Intel has announced the launch of OneAPI Gold and a new GPU product for the Android Cloud Game Market.
First, the new Intel server GPU This Xe-LP architecture snapshot has four different graphics chips and works together to handle video streaming workloads, video transcoding workloads and Android game streaming solutions. Xe-LP is the same architecture that Intel installs in Xe Max and Tiger Lake. Leopard Chips Performance Reviews Somewhat ahead of AMD’s Ryzen 4000 integrated GPU, and X Max is the same chip in TGL laptops that only runs at high clock speeds.
“We are moving from CPU to XPU,” Raja Coduri told ExtremeTech in a recent call: “Intel SVP and everything else (architecture, graphics and software) Engine SVP” “Intel has built our CPU architecture and played its part in enabling the entire computing world, but we know the workload has improved and we are trying to control more XPU architecture for graphics, media and AI. , Memory, Security and Networking ”
Like Lax laptops, the X-LP chip Intel is launching today for more media workloads than games, but in this case the company is talking about the solution as an Android gaming platform. Clearly, Intel is working with several companies, including Tencent and GameStream, to bring Android streaming services to market. Intel Argument Xe-LP makes it an attractive partner for xeon, enabling customers to become standard on all-Intel solutions for these products.
Intel can support 120 streams in dual GPU configurations and possibly support up to 160 players at a time, depending on the game in question. Since each card has 4 GPUs, Intel is supporting 15 GPU games at 30fps per GPU. Android games aren’t as hardware-intensive as PC titles, but they still run a total of up to 450 frames per 30 GPUs per GPU 15 streams. It is interesting to know how the company distributes inseparable streams between the card and the two GPUs. It also says that the user has some optimism here.
OneAPI gold comes out
Intel’s other major announcement is about the OnAPI initiative. OneAPI is a standardized programming model for standardized code development between various system components, including GPUs, FPSs, and CPUs. Intel always provides its own library and code development tools, and the company uses that expertise to create the most advanced framework for productive development. Support for features like AVX-512 and DLBoost is baked as you expect.
Intel has been talking about OneAPI for many years now, but the project is still only gold, with the expected kit kit expected in December. Free versions are available locally and in the cloud, and Intel will immediately switch to parallel Studio XE and Intel System suite to OneAPI products.
NAPI has produced some news for Intel over the past few years, but it is not clear if Intel’s overall opportunities in the AI market will change. As long as the high-definition X cards are available, Intel is working with one hand tied behind its back. Even with AVX-512 and DLBoost, a modern Xeon server cannot be compared to GPU raw performance.
With OneAPI, Intel is building infrastructure support that will help attract programmers to the platform as GPUs – not to mention consumer CPUs with AVX-512 support – for those who are concerned about CUDA support for mobile and desktop up to 2021 and above. There is background support that allows DPC ++ code to run on CUDA GPUs. DPP ++ is an Intel language implemented in OneAPI, which can accelerate FPGAs, CPUs, and GPUs and any other compatibility it brings to market.
This ad By 2020, Intel seems to be finalizing plans for GPU launch, which means we will have to wait until 2021 to see what the company brings to the table in consumer cards. Tiger Lake is gaining traction in mobile phones, which will show good things about Intel’s big cards in 2021.