Tag: VM

Red Hat announced an expanded strategic partnership with AMD, combining Red Hat's leadership in open-source and AI software platforms with AMD's deep expertise in high-performance computing architectures. This collaboration aims to provide enterprise users with more flexible and performance-advantageous AI and virtualization solutions to address the increasingly diverse needs of hybrid cloud and generative AI deployments. Regarding AI workloads, Red Hat stated that its OpenShift AI solution now fully supports AMD Instinct GPUs, allowing enterprises to accelerate performance through high-performance GPUs without significant resource investment and quickly import and deploy generative AI models across hybrid cloud architectures. The two companies also successfully demonstrated small and large language model inference capabilities based on Red Hat Enterprise Linux AI on the Microsoft Azure ND MI300X v5 platform, utilizing AMD Instinct MI300X GPUs. This demonstrates the ability to deploy multiple GPUs on a single virtual machine, effectively reducing the complexity and cost of multi-VM deployments. Regarding vLLM (a high-performance, open-source language model inference framework), Red Hat and AMD are actively collaborating with the upstream community to improve model inference performance and enhance collective communication and load scheduling in multi-GPU environments by optimizing the Triton core, supporting FP8, and upstreaming the core function library. Red Hat will also release its own enterprise-grade vLLM distribution, "Red Hat AI Inference..."

Netflix recently filed a lawsuit against Broadcom, alleging that VMware, the cloud computing company it acquired, infringed on its virtual machine technology patents. According to Reuters, Netflix claims VMware infringed on at least five of its patents related to virtual machine operation, including VMware's vSphere virtualization platform used for deploying and managing virtual machines. Neither Netflix nor Broadcom has responded to the allegations. Previously, Broadcom had also accused Netflix of infringing on its video streaming technology patents in 2018, filing lawsuits in California, Germany, and the Netherlands. The U.S. trial is scheduled for June 2025.

Following Parallels' announcement that Parallels Desktop 18 for Mac would allow Macs with Apple Silicon processors to install Windows 11 operating systems compatible with Arm architecture environments, VMware also announced that its new Fusion 13 can meet the same requirements, even supporting the OpenGL 4.3 API, allowing new Macs to handle higher display performance. With OpenGL 4.3 API support, Windows 11 running in a virtual environment can stably run DirectX 11-compatible applications and games. However, due to hardware limitations, Windows 11 installed on new Macs via Fusion 13 still cannot use external graphics cards. Windows 11 running in this configuration will be compatible with both Win32 and 64-bit architecture applications. The new Fusion 13 also comes in two versions: Player and Pro. The former is available for general consumer users and is sold for $149, or can be upgraded for $79. The latter is targeted at enterprise users and is priced at $199, with upgrades available for $99. Compared to the Player version, the Pro version adds features such as the ability to create and manage encrypted virtual machines, support for customized virtual networks, and connectivity to vSphere/ESXi server applications.

Following its 2020 collaboration with VMware on Project Monterey and the unveiling of the BlueField DPU design, NVIDIA announced at VMware Explore 2022 a new data center solution in partnership with Dell, designed for artificial intelligence applications. This solution integrates Dell PowerEdge servers, NVIDIA's BlueField DPU and GPU design, and leverages VMware's new vSphere 8 enterprise workload platform and NVIDIA AI Enterprise software resources to accelerate AI training and inference applications while meeting data processing, data science, and zero-trust security requirements. NVIDIA emphasized that with the assistance of the BlueField DPU, server performance can be fully utilized on the new vSphere 8 enterprise workload platform. Specifically, network resource allocation and security are handled by the BlueField DPU, allowing the server's CPU to handle more computational demands while ensuring sufficient performance. This also enables distributed workload operation, improving data transfer efficiency and reducing latency. With VMware announcing the new version of its vSphere 8 enterprise workload platform, NVIDIA also announced that its NVIDIA AI Enterprise software will fully support vSphere 8. This means that NVIDIA-certified Dell PowerEdge servers will be compatible with the new AI-ready enterprise platforms from both NVIDIA and VMware, meeting the needs of enterprises deploying data centers and public cloud applications. Manuvir Das, head of NVIDIA's Enterprise Computing Group, stated, "With the surge in compute and network workloads, AI and zero-trust security are powerful forces driving global enterprises to restructure their data center architectures. VMware vSphere 8 offloads data center infrastructure services to NVIDIA BlueField DPUs, offloading, accelerating, isolating, and providing better security on the DPUs, freeing up valuable compute resources to meet the needs of smart factories worldwide." Travis Vigil, senior vice president of product portfolio and product management for Dell Technologies' Infrastructure Solutions Group, said, "Dell and NVIDIA have already achieved tremendous breakthroughs by working together on next-generation GPU-accelerated data centers. Now, together, we are launching this new solution, combining NVIDIA's powerful BlueField..."

Google Cloud announced that it will integrate Ampere Computing's Altra processors into its services, thereby increasing the proportion of Arm-based processors on its cloud service platform. Last year, Google Cloud added a virtual machine project called Tau to its Compute Engine service, which included the Tau T2D virtual machine, running on AMD's third-generation EYPC server processors based on the x86 architecture. The newly announced Tau T2A virtual machine, powered by Ampere Computing's Altra processors, will further enhance the efficiency of cloud virtual machine operations while reducing power consumption. Google Cloud also announced a deeper collaboration with Ampere Computing, becoming a key partner. The Tau T2A virtual machine can emulate up to 48 CPUs in a single virtual machine, with each CPU configured with 4GB of memory, up to 32Gbps of network bandwidth, and storage capacity that can be expanded according to service needs. This will help users build various network services, containerized microservices, and data storage applications. Besides Google, companies including AWS, Microsoft, Oracle, Alibaba, Tencent, and Baidu have begun to adopt Arm architecture processors to improve server operating efficiency and reduce the previously significant power consumption.

At VMWorld 2019, VMware announced a new service portfolio called Tanzu, allowing users to build services natively based on Kubernetes within the existing vSphere environment. VMware also emphasized its continued embrace of Kubernetes applications and stated that it is currently one of the top three contributors to Kubernetes technology applications globally, seemingly indicating a gradual reduction in the use of virtual machines. In fact, looking at the Tanzu platform architecture, the parallel structure of Kubernetes and virtual machines within vSphere has been adjusted to a system where Kubernetes is the underlying layer, and virtual machines are configured on top of Kubernetes. This architectural shift allows users to directly build Kubernetes application services within vSphere, while also enabling single-point management through Tanzu Mission Control and the advantage of dynamic adjustments. Although other product infrastructures haven't changed yet, this may indicate that if VMware's application model with Tanzu is successful and widely adopted, it may be further applied to other products, potentially reducing the proportion of virtual machine technology used in the original product portfolio. However, according to Mark Lohmeyer, Senior Vice President and General Manager of the Cloud Platforms Group at VMware, both Kubernetes and existing virtual machine technologies are key areas of continued development for VMware, with no particular emphasis on either. Lohmeyer further explained that embracing Kubernetes offers more development opportunities; therefore, the Tanzu product portfolio design is more like an extension of the existing vSphere application model, while also allowing for greater development space for virtual machine technology. Given VMware's current market strategy of developing cross-cloud, multi-cloud, and hybrid cloud solutions, embracing Kubernetes is inevitable. Firstly, the benefits of traditional virtual machines are becoming increasingly limited, and unlike container services, they cannot adapt to changes in demand more quickly, and centralized management is also difficult. Secondly, to connect cross-cloud application services, containerization provides greater compatibility; therefore, embracing Kubernetes, which is currently widely used, will bring more advantages. In particular, the announced collaboration with NVIDIA, leveraging Kubernetes as a medium, enables the vSphere virtualization platform to quickly connect virtualized GPU computing resources. This allows users to rapidly deploy applications requiring large-scale computing, such as artificial intelligence, deep learning, and big data analytics. It also reduces the cost of building servers with integrated GPU acceleration resources and provides flexibility for applications that frequently require adjustments to computing resources based on research project changes. However, even with the increased opportunities brought by embracing Kubernetes, VMware cannot completely abandon its existing virtual machine technology application model. This is because many partners still use virtual machines for computing, and many application services remain built on virtual machine architectures. Therefore, virtual machine technology will remain a key focus for VMware. While the continued expansion of Kubernetes applications may gradually reduce the proportion of traditional virtual machine applications, it will not be completely eliminated. In fact, it may even lead to the development of new usage models for edge computing and IoT applications.