Good afternoon and welcome to eMemory's Fourth Quarter 2024 Webcast Investor Conference. Joining us today is our Chairman, Dr. Charles Hsu, President, Mr. Michael Ho, Head of IR, Ms. Li -Jeng Chen, and Director of the Finance Department, Ms. Teresa Kuo. The format of today's event will be as follows. First, eMemory's Chairman, Dr. Charles Hsu, will give an opening remark. Afterwards, President, Mr. Michael Ho, will summarize our operations, followed by our business outlook. Next, Dr. Charles Hsu will give a talk titled, "Why Post-Quantum Cryptography Needs PUF." Then, we will conclude today's conference with the Q&A section where our management team will answer your questions. Please feel free to submit your questions in the input box on the webcast window throughout the conference. As a reminder, this conference is being recorded, and a webcast replay will be available after the conference is finished.
For more information, please visit the company's website under the Investor Relations section. As usual, before we begin, we would like to remind everyone that today's presentation may contain forward-looking statements subject to risk factors associated with the semiconductor and IP business. Please refer to the cautionary statement on page three of today's presentation. Now, I would like to give the floor over to eMemory's Chairman, Dr. Charles Hsu.
Okay, thank you. Good afternoon, everyone. Thank you for joining our conference today. As we have mentioned in the previous quarters, we are currently entering a multi-year growth cycle. The recent licensing activities indicate that our technology has been adopted at an increasing rate, and we are very optimistic about the future. Recently, there has been a growing focus on the potential for faster AI deployment at the edge, which is a very encouraging trend for us. Our technology, including OTP and MTP, along with a variety of security IPs, enables chips to achieve improved performance, reduced costs, and enhanced security. Since edge computing demands low power, higher cost effectiveness, and stricter security, this will accelerate our customers' adoption of our IPs. This trend is already reflected in our recent licensing across various applications, such as networking, BMC, smart surveillance, and other edge-related applications.
Moreover, there is an ongoing discussion regarding the timeline for the commercialization of quantum computing, with opinions varying on whether it will occur in five years or 15. Quantum computers have the capability to quickly compromise current encryption standards that safeguard a significant portion of global tech data and infrastructure. Considering that hardware transitions typically take between five to 10 years at least, the U.S. National Institute of Standards and Technology has implemented post-quantum cryptography, which stands for standards to mitigate the cybersecurity risks associated with quantum computing. This transition presents a significant opportunity for our PUF-based solutions, including our newly launched PQC IP, post-quantum cryptography IP. Later, I will provide further insight on this topic. Next, I would invite our President, Michael Ho, to share our fourth quarter performance and future outlook. Michael.
Thank you, Charles. Good afternoon, everyone. Now, let's begin with our 2024 Fourth Quarter financial results. The Fourth Quarter revenue was TWD 1.01 billion, up 12.4% sequentially and up 12.4% year-over-year. Operating expenses were TWD 445 million, up 12.7% sequentially and up 23.8% year-over-year. Operating income was TWD 566 million, with an increase of 12.1% sequentially and an increase of 4.9% year-over-year. Operating margin decreased by 0.1 percentage points sequentially and decreased by 4 percentage points year-over-year to 56%. Our net income, amounting to TWD 515 million, experienced an increase of 24.3% sequentially and increased 27.5% year-over-year. This quarter's operating margin fell by 4 percentage points compared to the same time last year, mainly because of the rising salary and bonuses. Employee bonuses are calculated at 15% of the pre-tax profit.
Unlike the foreign exchange losses we experienced in the fourth quarter of 2023, this quarter, we benefited from the foreign exchange gain. This variation in non-operating results resulted in higher bonus payouts, which contributed to a decrease in the operating margin. EPS for the quarter was TWD 6.89, and ROE was 62.5%. Next, let's move on to the revenue contribution by licensing and royalty. Licensing in the Fourth Quarter accounted for 31.2% of the total revenue, up 8.5% sequentially and up 15.1% year-over-year. Royalty in the fourth quarter contributed 68.8% of the total revenue, increasing 14.2% sequentially and increasing 11.3% year-over-year. Total revenue for the fourth quarter increased by 12.4% compared to the previous quarter and increased by 12.4% compared to the previous year. For the full year of 2024, total revenue increased by 18.2% compared to the previous year.
Total licensing and royalty increased by 22.5% and increased by 16.4%, respectively. With that, I will comment on our revenue contribution by specific IPs. NeoBit accounted for 20.9% of total licensing revenue in the fourth quarter, decreasing 27.2% sequentially and decreasing 8.3% year-over-year. Its royalties accounted for 24.8% of total royalty, up 9% sequentially and up 26.1% year-over-year. NeoFuse accounted for 39.1% of total licensing revenue in the fourth quarter, up 41.9% sequentially and up 5.1% year-over-year. In terms of total royalty revenue, NeoFuse royalties increased by 16.7% sequentially and increased by 7.8% year-over-year, accounting for 73.1% of total royalties. PUF-based security IPs contributed 22.6% of licensing revenue, increasing 104.1% sequentially and increasing 138.7% year-over-year, while its royalties accounted for less than 1% of total royalties.
MTP Technology accounted for 17.4% of total licensing revenue, down 30.1% sequentially and down 0.2% year-over-year. Royalty from MTP decreased 5.4% sequentially and decreased 13.8% year-over-year, accounting for 2% of total royalties. For the full year of 2024, the revenue by technology is as follows. NeoBit licensing revenue increased 22.2% year-over-year, and royalty increased 14.7%, accounting for 25.2% of the total revenue. NeoFuse licensing revenue increased 7.2%, and royalty increased 16.5% year-over-year, contributing to 61.3% of the total revenue. PUF-based security IP licensing revenue increased 23.4%, accounting for 4.5% of the total revenue. MTP Technology licensing revenue increased 59% year-over-year, and royalty revenue increased 30.3%, accounting for 9% of total revenue. Now, let's look at the royalties for 8-inch and 12-inch wafers. For 8-inch wafers, accounted for 40.8% of royalties, up 13.9% sequentially and up 20% year-over-year.
12-inch wafers contribute 59.2% of royalties, increasing 14.4% sequentially and up 6% year-over-year. In total, 181 product tapeouts were completed in the fourth quarter. We will provide more information in the management report. In the next section, I will address our future outlook. Regarding the licensing revenue, we anticipate the licensing revenue will continue its growth momentum due to the increasing demands from both foundries and the end chip customers. We continue to launch new IPs and available an increasing number of process nodes on worldwide foundries. As for royalty revenue, we expect the royalty revenue to continue its growth trend, driven by a robust number of tapeouts in the pipeline that are moving into production. In 2024, we achieved a record high in NTO numbers.
For new IP technologies, one, NeoFuse is advancing various derivative process at leading-edge nodes, having the secure design wins in 3, 4, 5, 6, and 7 nm process nodes. Two, RRAM is broadening into automotive grade and already has multiple customer design wins. Three, NeoFlash continues its progress in specialty process, aiming to replace embedded flash and external NOR flash. Four, developing 2 nm technology in correlation with leading foundries. About the business development platform, one, we have joined the Arm Total Design Program and introduced PUFrt as the hardware root of trust for the RSE in CSS. Two, we have developed PUFhsm, an embedded hardware security module solution for automotive chips and high-performance computing, HPC. Together with the hardware root of trust, PUFrt, PUFhsm provides a comprehensive secure enclave solution. This concludes my comments. Next, I will pass the time to Charles. Thank you.
Okay. Hi, everybody. Recently, there are quite a few demands on the PQC IP, post-quantum cryptography IP. So this time, I would like to spend a few minutes to share with you what is PQC and why PQC needs PUF. As we approach a future where quantum computers may be able to compromise our existing encryption techniques, the importance of post-quantum cryptography increases significantly, and PQC aims to create a cryptographic system that can withstand attacks from quantum computers, and to ensure PQC's effectiveness, we need reliable methods for key generation and maintaining the integrity of these cryptographic systems, which is where PUF comes into play. A PUF can effectively generate the long keys necessary for PQC.
As the security of PQC relies on these extended keys to defend against quantum computing threats, PUF can securely and efficiently produce the unique secrets needed to create these long keys, making it an ideal choice for the key provision in PQC. Additionally, PUF can also generate random numbers efficiently, which are crucial for preventing tampering in PQC. The essence of PQC lies in safeguarding data with long keys. And beyond being resistant to quantum attacks, cryptographic systems must also incorporate anti-tampering features to protect keys and encrypted data from physical threats. Randomness is vital for obscuring the data process. And since PUF naturally possess randomness and uniqueness, it can be utilized in a design that resists attacks and effectively implements anti-tampering strategies. So now, after knowing why PQC needs PUF, next, I would like to explain what is PQC.
PQC stands for the Post-Quantum Cryptography, which includes cryptographic algorithms designed to protect data from potential risks associated with quantum computers. These quantum computers are anticipated to outperform classical computers in solving specific mathematical problems, which pose significant risks to many cryptographic methods, particularly the widely used RSA and ECC algorithms. As their security relies on algorithms that the quantum computer can efficiently solve, PQC algorithms are intended to ensure secure communication and data protection, even in the face of quantum computing advancements, thereby maintaining security in our interconnected world when quantum computers become a practical reality. Why is PQC essential at this moment? As quantum computing progresses, the demand for encryption capable of resisting quantum attacks becomes critical. Many systems in use today retain encrypted data for an extended period, sometimes for years or even decades. Consider financial transactions, personal data, and sensitive government information.
If this data is encrypted with methods that are susceptible to quantum attacks, it could be at risk in the future. The sooner we implement PQC, the sooner we can guarantee the security of our data in the quantum futures. To prepare for a post-quantum environment, the National Institute of Standards and Technology has been spearheading the development of new cryptographic standards that can endure quantum computing threats. Following multiple evaluation phases, NIST officially announced in 2024 the adoption of lattice-based key encapsulation, lattice-based digital signature, and hash-based digital signatures as a standard for the post-quantum algorithms. Organizations globally are now embracing these new standards to safeguard our data against potential quantum risks. We also will be launching our PQC IP based on the first two standards in this quarter.
Given the necessity for our data to remain secure in the future and the time required to develop a new encryption system, it will be essential to begin transitioning to post-quantum cryptography now. This practical approach will ensure that we are prepared by the time quantum computers become capable of posing a threat. Typically, post-quantum cryptography algorithms involve increased computational complexity and more extensive key storage requirements. I will now outline how our PUF-based solution facilitates PQC. Our PUF-based root of trust comprises several components. First, there is a NeoPUF, which generates a unique ID for each device, enabling the creation of long, secure secret keys for the PQC. Next is NeoFuse PUF, OTP, a compact and secure method for storing PQC keys. In comparison to eFuse, OTP offers superior security, particularly in advanced nodes, while also being more area-efficient for the large key size.
And modern cryptographic systems, including PQC, are susceptible to side-channel attack. Therefore, our solution incorporates a built-in true random number generator, TRNG, that continuously supplies a substantial amount of random numbers. And these random numbers are crucial in thwarting attacks and at compromising PQC keys. Moreover, for a successful transition to PQC, the cryptographic system must accommodate both traditional and PQC algorithms, allowing the flexible selection between them or hybrid solutions. By integrating the PQC, by integrating the PUFrt into the security subsystem, it can effectively handle the large and complex keys needed for PQC algorithms alongside the small keys for traditional algorithms, eliminating the need for separate key generation for both systems. As previously mentioned, PUFrt can produce high-quality, unique PQC keys that are securely stored in OTP, safeguarding their confidentiality and integrity while preventing unauthorized access and tampering.
Additionally, the TRNG continuously generates significant volumes of random numbers, which are utilized for countering side-channel attacks, thereby enhancing the overall security of the key management within the system. So in conclusion, to ensure the long-term security for our data against quantum computing threats, the integration of a PUF-based security subsystem is vital. And this approach will allow us to maintain a high level of security and resilience as we transition to PQC. Okay. So thank you for your patience for the lectures.
Thank you, Charles. This concludes our prepared statement. We will now begin the Q&A session. Please submit your questions in the input box on the webcast window. All of our questions will follow the format of answering the Chinese version first, followed by the English version. We will now collect the questions and begin our Q&A session.
[Foreign language]
[Foreign language]
Our first question is, what is your perspective on the year-over-year decrease in operating profit margin during the fourth quarter of last year and your outlook for future operating profit margins, and this question is for Michael.
More than 80% of our operating expense is employee salaries, and 30% of the salary expense is linked to the employee profit sharing, which accounts for the 15% of the pre-tax income. Since pre-tax income includes non-operating items, and our main non-operating factors are the currency gains or losses related to our U.S. dollar positions, the currency losses from the fourth quarter of 2023 led to a decrease in the pre-tax profit due to the non-operating reasons. This, in turn, reduced the corresponding employee profit sharing expense and increased the operating profit margin. In Q4 of last year, on the other hand, currency gains were a contributing factor for the increase of expense, resulting in a 4% difference on operating margin year-over-year. Looking forward, we anticipate an upward trend in the overall operating margin. Thank you.
[Foreign language]
[Foreign language]
With the recent sharp decline in AI model costs, several CSPs and chip companies anticipate an acceleration in AI inference applications. What are your key advantages in the inference applications, Michael?
The decrease in AI costs benefits edge computing applications. In Edge AI, data training is used to establish an AI model, which is then used to infer possible results. This entire process, including the protection of input data, the model, and the generated result, requires hardware security IPs. We provide high-performance PUF-based security IPs to make Edge AI applications safer and accelerate industry development. Thank you.
[Foreign language]
[Foreign language]
Open Compute Project has defined Caliptra root of trust as recommendation for security architecture. eMemory has been preparing for this architecture for a long time. Could you elaborate on the revenue and profit contributions of related IPs? And Charles, please.
These three key components outlined in the Caliptra specification are the same three key features, OTP, PUF, and TRNG that our PUFrt was developed to provide over four years ago. Consequently, starting in mid-2024, we have indeed received inquiries from some customers regarding Caliptra. So in the fourth quarter, several projects have already been licensed and also in the process, and some of them are in the process of being finalized.
[Foreign language]
[Foreign language]
SoftBank and Arm have joined the U.S. Stargate project, and eMemory already has a partnership with Arm. Are there other opportunities for future collaboration, Charles?
We are optimistic about it.
[Foreign language]
[Foreign language]
The current utilization rate for mature processes is generally below 70%, and with China expanding its mature process capacity, this will inevitably exert long-term downward pressure on foundry process prices. Since a significant portion of your royalties come from mature processes, how do you plan to address this issue? Michael.
Our OTP and MTP are already in industry standard and widely applied in the mature process for major applications such as display driver, PMIC, ISP and various sensors. As foundries expand their capacity, they will need to license our technology, which will increase our royalties. As existing customers transition to the more advanced process, royalties per chip will increase, which can offset the downward pressure on foundry price due to the oversupply. To remain competitive, current foundries must focus on developing high-value-added specialty processes such as MTP, embedded flash and PUF, which is already underway. Royalties from MTP and security are much higher than those from OTP. Additionally, with advanced high ASP processes starting to contribute, we expect the average royalty per chip to continue increasing. Regarding the regional political factors, Europe, the U.S., and Japan are also expanding their capacities, which will require major customers to localize production.
Since we are developing our IPs across factories in the older regions, this makes it easier for chip customers to use our technology, strengthening the stickiness of our IPs and further broadening the scope of our business. Thank you.
[Foreign language]
[Foreign language]
What main function does NeoPUF technology serve in post-quantum encryption? Charles.
NeoPUF is capable of generating long random numbers reaching several megabit and even larger in length, which can serve as a secret key, and given the post-quantum cryptography, to have need to have secret keys with 20-60 times longer as compared to the conventional encryption techniques, so NeoPUF can effectively meet this demand for the key generations. As a result, PQC application depends on NeoPUF to deliver high-quality secret keys, enhancing the security of the encryption systems.
[Foreign language]
[Foreign language]
In previous earnings calls, it was mentioned that the 3 nm CPU project, which began last October, has already had a positive impact on revenue. Can we infer that major IP companies in hardware security do not provide these solutions directly but prefer to collaborate with you instead? Will these big IP companies actively endorse PUF technology and suggest it to their clients? Charles, please.
Okay, this trend is consistent among most of these IP companies as they have not pursued the development of hardware security continuously. Nevertheless, we are committed to enhancing root of trust technology ranging from 55, 40 nm and down to 5, 4, and 3 nm. Consequently, leading IP companies and ASIC design service providers are working with us across different processes to promote PUF-based solutions to their customers.
[Foreign language]
[Foreign language]
What are the primary uses for MRAM and RRAM? Michael, please.
MRAM and RRAM are both emerging nonvolatile memory technologies, each offering unique benefits and use cases. MRAM is characterized by its high speed, low power consumption, and the ability to handle the numerous rewrite cycles, making it well suited for applications that demand data reliability and quick access, such as IoT devices, smartphones, automotive electronics, and industrial automation. In contrast, RRAM has a simpler design, is cost-effective, and consumes less power, making it ideal for the IoT applications, microcontrollers, and wearable technology. Additionally, as RRAM technology advances, it may find applications in automotive electronics and potentially in AI accelerators and neural network chips in the future. Thank you.
[Foreign language]
[Foreign language]
Would you provide more insights into SRAM repair technology? We understand it will be deployed in advanced nodes such as 5 nm, 3 nm, and 2 nm. When do you expect to see revenue contributions from this technology? Michael, please.
Using our OTP for SRAM repair has always been the main reason why DDI customers choose our OTP IP. They rely on it because their DDI ICs use large capacity SRAM, and to maintain high yield and high good product performance, OTP is necessary. Now we are seeing a similar trend in advanced node for HPC and AI applications. As these high-end digital ICs require large capacity SRAM for fast computation, they will also start adopting OTP as a standard, just like DDI customers do to repair the SRAM. Across various advanced nodes, we have the demands that customers incorporate our OTP in their design for SRAM repair. Thank you.
[Foreign language]
[Foreign language]
With CXMT and Samsung stocking up on DDR4, the resulting oversupply has driven prices way down. This might push DDR4-focused companies to speed up DDR5 developments. Since eMemory's MTP technology is already being used in DDR5 PMICs and SPDs, do you think this shift in the market boosts your business? Has eMemory put any plans or strategies in place for this trend? Michael?
Our OTP and MTP technologies have been adopted by several major customers in DDR5 PMIC and SPD ICs. As DDR5 adoption continues to grow, our licensing fees and royalties will increase accordingly. Thank you.
[Foreign language]
[Foreign language]
Security is a significant trend for decades, and numerous companies are providing cybersecurity software and services. How does your company fit into the broader security landscape? Charles, please.
We offer hardware security solutions based on PUF technology to protect data usage, while many security software and service companies use software for data encryption and security key generation. The use of software method to secure data usage is easy to be attacked through the internet. For a system to be secure, secret keys must be generated randomly and stored in the hardware. Our cutting-edge technology combines both generation and storage of secret keys, providing the most secure foundation for security applications and ensuring data protections.
[Foreign language]
[Foreign language]
EDA companies are now leveraging AI to assist customers in IC design. When engineers select IP and design modules, AI can automatically generate a design layout, which saves a lot of time. As a result, being included as a module in the EDA software becomes important. EDA companies tend to prioritize their own IPs, and unless customers insist on external IPs, the automatically generated design layout will likely default to using their in-house IPs. How do you view this trend? Michael?
The EDA companies primarily apply the AI assistant design to the purely digital IPs using the standard transistors in these designs. In contrast, our IP design uses our own OTP transistors, which are protected by our patents. Unlike the digital IPs, OTP and PUF IPs are specialized analog IPs, making them challenging to design using AI. This is because AI relies on a large database for training, and they don't have sufficient OTP data to enable the effective AI-driven design. Thank you.
[Foreign language]
[Foreign language]
In the interest of time, this is the last question. Trump highlights the importance of making America great again, which may lead to an increase in customers relocating their production to U.S. foundries like Intel or GlobalFoundries. What is the current situation regarding your OTP or PUF implementation in these U.S. foundries? Are your IPs available for customers at all process nodes? Michael?
We have licensed our technology and IPs to foundries I and G for all their process nodes. Furthermore, other foundries like T and U and others also utilize our technology and IP in their process nodes at their fabs in the United States. Thank you.
Thank you, Michael. We will begin the closing comments. Charles, please proceed.
Okay, thank you once again for your patience and support for eMemory. We will continue to work hard on technology and IP innovation and PUF-based hardware security solutions for our customers and bring higher returns for our shareholders. For more information about our PUF-based security IP and technology, we encourage you to visit our PUFsecurity website at www.PUFsecurity.com and check out our articles and other materials. Thank you.
Thank you, ladies and gentlemen. Please be advised that the conference recording will be accessible within the next three hours. Thank you, everyone, for joining us today. We hope you will join us again next quarter. You may now disconnect. Goodbye and have a good day.