[Featured Stock] OpenEdgeTech Targets Nvidia and Mobileye with NPU Performing 250 Trillion Operations per Second

Open Edge Technology is showing strong performance. News that the company is developing a high-performance 'Neural Processing Unit (NPU)' semiconductor design asset (IP) capable of 250 trillion operations per second appears to be influencing its stock price.

As of 2:07 PM on the 31st, Open Edge Technology is trading at 22,250 KRW, up 7.23% from the previous trading day.

The day before, an IT specialized media outlet reported that Open Edge Technology is developing an NPU IP with 250 TOPS, aiming for release in the first half of next year. TOPS is a unit indicating computational speed, and 250 TOPS means it can perform 250 trillion operations per second.

Open Edge Technology supplies IP and solutions necessary for AI semiconductor design. It primarily develops IP related to technologies that increase the speed at which system semiconductors process data received from external sources, providing these to foundries, fabless companies, and design houses. The company has secured globally leading technology mainly within the Samsung Foundry ecosystem.

Lee Jong-wook, a researcher at Samsung Securities, stated, "The recent increase in demand for HBM DDR PHY, driven by the need for higher transmission speeds, creates a favorable environment for Open Edge Technology," adding, "The generation of sales from HBM DDR PHY will help improve the product mix." He also noted, "As AI semiconductors begin to be produced mainly by TSMC, Open Edge Technology, which operates primarily within the Samsung Foundry ecosystem, will have an opportunity to expand its domain to TSMC."

The researcher introduced, "Autonomous driving NPUs are expected to be a long-term growth driver along with chiplets," and added, "We developed a 16TOPS-level autonomous driving NPU equipped with NPU design and memory systems through a government project." He continued, "We plan to have customer sample products in 2023," emphasizing, "We are also participating in the consortium development of the third-generation NPU with 250TOPS level, supporting autonomous driving level 4 and above."

By utilizing Open Edge's NPU IP, semiconductors capable of implementing autonomous driving functions based on AI can be created. The 250 TOPS performance is close to that of server chip computational power, enabling support for autonomous driving levels 3 to 4. Level 4 is the stage where the vehicle independently recognizes and judges surrounding situations without driver intervention, just one step below level 5, which is 'full autonomous driving.'

These are called 'edge solutions' or 'edge AI semiconductors,' meaning that data is processed and analyzed independently at the device level?such as in cars or robots?without passing through servers or data centers. Autonomous driving chips are currently dominated by companies like NVIDIA, Mobileye, and Tesla. Attention is focused on whether domestically competitive autonomous driving semiconductors will emerge.

Earlier, the government announced in March that it would invest 3.2 trillion KRW in developing core technologies in next-generation semiconductor fields such as power, automotive, and AI semiconductors. The Ministry of Trade, Industry and Energy identified three promising semiconductor technology areas: power semiconductors, automotive semiconductors, and AI semiconductors. On the 21st, it announced an expansion of the preliminary feasibility study budget for power and automotive semiconductors to 1 trillion KRW.