한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

from "quantity" to "quality", breakthrough technologies and operating models are needed:

fan congming said that the investment payback period of intelligent computing centers is generally about five years, but the market changes rapidly, and factors such as electricity charges and operating costs make it difficult to control the end-user cost. du yunlong believes that in the long run, operators of intelligent computing centers should formulate long-term plans to reduce end-user usage costs, cultivate usage habits, and resume pricing in the future; they should focus on cultivating application cases and gradually expand industry coverage.

technology and collaboration promote high-quality development:

the industry's first high-quality computing power evaluation system "artificial intelligence computing power high-quality development evaluation system report" released by inspur information and the academy of information and communications technology pointed out that high-quality computing power needs to be based on the latest artificial intelligence theory, adopt advanced artificial intelligence computing architecture, and high-level computing power that combines algorithms and data depth. however, at present, the gap between the measured performance and theoretical performance of most computing power clusters is too large, and the actual performance of some computing power clusters is less than 10% of the theoretical performance.

"chip" is the key, but the system level is more important: zhang dong believes that simply focusing on chip indicators cannot reflect the true value of the computing center, and the final application scenario needs to be considered from a system perspective. guo liang also said that there are not many chip options at present, but computing and network integration are hot areas. in the future, the number of cards in ai servers will be expanded to improve cluster capabilities and improve overall operation and maintenance efficiency.

the intelligent computing center still needs to break through:

with the development of science and technology, the development model of intelligent computing centers requires continuous innovation. how to balance "quantity" and "quality" and achieve success tests the technical strength, resource advantages and industrial collaboration capabilities of intelligent computing center operators.

future outlook:

as the global intelligent computing market continues to grow, china's intelligent computing market is also showing a rapid development trend. from "quantity" to "quality", we need breakthrough technologies and operating models, and rely on technological innovation and collaborative development to promote the china intelligent computing center to truly play its core role and promote the application of artificial intelligence technology.