Related Topics:
Tensor Core Graphics Nvidia-
NVIDIA s latest AI server chip
At the 2026 Nvidia GTC conference, Jensen Huang announced an inference-specific chip, the Groq 3 LPU. The LPU will work in concert with the Rubin GPU to accelerate AI workloads. According to TrendForce's latest findings on AI servers, NVIDIA's high-end AI chip shipment mix is expected to change in 2026. This week, over 30,000 people are descending upon San Jose, Calif., to attend Nvidia GTC, the so-called Superbowl of AI—a. Nvidia's Blackwell Ultra chips, the company's next-generation graphics processor for AI, have been commercially deployed at CoreWeave, the companies announced on Thursday. CoreWeave historically has a close relationship with Nvidia, which owns a stake in the cloud provider. CoreWeave went public. The Rubin platform harnesses extreme codesign across hardware and software to deliver up to 10x reduction in inference token cost and 4x reduction in number of GPUs to train MoE models, compared with the NVIDIA Blackwell platform. NVIDIA Spectrum-X Ethernet Photonics switch systems deliver 5x.
[PDF Version]
-
Multiple network cables stacked on the core switch
This is achieved by using stacking-capable switches which have dedicated ports and use dedicated cables to connect to other switches in the stack. Up to eight. Switch stacking is a feature of certain Cisco access layer switches which allows for the creation of a single logical device from many individual devices via a backside stack port connected by several stack cables. The major benefits of stacking. This article explains what switch stacking is, how stacking works, its advantages and disadvantages, why Asterfusion is moving away from stacking, and alternative solutions — and shows how we address the challenges modern network designs face due to stacking. Each switch will use its own MAC address table to make frame-forwarding decisions.
-
How to make a joint for optical fiber and copper core cable
Fiber optic splicing creates an accurate connection between fiber cores and involves delicate operations such as fiber stripping, fiber cleaving, core aligning and coupling, etc. However well you plan your installation, fiber cable is rarely the right length for each run, and is inherently difficult to join. Consequently, cables have to be connected or cut in the field, with the potential issues this entails. This blog post looks at the various options available to. There are two methods of fiber optic splicing, fusion splicing & mechanical splicing. Either joining method must have three primary characteristics. At the heart of any robust fiber optic network lies a crucial process: Preparing a fiber cable for termination of a connector or splice. What is Fiber Optic Splicing and Why is it Needed? – #1.
[PDF Version]
-
Can a core switch be configured with two IP addresses
The switch can have multiple IP addresses. Each IP address can be assigned to specified interfaces or ports, Link Aggregation Groups (LAGs), or Virtual Local Area Networks (VLANs). Yes, it is possible to have two core switches with the same SVIs (Switched Virtual Interfaces) configured. Firewall. This document describes the configurations of IP Service, including IP address, ARP, DHCP, DHCP policy VLAN, DNS, mDNS gateway, mDNS relay, UDP Helper, IP performance optimization, IPv6, DHCPv6, IPv6 DNS, IPv6 over IPv4 tunnel, and IPv4 over IPv6 tunnel. This is useful when deploying IP phones! To establish if your core switch is providing DHCP, login to it and enter: sh run | s dhcp Example with two pools for two TR's. Location names are factory-1 and factory-2. Access switches of the two locations are connected via fiber to their. In networking, switches play a crucial role in ensuring seamless communication between devices within a local area network (LAN). Layer 2 switches, in particular, operate at the data link layer of the OSI model and are primarily responsible for forwarding data packets based on MAC addresses.
[PDF Version]
-
How to access Huawei s core switch
Connect to the Huawei switch using SSH or the console port. Use the following commands to set a port to trunk mode. For example: Replace USERNAME with the new username, set the password, define service-type (telnet, ssh, etc. ), and specify. This document describes how to log in Huawei S series switches for the first time and fast configure the switches, and describes the procedures of configuring unconfigured switches in three scenarios: Small-Sized Campus Networks, Small- and Mid-Sized Campus Networks and Mid-sized Campus WLANs. Before You Start This document will help you log in to and quickly configure Huawei S series switches.
-
What type of cable is used in the core switch
What type of cabling is typically used with a core switch? Fiber optic cabling is typically used with a core switch due to its high bandwidth capacity and long-distance capabilities., Cat6a or Cat7) may be used for shorter distances or in smaller networks. The hierarchy Ethernet network is a three-layer integrated setup of networking devices. Choosing the right cable ensures reliability, reduced interference, and future-proof. Q: What is a core switch, and how is it different from a standard switch? Q: What are the principal distinctions between a core switch and an ordinary switch? Q: What does a core switch do in a high-capacity core network infrastructure? Q: What is the role of the core layer in the network? Q: Why. It is a powerful backbone switch in the center of the network core layer, which centralizes multiple aggregation switches to the core and implements LAN routing. In these switches, the data routed and switched.
[PDF Version]
-
AI is a server
An AI server is a specialized computing system built to handle machine learning workloads – model training, inference, and data processing – at a scale that standard servers can't support. They provide the hardware environment —. AI, or artificial intelligence, is changing the way organizations and businesses handle data by incorporating automation of complex calculations, introducing new advanced applications, and fulfilling computational demands like never before. This is where AI server clusters stand out, crafted for. Modern AI models are data-hungry, computation-heavy beasts that need specialized hardware just to function, let alone perform at their best. If you're running LLM inference, computer vision pipelines, or anything that touches GPU-accelerated compute. Unlike traditional servers designed for general-purpose computing tasks such as hosting websites or managing databases, AI servers are specialised systems engineered to handle the specific computational demands of AI workloads. These supercomputing systems are designed to execute complex.
[PDF Version]
-
Huijue Information AI Server Chip
The system, launched at the World AI Conference in Shanghai, uses 384 Ascend 910C chips, significantly outnumbering Nvidia's 72 B200 GPUs in the GB200 NVL72. China's domestic AI chips took 41% of the accelerator server market in 2025. New data shows Huawei alone shipped roughly 812,000 AI chip units last. Dozens of Chinese hi-tech manufacturers - from Lenovo Group and Huawei Technologies to Inspur Group - are pushing new "all-in-one" servers that include DeepSeek 's advanced artificial intelligence (AI) models to private and public enterprises across the country, ramping up democratisation of the. Huawei has started reclaiming its growth and influence in Chinese server business due to increasing demands for its AI chips. The firm is once again coming into power for its server business and pushing back its rivals like Digital China Group. A few industry analysts reported that Huawei is. Huawei's computing business includes Kunpeng for general servers and Ascend for AI computing.
[PDF Version]
-
Core Switches in the Data Center Layer 3
Core layer—Provides the high-speed packet switching backplane for all flows going in and out of the data center. A core switch is a high-capacity, high-performance Layer 3 switch positioned at the physical backbone of an enterprise network. Engineered to aggregate massive volumes of data from distribution switches, it provides ultra-low latency and maximum throughput to ensure uninterrupted routing and packet. Important—Updated content: The Cisco Virtualized Multi-tenant Data Center CVD ( This determines network efficacy, dependability, and the speed at which. Data center-grade switches are characterized by high-quality business assurance and control recognition capabilities. They feature end-to-end flow control and backpressure mechanisms, ensuring stable and reliable data transmission, and smoothing out network surges.
[PDF Version]
-
Where to connect the fiber optic quick connector core
Inserting the Fiber: Carefully insert the cleaned fiber core into the LC fiber connector, ensuring it fully enters the connector and aligns with the internal metal contact faces., V-groove clamp) to secure the fiber firmly inside the connector. It eliminates the need for time-consuming and complex fusion splicing techniques, making fiber optic fast connec. A correct installation creates a low-loss, reliable connection essential for high-speed data transmission. While fiber optics enable speeds and distances copper can't match, the system's performance hinges. A Fiber Optic Fast Connector is a revolutionary component in the telecommunications industry, designed to simplify the process of terminating fiber optic cables in the field.