, from drug discovery, med-tech and digital health products. Nvidia announced 25 NIMs at its developer conference, GTC 2024, offering advanced imaging, natural language and speech recognition, digital biology generation, prediction ...
EDA and IP
Nvidia offers Omniverse Cloud as APIs
A new take on the concept of digital twins was announced by Nvidia at its developer conference in San Jose this week, as founder and CEO, Jensen Huang announced that the company’s Omniverse Cloud will be available as APIs (application programming interfaces). Customers will be able to use these APIs to design, simulate, test and build physically based digital twins ...
Network-on-chip supports Arm v9 and automotive functional safety
Arteris is supporting Armv9 automotive cores and ASIL D functional safety with release 3.6 of its Ncore cache coherent network-on-chip. “In 2020, Arteris announced a partnership with Arm and the alignment of our roadmaps to support CHI-E and ASIL B and ASIL D safety,” Arteris told Electronics Weekly. “We confirm the expansion of that partnership with a pre-validation of our ...
Emulation and prototyping trio addresses AI, ML and hyperscale computing
Siemens EDA has announced three products, for emulation, evaluation and software prototyping for data centres and servers. The Veloce CS consists of the Strato CS hardware, for emulation, the Primo CS hardware for enterprise prototyping and proFPGA CS hardware for software prototyping. According to the company, the hardware-assisted verification and validation system can accelerate verification and validation cycles by up ...
Training-grade PDK for 2nm gate-all-around ICs
Imec has launched an early-access process design kit for the N2 (2nm ICs) node. “The PDK will enable virtual digital designs in imec’s N2 technology, including backside power delivery network,” according to the Belgion semiconductor research lab. “This will give academia and industry the tools to train the semiconductor experts of tomorrow and enable the industry to transition their products ...
IP block secures FPGAs with one external IC
Protecting against IP theft, and to prevent contract manufacture hardware counterfeiting, ‘FPGA Lock’ from Nial Stewart Developments is an intellectual property block for FPGAs that communicates with Microchip’s SOT-23 ATSHA204A crypo-authentication IC. The cores uses generic Intel altsyncram and Xilinx 7 series blockram macros and occupies ~720 registers and two ram blocks. Only one pin on the FPGA is needed, ...
Faraday Technology plans 64 core ARM design on Intel 18A process
Taiwanese SoC design house Faraday Technology is collaborating with Arm and Intel on a 64-core system-on-chip, intended to be made with Arm Neoverse data-centre-grade cores using Intel’s 18A (nominally 1.8nm) foundry process – the latter due to become available at the end of this year. It will use Arm Neoverse Compute Subsystems (CSS, pictured), where “Arm delivers validated, performance-optimised compute ...
Try a network-on-chip configuration tool?
SignatureIP is offering a trial of its processor agnostic network-on-chip (NoC) configuration software. Called iNoCulator, the “tool enables users to find the optimal NoC configuration, and the company believes that it is the fastest on the market”, it said. “This try-before-you-buy offer will enable engineers to see for themselves just how easy it is to design a NoC with our tool.” ...
Cadence adds apps for billion gate SoCs to emulation hardware
Cadence has added three apps to its Palladium Z2 emulation hardware for SoCs. “These domain-specific apps allow customers to manage increasing system design complexity, improving system-level accuracy and accelerating low-power verification for applications such as artificial intelligence, machine learning, hyperscale and mobile,” according to the company. They are: Four-state emulation for simulations requiring X-propagation, such as for low-power verification of ...
Synopsys extends GenAi across full stack
Synopsys is expanding its Synopsys.ai EDA suite to bring GenAI across the full stack to improve engineering productivity for the semiconductor industry. This expands upon Synopsys’ recent announcement to deliver Synopsys.ai Copilot, the first in a series of GenAI capabilities for chip design. Early collaborations with AMD, Intel, and Microsoft have substantiated the power of generative AI for chip design. ...
Electronics Weekly