Intel is launching a series of server processors that it says are 1.65 times faster than its previous generation of chips.
The Intel Xeon Scalable chips range from 8 cores, or processing brains, to 28, and they represent the chip giant’s biggest effort to stay ahead of rival Advanced Micro Devices, which recently launched its Zen-based server chips. Both companies say that the new chips are their most competitive in a decade.
Lisa Spelman, vice president and general manager of the data center marketing group, said that the new data center platform will enable almost every industry to improve its core computing foundations through better artificial intelligence processing and other benefits. With this platform, Intel is targeting compute, storage, and communications applications across the data center.
Spelman said the Intel Xeon Scalable platform can support up to 4.2 times more virtual machines and up to five times more transactions per second than four-year-old systems. Configured for storage, it delivers up to five times more operations at up to 70 percent lower latency, or interaction delays. For communications, the platform can power 5G networks with 2.5 times better performance for key networking applications.
AT&T, an early customer, is running production traffic and has seen a major performance improvement using 25 percent fewer servers per cluster. Five customers on the Google Compute Platform have seen up to 40 percent more performance, and more than 100 percent more performance when the application includes optimization for Intel AVX-512 instructions.
Meanwhile, leading content creator Technicolor made it possible to render virtual reality content almost three times faster, freeing up valuable time for its most creative employees.
Intel said that its new Intel Xeon Scalable processors (specifically, the Intel Xeon Platinum 8180) deliver 28 percent faster performance and percent higher core performance than Advanced Micro Devices’ latest Zen-based server chip, the Epyc 7601.
Intel assigned thousands of employees to work on the processors for several years. In that respect, it represents a huge investment.
“This launch signifies an expansion and broadening of Intel’s data center efforts and puts an exclamation point on targeted workloads for enterprises, carriers, HPC, cloud and AI,” said Patrick Moorhead, analyst at Moor Insights & Strategy. “It is important to note that Intel is bringing more than a central processing unit (CPU) to these markets. They are bringing CPUs plus chipsets, accelerators (ie FPGA, QuickAssist, AVX-512), Optane SSD, buses (ie OmniPath) and optimized software (ie DPDK) required for workload-optimization. Intel is even building near-engineered solutions to the table with Select Solutions, an indication that it is moving up the food chain and also enabling a much larger market-basket.”
He added, “By increasing the ‘market basket’ with workload-optimized technologies and Select Solutions, I see growth potential for the company in spaces they haven’t historically achieved high degrees of market share while defending their CPU turf.”