In the early hours of this Monday (26), Intel revealed its new Xeon line processors, aimed at next generation computing, during MWC 2024, in Barcelona. They consist of Sierra Forest and Granite Rapids-D chips.
In addition to them, the company also presented a new network edge platform, aimed at scalable Artificial Intelligence applications. TudoCelular followed the company's news and details it for you.
Sierra Forest
Sierra Forest processors will have up to 288 cores and will use Intel's E-core technology. It will be focused on helping vRAN network operators gain more performance for 5G workloads. The estimate is 2.7 times greater performance per rack.
These CPUs will also have an updated version of the Intel Infrastructure Power Manager software, so that customers will be able to explore integrated telemetry in order to reduce energy consumption, without losing performance in metrics such as transfer rate and latency.
According to Intel, Sierra Forest chips are scheduled to become available in the second half of this year. There are already network operators and partners interested in them, such as BT Group, Dell, Ericsson, HPE, Lenovo, KDDI Corp. and SK Telecom Co., Ltd.
Granite Rapids-D
Designed for servers of the future, Granite Rapids-D will support virtual radio access network workloads. These chips focus on doubling the capacity of vRAN payloads compared to their predecessor.
In practice, this means that network operators will be able to double the number of cell sites operated or subscribers supported, while being able to reduce energy expenditure by up to 20% in vRAN computing.
According to Intel, the considerable increase in performance with these chips will make customers expand networks on a larger global scale, as vRAN costs will drop. To achieve this improved performance, the processor will utilize Intel AVX and integrated Intel vRAN Boost acceleration.
The expectation is that Granite Rapids-D CPUs will only be available in 2025. For now, they have been tested in laboratories by customers such as Samsung, Dell, HPE, Lenovo and Red Hat.
Until the debut occurs, the company will release the vRAN AI development kit in early access to select partners. Participants will be able to help customers create, train, optimize, and deploy AI models for use cases on supported servers.
Tech 07 Dec
Economy and market 30 Mar
Intel Edge Platform
Completing the announcements is the Intel Edge Platform, a platform designed for deploying Artificial Intelligence at the edge of the network. It focuses on enabling customers to leverage vast amounts of data generated at the edge to improve customer experience and scale business operations through automation.
It is a modular open source platform to make developing, deploying and managing cloud-like AI applications simpler.
The intention is for developers to be able to use the platform to run latency-sensitive workloads on the most suitable device, remotely, for better performance.
Among the services possible with the technology are data annotation for building more AI models, as well as specific industrial applications in video, temporal data and so-called digital twins.
The Intel Edge Platform will begin general availability before the end of this quarter, as estimated by the company.
Did you like Intel's announcements at the start of MWC 2024? Comment with us!