Skip to content
The seminar on the implementation plan of the benchmark testing method in microcomputer procurement, hosted by the Chinese Academy of Metrology, was held in Beijing.
This report is a news release by reporter Zhang Mingliu Benchmark testing of computer performance has been widely recognized internationally, but it has not been well accepted domestically. Now, the maturity of technical tools and certification service markets is expected to promote the implementation of benchmark testing in China.
Recently, at the seminar held by the Chinese Academy of Metrology, technical experts and representatives from various institutions such as the National Administration of Government Affairs, the State Administration of Taxation, the Chinese Academy of Sciences, Renmin University of China, the National Quality Certification Center, and the National Electronic Computer Quality Supervision and Inspection Center discussed the practical application of benchmark testing in procurement. The meeting formed the “Implementation Suggestions on the Use of Benchmark Testing Methods in Microcomputer Procurement” (hereinafter referred to as the “Implementation Suggestions”). The “Implementation Suggestions” believe that benchmark testing can be used to determine the minimum performance threshold for shortlisted products in microcomputer procurement, while adopting a “multi-brand” method to select shortlisted products. This approach allows for a comprehensive and intuitive assessment of product performance while providing users with more autonomy in their choices, achieving the procurement goal of “value for money”.
According to the reporter, the current procurement process continues to use hardware configuration-based evaluations, such as CPU core number, frequency, and memory, to assess the performance of procurement objects. Over a decade ago, this method provided an effective objective evaluation system for microcomputer procurement. However, with technological advancements, mainstream hardware manufacturers have adopted comprehensive innovative methods in areas such as process and packaging, architecture, memory and storage, interconnect, security, and software to achieve exponential performance growth. The original evaluation methods can no longer adapt to the performance improvements brought about by innovative overall design concepts, making it difficult to predict and evaluate a computer’s performance in various application scenarios. To address these issues, benchmark testing has emerged and gradually become the mainstream evaluation system for computing and performance.
The “Implementation Suggestions” propose that benchmark testing can be used to determine the minimum performance threshold for shortlisted products. Furthermore, considering that the CPU is the core component affecting performance, and there are two mainstream manufacturers, the “Implementation Suggestions” suggest a multi-brand method based on dual-brand CPUs for product selection, giving users the right to choose.
The specific approach is for the purchaser to determine a reference document list for each budget range based on procurement needs, with each budget range corresponding to a bidding package, allowing multiple brand products to undergo benchmark testing. Products that meet or exceed the threshold value specified for the corresponding package and meet all other requirements listed in the reference document for that package will enter the bidding process.
Experts at the meeting believe that this model can not only achieve the procurement goal of “value for money” but also enhance procurement satisfaction. For suppliers, competition becomes fairer; for purchasers, more choices are available, and personalized needs are met.
It is reported that benchmark testing has been widely applied internationally. However, foreign testing tools have certain limitations in China and have not been widely promoted. Now, the Chinese Academy of Metrology has independently developed various performance benchmark testing tools, among which the benchmark testing tool CPSmark for microcomputers has achieved international advanced levels in various indicators, providing the capability to objectively evaluate the performance of various types of computers in China and offering strong support for performance improvement, optimization, and procurement.
In addition, existing testing service platforms in China are stable and mature, with numerous third-party certification agencies for IT products in the domestic market, which have been providing various types of testing services for various IT products, including microcomputers. Each testing agency has a stable and mature testing technical team. After simple training for technical personnel from the selected testing agency based on government procurement needs, benchmark testing services based on CPSmark can be quickly provided. All of these provide a foundation for the application of benchmark testing in procurement.
To ensure the smooth, fair, and just implementation of benchmark testing plans, the “Implementation Suggestions” also propose several opinions, including opening testing institutions, publicizing testing tools, appropriate supervision of testing institutions, and timely unified upgrades of evaluation tools. Additionally, suggestions were made to use benchmark testing methods to establish a microcomputer e-marketplace product library for this new type of procurement format.
★ This report holds the copyright to this article. If you need to reprint or copy, please indicate that it is sourced from the China Government Procurement Report, credit the author, and maintain the integrity of the article. Otherwise, legal responsibility will be pursued.
