Floadia Corporation (hereinafter referred to as "Floadia''), which develops embedded flash memory IP cores based in the Tokyo suburb of Kodaira-shi, has raised approximately 1.05 billion yen in Series D round.

Logo: https://cdn.kyodonewsprwire.jp/prwfile/release/M104461/202310050700/_prw_PI1fl_9kEF385a.jpg

As a first closing, Inabata & Co., Ltd. (hereinafter "Inabata") headquartered in Chuo-ku, Osaka, and Cypress Capital, a Hong Kong investment company, underwrote an allotment of capital for a total of 850 million yen. Including the 200 million yen raised through venture debt (loan with stock acquisition rights) from Japan Finance Corporation in March 2023, the total amount raised in this round was approximately 1.05 billion yen.

Floadia will use the funds raised to strengthen sales activities promoting its current primary business, embedded flash memory IP cores, to semiconductor manufacturers, and to enhance developing ultra-low power consumption AI accelerator chips that utilize flash memory devices. Inabata, which participated in this round, operates a wide range of businesses overseas in the information & electronics, chemicals, consumer goods industry and plastics sector, and plans to utilize the network to support Floadia's sales overseas.

Floadia is in the business of licensing the processes and circuit design necessary for manufacturing flash memory, a type of non-volatile memory (memory that stores data even when the power is turned off), to semiconductor manufacturers as intellectual property (IP). Floadia's flash memory IP cores are mainly used in microcontrollers, power semiconductors, sensors, etc., and require extremely lower power consumption than conventional flash memory IP cores, and can be integrated into chips at low cost. "G1," the first flash memory IP developed by Floadia, has been used in microcontrollers at Toshiba Electronic Devices & Storage Corporation, and passed quality standards that allow data retention for 10 years at 125 C after 10,000 Program and Erase operations at Taiwanese foundry TSMC. The IP was tested at even higher temperature of 200 C, and was confirmed to retain data for 10 years, the highest quality in the world.

Floadia is developing Computing in Memory (CIM) chips, which use flash memory devices as information transmission media, as a pillar of the next business following flash memory IP cores. Current AI calculation is mainly performed on large-scale servers located in the cloud such as data centers, but in addition to issues such as data security and latency, there are concerns about increased power consumption at data centers. Therefore, AI calculation processing on the terminal (edge) is required. Floadia's flash memory solves these issues with analog memory technology and enables advanced AI inference calculations in edge environments with significantly low power consumption.

Floadia has so far received approximately 3.9 billion yen from INCJ (Japanese government fund), UMC Capital (Taiwan), an affiliated company of major foundry UMC, Faraday Technology (Taiwan), a cutting-edge semiconductor design company, and Teijin Limited, a major Japanese chemical firm. Adding the amount raised in this round, the cumulative total raised to date is approximately 4.95 billion yen.

For inquiries regarding this press release: https://floadia.com/contact/

About Floadia Corporation

Floadia was established in 2011 by experienced engineers who had been developing embedded non-volatile memory for over 20 years at Hitachi, Ltd., and Renesas Technology (currently Renesas Electronics Corporation). The company licenses to semiconductor manufacturers as intellectual property (IP) the processes and circuit designs necessary for manufacturing embedded non-volatile memory (memory that retains its contents even when the power is turned off) used in microcontrollers, power semiconductors, sensors, etc. Floadia's non-volatile memory technology consumes only 1/1,000,000 of the power required to write and erase data, as compared to competitors' memory technologies, and has excellent high temperature tolerance, and requires only 1/3 of the additional cost to integrate into a chip. Because of these features, G1 has already been installed in automotive microcontrollers by domestic semiconductor manufacturers and has also been adopted by a foundry in Taiwan, where it is used as embedded memory in smartphone components.