Key Highlights:
- In Bluefield, 2 Nvidia paired its DPU with the NGFW VM Series.
- Intel has launched a chip to process its infrastructure.
- 80% of network traffic in a data center cannot be thoroughly checked by a firewall
Nvidia Technology’s DPU Industry Accelerates
In larger IT companies, DPU technology is garnering greater attention. Although Nvidia has taken a lot of attention, it is facing increasing competition. Nvidia combined its Data Processing Unit (DPU) in the Bluefield 2 in the third week of July with NGFW VM Series. From Palo Alto Network to dramatically speed up virtual firewall performance.
Intel launched its infrastructure processing unit chip the previous month. A configurable network gadget that allows cloud and communications companies to discharge some CPU operations. So, this chip increases the performance of excessive CPUs to provide a useful structure between computing, memory, and protection or networking.
Dan Newman, Futurum Research’s senior analyst, and founder says: “There will be more firms joining Nvidia, Intel including Marvell and AWS with Nitro. These enterprises will sell in their advantages, but they can all do more for general CPUs by downloading workloads such as security, storage, and networking from the CPU. New technologies are simply placing too much load on CPUs.”
Author’s Advice on Next-Generation Firewall
“Telcos and cloud service providers are likely to be the early adopters,” said Deierling. We expect it to grow when people are moving into a society in general. Where each server needs a next-generation firewall, from defending the data center perimeter using traditional devices.
A firewall does not need or cannot adequately examine 80% of network traffic in a data center, officials in both firms stated. In this way, the Intelligent Traffic Offload (ITO) Service is jointly developing to examine network traffic and determine which sessions will later benefit from a safety evaluation.
Whenever the firewall determines that it is not safe to investigate a session. So, the ITO system allows the Bluefield-2 DPU to tell that all packets. So, these packets should be sent straight to their destination in that session and not to the firewall. Deierling added, Nvidia will also be targeting other areas, like financial and healthcare communities. Which places a strong emphasis on data protection.
Nvidia Focuses on Three Factors
Nvidia worked on three things offloading, acceleration, and isolation.
- The offloading portion assumes the duty of the x86 chip and runs it on the DPU.
- The increasing acceleration derives from the processing tasks of CPU and DPU sharing.
- The isolation consists of the creation of a new DPU processing infrastructural layer on the GPU, which isolates it from the interface of the application.
Also Read: ON Semi-Conductor collaborates with NVIDIA for Cloud based Autonomous Vehicle Simulation
