Nissan is turning to the cloud to help
it bring new car models to market faster. As a result, it is migrating its
on-premises, high-performance
computing (HPC) workloads
to Oracle Cloud Infrastructure.
Nissan says it relies on a digital
product design process to make quick and critical design decisions to improve
the fuel efficiency, reliability and safety of its cars. By moving its
performance and latency-sensitive engineering simulation workloads to Oracle Cloud,
it will therefore be able to speed the design and testing of new cars.
Specifically, Nissan uses software-based
Computational Fluid Dynamics (CFD) and structural simulation techniques to
design and test cars for external aerodynamics and structural failures. Oracle
Cloud Infrastructure’s compute, networking, and storage services optimised for
HPC applications will allow Nissan to benefit from a “bare-metal HPC”
solution as it innovates cars. Nissan says it anticipates higher performance
and lower costs with the ability to run engineering simulation workloads
in the cloud.
“Nissan is a leader in adopting
cloud-based high performance computing for large scale workloads such as safety
and CFD simulations,” said Bing Xu, general manager of the engineering systems
deepartment at the Nissan Motor Company. “We selected Oracle Cloud
Infrastructure’s HPC solutions to meet the challenges of increased simulation
demand under constant cost savings pressure. I believe Oracle will bring
significant ROI to Nissan.”
Oracle
Cloud Infrastructure gives companies very high performance compute, fast
networking, and fast memory access, which all suit HPC workloads, such as
Computational Fluid Dynamics and structural simulations, that help design and
test cars for aerodynamics and structural failures.
Running HPC workloads in the cloud helps companies rapidly deploy HPC
infrastructure wherever it’s needed, and only pay for what they consume. It is
also beneficial for companies that need to ship data between cloud regions or
from the cloud to their own data centers.
Nissan has adopted a cloud-first
strategy for its HPC platform to ensure its engineers always have the compute
capacity needed to run their complex simulations. While the HPC market has been traditionally underserved by
public cloud providers, Oracle Cloud Infrastructure delivers an Intel
Xeon-based bare-metal compute infrastructure with RDMA cluster networking, a
set-up that promises latencies of under two microseconds and 100 Gbps
bandwidth, enabling large scale HPC migrations to the cloud.
Nissan is one of the first automotive
OEMs to leverage GPU technology in Oracle Cloud Infrastructure for structural
simulation and remote visualisation. By using bare-metal GPU-accelerated
hardware, says Oracle, Nissan reduces the cost and overhead of large data
transfer, while ensuring that all the data generated by simulation jobs can
easily be viewed in 3D OpenGL format in the cloud.
In addition to HPC workloads, Oracle
Cloud Infrastructure supports a mature and diverse ISV application ecosystem
across different domains, such as CFD and structural simulation. This helps
deliver a price/performance ratio that is more compelling than running
on-premises or compared to other public cloud providers.
With Oracle Cloud Infrastructure, Nissan
is able to launch tens of thousands of cores and GPU-based high-end
visualisation servers with tremendous flexibility, enabling them to dynamically
change compute and remote 3D visualisation, based on the needs of its
engineers.
“Oracle is excited to work alongside
Nissan to change digital product design and development, and help them build
the next generation of award-winning vehicles,” said Clay Magouyrk, executive
vice president of Oracle Cloud Infrastructure. “Our mission has always been to
build the best cloud infrastructure for enterprises, including computationally
intensive and extremely latency sensitive workloads that organizations like
Nissan need to build the next generation of vehicles.”
