NetLogic Microsystems, a provider of intelligent semiconductor offerings, has extended its XLP processor family with the introduction of XLP316 processor.
The XLP316 processor features quad-core, 16-issue, 16-threaded, superscalar processor architecture at up to 2.0GHz with out-of-order execution and superior L3 cache to address the control-plane processing requirements in communications, networking, storage and security applications.
The new processor is suited to address new levels of control-plane performance requirements by combining a multi-core processor architecture with a high-performance floating point unit and a low latency tri-level cache architecture, said the company.
In addition, when compared to competing quad-core processors with dual-issue and single-threading per core, in-order execution and missing L3 cache, the XLP316 delivers higher control-plane processing performance, the company claimed.
The XLP316 multi-core processor integrates 16 NXCPUs, which are fully cache and memory coherent for software applications to run in Symmetric Multi Processing (SMP) or Asymmetric Multi Processing (AMP) modes.
The 16 NXCPUs are interconnected via the company’s high-speed, low-latency Enhanced Fast Messaging Network to support billions of in-flight messages and packet descriptors between all on-chip elements, said the company.
Further, the processor offers a tri-level cache architecture with 4 Mbytes of L3 cache and over 6 Mbytes of fully coherent on-chip cache which delivers 40 Terabits per second (Tbps) of extremely high-speed on-chip memory bandwidth.
The XLP316 multi-core processor integrates networking interfaces including: PCI-Express Gen2 4×1 or 1×4; eight Gigabit Ethernet (SGMII) channels; dual XAUI ports; USB2.0; XEN Hypervisor virtualisation and I/O virtualisation (SR-IOV).
NetLogic Microsystems marketing vice-president Chris O’Reilly said having the only quad-issue, quad-threaded communications processor family with on-chip L3 cache gives them an enormous competitive advantage to address the most demanding control-plane processing requirements.
"The combination of the superior and highly optimised architecture along with the advanced 40nm process gives us further advantage in delivering the lowest power profile to meet our customers’ stringent requirements," O’Reilly said.
