Many innovations in memory, whether it's a new technology or an existing one, come as a result of new materials, but one company is looking to improve on current DRAM technology without new materials or complex process flows.

Kilopass Technology, Inc., maker of one-time programmable (OTP) memory and provider of intellectual property for semiconductor embedded non-volatile memory (NVM), has introduced its Vertical Layered Thyristor (VLT) technology for DRAM applications. VLT eliminates the need for DRAM refresh, is compatible with existing process technologies and offers significant other benefits including lower power, better area efficiency and compatibility, company CEO Charlie Cheng said in an interview with EE Times.

Kilopass' VLT is based on thyristor technology that has a structure electrically equivalent to a cross-coupled pair of bipolar transistors that form a latch. Cheng said the latch lends itself to memory applications since it stores values and, as opposed to current capacitor-based DRAM technology, does not require refresh. Part of Kilopass' approach to VLT was to implement the thyristor structure vertically.

Since VLT does not require complex performance- and power-consuming refresh cycles, he said the main benefit of a VLT-based DDR4 DRAM is that it lowers standby power by 10 time when compared with conventional DRAM at the same process node. Given that the DRAM market is being driven by strong demand in the server and cloud computing market, Kilopass sees a great deal of potential for its VLT DRAM, said Cheng, as servers and server farms consume a tremendous amount of energy with memory being a major contributor.

Ideally, he said, current generation of 20nm DRAM would migrate to sub- 20nm processes to further reduce power consumption, but DRAM industry is struggling to increase memory performance while reducing power consumption because of the physics at play with the current 1 transistor, 1 capacitor (1T1C) bitcell technology. Cheng said 1T1C is difficult to scale since the smaller transistors exhibit more leakage and the smaller capacitor structure has less capacitance, resulting in the need to reduce the time between refresh intervals.

This challenge is what led Kilopass to develop VLT DRAM and a new architecture that requires fewer processing steps and is designed to be built using existing processing equipment, materials and flow. Cheng said one of the reasons the company wanted to create a new DRAM technology without the need for new materials is that it doesn't have its own fab.

In addition, he said, the company has gained a lot varied experienced in different technologies thanks in part to its OTP memory business, which has seen a lot of traction and mobile wallet applications. "Unlike most development teams that are specialized in one area for one time, we end up with expertise in DRAM and CMOS processes. It gives us a broad view of what is possible."

And while the mobile segment has grown significantly, Cheng sees the server, cloud computing, computational and graphics markets as areas that can best benefit from Kilopass' VLT DRAM. "Since 2010, DRAM scaling has significantly slowed down compared to SoC and NAND flash." Back in 1985, the refresh cycle time was insignificant, he added, but now it's becoming unpredictable and a large problem for servers, cloud computing and graphics applications."

Kilopoass is aiming to commercialize its VLT DRAM in early 2018, and unlike its OTP intellectual property, Cheng said the company is looking at a "protective licensing" model and limiting the technology to three licensees that have yet to be selected.

By: DocMemory
Copyright © 2023 CST, Inc. All Rights Reserved