Features - xFusion Digital Technologies FusionServer CH242 V5 Manual

FusionServer CH242 V5 Compute Node
Technical White Paper
Performance and Scalability
⚫
⚫
⚫
⚫
⚫
⚫
⚫
⚫
2022-08-12
®
Powered by Intel Xeon
provides up to 28 cores, 3.8 GHz frequency, a 38.5 MB L3 cache, and three 10.4 GT/s
UPI links between the processors, which deliver supreme processing performance.
− It supports four processors with 112 cores and 224 threads to maximize the
concurrent execution of multithreaded applications.
Intel Turbo Boost Technology 2.0 allows processor cores to run faster than the
−
frequency specified in the Thermal Design Power (TDP) configuration if they are
operating below power, current, and temperature specification limits.
− Intel Hyper-Threading Technology enables each processor core to run up to two
threads, improving parallel computation capability.
− The hardware-assisted Intel® Virtualization Technology (Intel® VT) allows
operating system (OS) vendors to better use hardware to address virtualization
workloads.
− Intel® Advanced Vector Extensions 512 (Intel AVX-512) significantly accelerates
floating-point performance for computing-intensive applications.
− The Cascade Lake processors support Intel® Deep Learning Boost vector neural
network instructions (VNNI) to improve the performance of deep learning
applications.
The CH242 V5 fully configured with 48 LRDIMMs provides supreme speed, high
availability, and a maximum memory capacity of 6 TB.
The maximum theoretical memory bandwidth is 550 GB/s (64 bit/8 x 2933 MHz x 6
channels x 4), which is 83.3% higher than that of the previous-generation platform.
The use of all solid-state drives (SSDs) is supported. An SSD supports up to 100 times
more I/O operations per second (IOPS) than a typical hard disk drive (HDD). The use of
all SSDs provides higher I/O performance than the use of all HDDs or a combination of
HDDs and SSDs.
The compute node supports 192-lane PCIe 3.0 (8 GT/s per lane) to provide 20% higher
I/O bandwidth than the previous 160-lane PCIe.
With Intel integrated I/O, the Intel
controller to shortens I/O latency and improve overall system performance.
The compute node supports mezzanine cards with a variety of network ports to meet
requirements for flexible networking.
The compute node supports standard PCIe SSD cards.
®
Scalable Skylake or Cascade Lake processors, the server
® ®
Xeon Scalable processors integrate the PCIe 3.0

2 Features

2
Features
8