Accelerate Finite Element Analysis with Next-Generation Server Hardware

Faster completion of complex finite element workloads saves time and money for simulation-based prototyping of everything from tiny electronics components to massive industrial machinery.

The combination of NX Nastran*—part of Siemens PLM Software’s SimCenter* portfolio—and Intel® Xeon® Scalable processors enables an advanced design environment.

Design and engineering processes based on finite element analysis continue to grow in size and complexity, in an effort to provide greater fidelity to products in the physical world. With that progression, users of tools such as NX Nastran have an ongoing need for greater compute power to enable larger simulations within shrinking product-development timeframes. They must identify synergies between software and hardware that can keep pace with the innovation their success depends on.

Siemens PLM Software and Intel drive solutions to that challenge through a co-engineering relationship that helps ensure that NX Nastran takes full advantage of each successive generation of Intel® Xeon® processors. That collaboration continues with the introduction of the new Intel® Xeon® Scalable processors. This latest generation drives more robust results from the NX Nastran environments operated by manufacturing and engineering firms in aerospace, automotive, electronics, heavy machinery, medical device, and other industries.

Faster Results from More Complex Models
In a test environment, Intel performance engineers demonstrated up to a 1.52X speedup on NX Nastran V11 standard benchmarks when comparing the Intel® Xeon® Gold 6134 processor to the Intel® Xeon® processor E5-2680 v3.1 Given that the older processor was introduced just three years prior to the newer one, in 2014, the systems that underlie this performance difference may be similar to those in use by many customers. In fact, even compared to the Intel® Xeon® processor E5-2697 v4 (introduced in 2016), the newer processor provided an increase of up to 1.3X. 1

In addition, it is noteworthy that where the baseline processor has 12 cores per socket and the next-generation one has 18, each Intel Xeon Gold 6134 processor has just eight cores per socket. The ability to deliver the performance increases found in this testing with a lower number of processor cores demonstrates the significant increase in performance per core with the new processor.

These performance increases could allow customers to run more complex simulations in the same amount of time as they previously ran simpler ones, or else to run their existing simulations in less time than previously. In either case, the increased efficiency helps keep companies competitive as they strive to produce safe, reliable, and optimized designs under continuing pressure to compress development cycles and get products to market more quickly.

Synergies with a New Hardware Microarchitecture
Intel Xeon Scalable processors are built on a new microarchitecture, based on 14-nanometer process technology that represents the greatest platform advance in the past 10 years. The processors support server architectures with two, four, or eight sockets, for a new level of flexibility to scale systems up cost-effectively as business needs grow.

A redesigned memory subsystem and cache hierarchy helps improve the efficiency of data handling at both the processor and the system levels. Six memory channels per socket (up from four in its predecessor) increases bandwidth to system memory, helping keep data close to the processor where it is needed. Support for up to 24 DDR4-2667 DIMMs provides the Intel Xeon Gold 6134 processor with a capacity of up to 1.5 TB of RAM per socket.

The new Intel® Ultra Path Interconnect provides greater inter-processor bandwidth and reduced latency compared to predecessors, adding further to the data handling improvements available from the new platform. Up to 48 PCI Express* lanes per socket (an increase of 20 percent over its predecessor) enhances system-level I/O to complement the processing and memory advances and helping to deliver a balanced platform overall. Customers running NX Nastran benefit from system refresh with the ability to generate more sophisticated simulations based on expanded data sets, for a more robust product-engineering process overall.

Siemens PLM Software’s NX Nastran has been an industry standard finite element solver for more than 40 years. Companies that depend on it as a foundational part of the design and engineering toolchain now have the ability to drive up performance dramatically through system refresh using servers based on the Intel Xeon Scalable processors. Flexible system scale-up in the future helps solidify the value of this approach to future-proofing operations.

Informações de produto e desempenho


NX.Nastran V11 standard benchmarks tested by Intel, June, 2017.

Baseline: 2x Intel® Xeon® processor E5-2680 v3 @ 2.5 GHz, Intel® Turbo Boost Technology and Intel® Hyper-Threading Technology enabled, BIOS SE5C610.86B.01.01.0016.033120161139, 128 GB total memory (8x 16 GB, 2333 MT/s, DDR4), Red Hat Enterprise Linux* 7.1 kernel 2.6.32-431.el6.x86_64, 1x Intel® SSD DC P3700 Series (800 GB).

Next-generation: 2x Intel® Xeon® processor E5-2697 v4 @ 2.3 GHz, Intel Turbo Boost Technology and Intel Hyper-Threading Technology enabled, BIOS SE5C610.86B.01.01.0016.033120161139, 128 GB total memory (8x 16 GB, 2400 MT/s, DDR4), Red Hat Enterprise Linux* 7.3 kernel 3.10.0-229.20.1.el6.x86_64.knl2, 1x Intel® SSD DC P3700 Series (800 GB).

New: 2x Intel® Xeon® Gold 6134 processor @ 3.2 GHz, Intel Turbo Boost Technology and Intel Hyper-Threading Technology enabled, BIOS version 412, 256 GB total memory (16x 16GB, 2666 MT/s, DDR4), Red Hat Enterprise Linux* 7.3 kernel 3.10.0-514.el7.x86_64 Intel® SSD DC P3700 Series (800 GB).

2Performance results are based on testing as of June 2017 and may not reflect all publicly available security updates. See configuration disclosure for details. No product can be absolutely secure.