[Acompany presents a paper at an academic conference]
Acompany develops secret ID matching method “SGX-EIM”! Enables secure calculation of 10 million pieces of data in about 11 seconds
*View in browser* *Acompany Co., Ltd.*
Press release: February 9, 2024
Acompany develops secret ID matching method “SGX-EIM”! Enables secure calculation of 10 million pieces of data in about 11 seconds *Presentation by R&D researcher Aoi Sakurai at “2024 Cryptography and Information Security Symposium (SCIS2024)”*
Acompany Co., Ltd. (Nishi-ku, Nagoya City, Aichi Prefecture, Representative Director and CEO), which provides data clean rooms Ryosuke Takahashi (hereinafter referred to as Acompany) developed the secret ID matching method “SGX-EIM”.
SGX-EIM combines high-speed computing power and security. The calculation speed is about 11 seconds for ID matching and
cross-tabulation of 10 million data items using secure calculation, which is fast enough. In addition, safety is provided by Intel(R) By utilizing the protection area provided by SGX, it serves as a strong security and safety control measure. These two features contribute to the need to safely calculate and utilize large amounts of personal data.
This result was announced at the “2024
Presented at the Symposium on Cryptography and Information Security (SCIS2024).
In addition, Acompany is promoting the utilization and proposal of Privacy Tech (PETs) such as secure computation and federated learning for business operators as “PETs
We provide this as an R&D support service. The SGX-EIM announced this time can be used in this PETs R&D support service.
PETs R&D support service * ■What is “SGX-EIM” *
“SGX-EIM” is “Intel(R)
This is a secret ID matching method that uses the protection area provided by SGX to quickly and safely analyze and process large amounts of data. It is possible to horizontally combine data and generate cross-tabulation tables from data held by two businesses. * Enables high-speed secure calculation of large amounts of data* SGX-EIM can match data and perform cross-tabulation in 1.8 seconds for 1 million records and approximately 11 seconds for 10 million records.
In addition, for this presentation, we measured the execution time for horizontal data joins and cross-tabulation table generation for 1,000 to 10,000,000 items. As a result, in the case of matching processing between 10 million items, which is the maximum amount of data, we succeeded in completing all processing, including processing, encryption processing, and data transfer, in 13.6 seconds for horizontal data joins and 11.5 seconds for cross tabulation tables. Did.
* Analytical processing is performed in a safe environment and is useful as a safety management measure*
SGX-EIM uses secure calculation in the calculation process, so that not only the provider of personal data but also the engineer performing the analysis cannot view the analysis process of personal data. Also, Remote
The protocol * allows us to guarantee that the program defined to perform secure calculations with SGX-EIM will work. Therefore, it is possible to prevent third-party attackers from using it for unexpected purposes, such as changing the program.
By using SGX-EIM in this way, it is possible to safely analyze and process personal data, and it is considered a method that can achieve higher privacy protection than normal security management measures.
What is Attestation? Check whether a user using a remote non-SGX machine is performing any unauthorized operations such as tampering with the verification of the remote SGX machine, or whether the machine itself has any SGX-related vulnerabilities. This is to verify that you do not have any problems. Detail is,
“Data clean room company Acompany releases Intel(R) SGX’s TEE technology, RA and LA OSS”
Check out * ■What is Privacy Tech (PETs)?*
Privacy Tech is a technology to protect individual privacy. In modern times, as large amounts of personal data are held by companies and other organizations, it has become necessary for individuals to securely preserve and utilize their data. Privacy Tech was developed to solve these issues. For example, technologies such as “secure computation” that allows advanced analysis to be performed while encrypting data, “synthetic data” that generates similar data from the original data, and “k-anonymization” that makes it difficult to identify individuals. there is.
Privacy Tech Lab: https://acompany.tech/privacytechlab/
* ■Company profile*
Company name: Acompany Co., Ltd.
Representative: Ryosuke Takahashi, Representative Director and CEO Address: Nagono Campus, 2-14-1 Nagono, Nishi-ku, Nagoya, Aichi Prefecture Established: June 2018
Business content: Development and provision of data clean room “AutoPrivacy” where all data can be used safely, consulting
* ■Inquiry regarding this matter*
Please contact us using the inquiry form or the email address below. Contact form: https://acompany.tech/contact/
*About details about this release*
*Download press release materials*
Unsubscribe HTML email