Chip maker Intel announced to launch their GPU’s for servers running in data centers. So, Intel rivals NVIDIA and AMD with its coming up with the GPU’s for the data centers. This might further fuel the competition in the GPU enabled servers for data centers. Intel is the first manufacturer of its first discrete GPU for data centers and it is designed for the Android-based
cloud gaming and media streaming segments.
The newly introduced GPU is for Data Centers and after the launch its GPUs Intel will give tough competition to already established GPU makers AMD and Nvidia.
Intel’s GPU for servers is a low-power chip which will be used in PCIe graphics cards and it can be installed on the servers. Intel also posted a message on its official Twitter account regarding its upcoming GPU’s for the server.
As of now the GPU and graphics card market is dominated by already established AMD & Nvidia. It is noted that Intel has already launched Iris Xe Max, a discrete GPU for thin laptops. The Iris Xe Max is designed to work well with the Intel CPU and give high performance. The power consumption is also optimized for Iris Xe Max GPUs. So, Intel is already started giving competition to AMD and Nvidia in the laptop GPU market.
According to the official statement, Intel would release oneAPI toolkit for the developers which will include vector-matrix-spatial architectures. The new API will help developers in developing software that will utilize Intel’s GPUs in a much better way. This way the company is going to offer hardware and software stack to the developers for developing next generation compute intensive AI applications.
Intel is bringing Server GPU for the data centers and this move will expand existing offerings to the next level in Server computing. The company is using its Xe-LP microarchitecture, which is said to be the most energy-efficient graphics architecture for intensive graphics processing. Intel is aiming to provide high-density, low-cost solutions for Android cloud gaming and video streaming.
Intel’s solution will further enable the service providers to scale graphics processing infracture separately from the servers which can support large numbers of streams and subscribers per system. All these solutions will help the solution providers to offer a wide range of services to the clients.
Intel is already partnering with the various service providers, including Gamestream, Tencent Games, and Ubitus for the release of their data center server solutions.
It is noted that Intel also releases a separate enterprise grade graphics card XG310, which is manufactured for the purpose of cloud streaming. This card is designed to accommodate 4 Server GPUs and offers a base clock speed of 0.9GHz. This can be run at a boosted clock speed of 1.1GHz. The XG310 card is energy efficient and can run on low power costs as compared to counterpart Nvidia RTX 3080 card. Clients can use Intel’s XG310 card to consolidate their server in small form factors and get much more computing power with extra power savings.
The Intel oneAPI initiative, which was first introduced at the SuperComputing 2019 event will further help developers in developing programs to take full advantage of Intel’s GPU Servers. The oneAPI is an open initiative and developers can take full advantage of the hardware without vendor lock-in.
