Meeting the Growing Demand for Computing Power in the Era of AI with 230% Performance Increase per Watt, Enhanced Rack Density, and Lower Power Consumption.
KAYTUS, a leading IT infrastructure provider, announces that its servers now support the newly released Intel® Xeon® 6 Processors, which offer outstanding performance and energy efficiency, setting a new standard in the industry. Compared to servers equipped with 3rd Gen Intel® Xeon® Scalable Processors, the updated KAYTUS servers deliver a 200% performance boost and a 230% improvement in performance per watt, rigorously verified in virtualization scenarios and tasks such as decompression, video transcoding, and pgbench. The significant enhancement in performance per watt provides greener and more powerful computing support for diverse applications, including AI, cloud-native environments, and data analytics. KAYTUS continues to ensure that its clients benefit from the latest advancements in technology, achieving exceptional efficiency and performance.
With the continuous development of artificial intelligence, particularly the rapid expansion and widespread application of AI technologies such as Large Language Models, the demand for computing power is constantly increasing. It is essential to meet the diverse workload requirements for AI, web and microservices, and data analysis while addressing challenges in architecture design, power supply, rack density, cooling, and sustainability. Many data centers today house over 100,000 server devices, and the expansion in scale introduces challenges related to high density and high energy consumption. To optimize infrastructure utilization, data centers must increase rack density and improve computing power output per watt within limited space. This optimization is crucial to enhancing efficiency and sustainability in modern data center operations.
KAYTUS servers are now upgraded with the latest Intel® Xeon® 6 Processors with Efficient-cores. Compared to those with 3rd Gen Intel® Xeon® Scalable Processors, the updated KAYTUS servers, with up to 144 cores, deliver significantly increased performance and rack density. The servers increased memory bandwidth, and feature excellent input/output capabilities with full-line PCIe 5.0, four Intel UPI 2.0 links of up to 24GT/s, 64 CXL 2.0 channels, and DDR5 6400 memory of 17% speed increase compared to DDR5 5600. This combination provides higher performance and energy efficiency for high-parallel and high-throughput cloud-scale workloads. Numerous test data shows that in virtualization scenarios, KAYTUS servers based on the latest Intel® Xeon® 6 Processors have more than doubled virtual machine density and energy efficiency compared to the 3rd Gen Intel® Xeon® Scalable Processors across various workloads, ensuring unparalleled performance and energy efficiency, for meeting the demands of modern data centers and their diverse applications.
KAYTUS servers boost energy efficiency with cutting-edge innovations in system architecture, hardware, firmware, and cooling designs. In liquid-cooled data centers, the servers feature an intelligent CDU control system that dynamically adjusts the CDU pump speed based on real-time server workload, further enhancing energy savings and reducing emissions. For traditional air-cooled data centers, KAYTUS servers utilize optimized air ducts and cooling strategies to ensure efficient cooling of critical components such as the CPU, GPU, memory, and hard drives. This design reduces overall system power consumption by 13%, markedly improving the servers' overall energy efficiency.
About KAYTUS
KAYTUS is a leading provider of IT infrastructure solutions, delivering a diverse range of innovative, open, and eco-friendly products for cloud, AI, edge computing, and other emerging applications. With a customer-centric approach, KAYTUS is agile and responsive to user needs through its adaptable business model. Discover more at KAYTUS.com
View source version on businesswire.com: https://www.businesswire.com/news/home/20240617279013/en/
Contacts:
Media
media@kaytus.com
