Intel Foundry Providers (IFS) this week soft-launched their new Intel 16 course of know-how, a 16nm-class node that might be used for the manufacturing of low-power chips for on a regular basis workloads. The up to date legacy node, derived from Intel’s present 22FFL course of tech, is geared toward cost-conscientious clients who’re producing easier chips that do not require the efficiency provided by cutting-edge course of nodes. Set to compete towards nodes comparable to TSMC’s N12e, finances nodes like Intel 16 usually see extensive use in a wide range of fields, starting from aerospace and protection to IoT and radios.
First revealed by Intel a few years again, the ramp-up of Intel 16 comes as IFS is within the technique of increasing its foundry choices with the intention to provide the complete vary of course of nodes that chip designers have come to count on from a contract fab. Whereas not on the forefront of fab dialogue as innovative nodes are, trailing-edge and mature course of nodes are nonetheless used to supply an unlimited variety of chips yr after yr, usually easy chips that want few upgrades and will find yourself in manufacturing for a decade or longer.
TSMC, for its half, earns round 25% of its income by making lots of of tens of millions of chips utilizing 40 nm and bigger nodes. For different foundries, the income share share of 40nm and comparable course of applied sciences is even larger: SMIC and UMC earn over 80% of their income on mature nodes. These chips have very lengthy lifecycles – to the purpose that TSMC has a tricky time persuading its clients to begin utilizing 28nm – underscoring their significance within the general chip ecosystem, and the necessity for IFS to supply some cheaper, much less superior nodes to court docket these chips and their designers.
IFS’s Intel 16 node targets all kinds of purposes, together with software processors, analog, shopper electronics radio frequency (comparable to Wi-Fi and Bluetooth), mmWave, storage, navy, aerospace, and government-usage chips. This FinFET-based know-how is designed to hit a candy spot between efficiency and value, providing appreciable transistor density and excessive efficiency due to its use of FinFETs, whereas nonetheless costing far lower than modern nodes the place EUV machines and multi-patterning shortly drive up prices. Intel says that its 16nm-clss know-how requires fewer masks in addition to comparatively easy back-end design guidelines. In essence, we’re taking a look at a cost-optimized amalgamation of Intel’s 22nm and 14nm nodes, bringing 22FFL a bit extra ahead with out using a few of the costlier features of 14nm.
The majority of this week’s bulletins, in flip, are centered on chip design for the brand new node. Main digital design automation (EDA) and IP suppliers — Ansys, Cadence, Siemense EDA, and Synopsys — are all asserting their backing of the Intel 16 course of know-how with their licensed software program flows and IP. For instance, Cadence has ported a variety of its IP blocks to Intel 16, together with PCIe 5.0; 25G-KR Ethernet multi-protocol PHY; multi-protocol PHY for shopper purposes that help requirements like PCIe 3.0 and USB 3.2; multi-standard PHY for LPDDR5/4/4X reminiscence; and MIPI D-PHY v1.2 for cameras and shows. Moreover, Synopsys additionally has its AI-enabled Synopsys.ai toolkit that now helps Intel 16.
With the required tooling now in place, chip designers can begin utilizing design, verification, and simulation instruments to develop their chips on IFS’s 16nm-class fabrication know-how. At this level it stays unclear when Intel expects to begin fabbing chips on the Intel 16 course of, although because it’s based mostly on mature designs, it is seemingly extra a matter of when the primary buyer chip designs might be prepared.
Sources: Intel, Ansys, Cadence, Siemense EDA, and Synopsys.
