Intel’s New Xeon E-2300 CPUs Bring Rocket Lake To Entry-Level Servers
The chipmaker says the Xeon E-2300 CPUs are a good fit for on-premises servers used by small businesses and entry-level bare metal instances provided by cloud service providers. The new chips offer up to a 17 percent improvement in performance over the Xeon E-2200 processors that launched in 2019 and double the memory enclave capacity for Intel Software Guard Extensions, which isolates confidential data in private memory regions, according to Intel.
Intel has adapted its Rocket Lake desktop CPU architecture for a new line of Xeon E processors that are designed for entry-level servers in the small business and cloud services segments.
The Santa Clara, Calif.-based company launched the 10 Xeon E-2300 CPUs for single-socket servers on Wednesday, saying they provide up to a 17 percent improvement in performance over the Xeon E-2200 processors that were made generally available in 2019. The processors are being supported by Dell Technologies, Hewlett Packard Enterprise, Lenovo, Supermicro and other server vendors. Intel offered no performance comparisons with its x86 CPU rival, AMD.
[Related: Intel Heralds New Era With Alder Lake Hybrid CPU Architecture]
The processors are available in four-, six- and eight-core configurations — unchanged from the previous generation — with thermal design power ranging from 65-95 watts. They support base clock speeds of up to 3.7GHz and turbo boost speeds of up to 5.1GHz using Intel Turbo Boost Technology 2.0.
The recommended customer pricing ranges from $539 for the top-line eight-core Xeon E-2388G to $182 for the four-core Xeon E-2314. The two lowest priced CPUs, which includes the Xeon E-2314 and the $209 Xeon E-2324G, don’t support hyper-threading and represent new price points for the Xeon E brand.
New features for this Xeon E generation include CPU-level support for up to PCIe 4.0 connectivity, two channels of DDR4-3200 ECC memory and a single digital display, thanks to the inclusion of Intel’s Gen12 integrated graphics within the processors. The total memory capacity supported is 128GB.
The new processors also feature Intel Active Management Technology, a remote management capability part of the Intel vPro platform. Another crucial feature is Intel Software Guard Extensions, or SGX, a silicon-level security feature that isolates applications with confidential or sensitive data in private memory regions.
While Intel SGX appeared in Intel’s Xeon Scalable CPUs for the first time with the release of the third-generation lineup earlier this year, the feature was present in the previous-generation Xeon E-2200 chips. The difference with the new Xeon E-2300 CPUs is that they support 512MB in memory enclave capacity, double what the previous Xeon E generation offered.
The Intel C250 Series chipset supports up to 24 additional lanes of PCIe 3.0 connectivity, up to eight SATA 3.0 ports, up to 10 USB 3.2 Gen 2x1 ports and up to three USB 3.2 Gen 2x2 ports. The latter two are new to this Xeon E chipset and support throughputs of 10Gbps and 20Gbps, respectively, which are double what the previous chipset generation supported.
Intel said the Xeon E-2300 CPUs, code-named Rocket Lake E, are a good fit for on-premises servers used by small businesses and entry-level bare metal instances provided by cloud service providers.
Like the generation before it, the new Xeon E-2300 lineup is based on a desktop CPU architecture that has been modified for server use. In this case, it’s based on Rocket Lake, which powers the 11th-generation Intel Core S-Series. The architecture uses Intel’s Cypress Cove microarchitecture, which is a variant of the Sunny Cove architecture developed for 10-nanometer sever and laptop chips that has been backported to the company’s old 14nm manufacturing process.