The wait for the Intel Core Ultra 200S has extended for monthsthroughout which some very interesting leaks have been produced, which told us what Intel had planned for this new generation of integrated devices for desktop systems. Last month, for example, we already provided you with a lot of information regarding Arrow Lake-S, which has been its unofficial name in this period, and today we reported on a new leak about specific models.
Now, whether they were more or less reliable, until now we only had leaks, but Intel has finally presented the Core Ultra 200S platformconfirming some of them and substantially expanding the information we have, which finally allows us to review, safely and in depth, the improvements and changes that come from the hand of this new generation of integrated, with the that Intel will try to begin to leave behind the bad period it has recently gone through.
Intel has launched the Core Ultra 200S series in a time of significant change in the desktop processor industrymarked by growing demand for energy efficiency and advanced processing capabilities. With this new generation, Intel seeks to regain ground against its competitors and address the expectations of the most demanding users, introducing key improvements in performance and efficiency.
Among the most notable new features, the Core Ultra 200S include an integrated neural processing unit (NPU)designed to improve artificial intelligence applications. Another significant improvement is the integration of the new Xe-LPG GPUwhich doubles the graphical performance compared to previous generations. The processors are also optimized for lower power consumption, reducing power usage and operating temperatures, making them more efficient for high-performance systems.
Cores
The Arrow Lake architecture in the Intel Core Ultra 200S processors maintains the hybrid approach that we already saw debut in the 12th generation, with the combination of Lion Cove (P-cores) and Skymont (E-cores) cores. The orchestrator, as you already know, is responsible for assigning highly demanding tasks to the P-cores, while the E-cores handle less demanding parallel processes, optimizing the overall efficiency of the system. The key is in the refinement of the P-cores, which have been restructured to increase their performance per clock cycle (IPC), improving both execution in single-threaded and intensive multithreaded tasks.
In this generation, Intel has introduced improvements at the microarchitecture level in the P-coressuch as optimizing the memory subsystem and using a broader out-of-order execution engine. Additionally, the increase in L2 cache per core (up to 3 MB) and shared L3 cache (up to 36 MB) provides greater storage capacity for data close to the cores, resulting in reduced latencies and better performance. handling tasks with high memory demands. These improvements have a direct impact on the cores’ ability to execute more instructions per cycle, increasing efficiency per watt in high-demand scenarios.
Another of the most relevant changes is Removing Hyper-Threading in the Lion Cove cores. This decision responds to the need to improve energy efficiency and maximize the use of physical cores, which reduces the overhead (additional management burden to distribute tasks among threads and coordinate access to shared resources) associated with thread duplication in workloads that do not significantly benefit from concurrent multithreading. With this approach, Intel seeks to gain greater control over power consumption in intensive loads, while improving the raw processing capacity of individual cores.
NPU
The NPU (Neural Processing Unit) in the Intel Core Ultra 200S is one of the most innovative features of this generation. Specifically designed, as you already know, to accelerate artificial intelligence tasks, since after the emergence of AI-based services, the demand for devices to be capable of carrying out at least part of these tasks has begun to grow. without depending on the cloud, an approach in which we find two models: the one that relies on NPUs versus the one that relies on GPUs. In this case, Intel opts for the first option.
The NPU integrated into these new chips is capable of providing up to 13 TOPS in processing associated with artificial intelligence and works in conjunction with the processor’s memory subsystem, optimizing access to large volumes of data through prediction techniques and specialized caching to minimize latency. This integrated system also allows for high parallelization in workflows, thus maximizing performance in tasks such as real-time inference and the acceleration of deep neural networks.
Now, for those users for whom the 13 TOPs of the NPU fall short, Intel also focuses on the cumulative computing capacity of NPU+GPU+CPU and that, as you can see in the image above, It is 36 TOPs (13+8+15). Thus, all applications that are capable of making combined use of these three integrated AI tasks will find a fairly fast and reliable response from those workloads that have been designed and/or optimized to run locally.
GPU
In terms of graphics and multimedia capabilities, The GPU of the Intel Core Ultra 200S processors are based on the Xe-LPG architectureoffering significantly improved graphics performance compared to previous generations. This integrated GPU is designed to handle demanding graphics tasks, such as high-resolution gaming, image rendering and video editing, doubling the graphical computing capacity compared to its predecessors. Support for advanced technologies, such as DirectX 12 Ultimate and ray tracing, also improves visual quality and performance in applications that require high graphics processing power.
One of the key innovations is the optimization of energy consumption in graphic tasks. Intel has implemented power management improvements within the Xe-LPG GPU, enabling a more efficient balance between performance and power consumption. This is especially relevant for users who require sustained graphics power during long work or gaming sessions, as it reduces both consumption and operating temperatures, prolonging system durability and improving stability.
The processors are also prepared to support the latest multimedia standards. The architecture supports playback of content in 4K resolutions at 60 Hzas well as the possibility of managing multiple high-resolution screens simultaneously. In terms of video encoding and decoding, the Core Ultra 200S series includes support for the latest codecs such as AV1 and HEVC, which optimizes the transmission and editing of multimedia content, facilitating a smooth experience in streaming and content creation.
Foveros 3D Packaging
Intel Core Ultra 200S processors are manufactured using Intel’s Foveros 3D packaging technologywhich we already saw last year in the Meteor Lake for portable systems. a technique, an evolution of the one presented by the company back in 2018, that allows different components to be stacked vertically. This approach allows logic, graphics, memory chips and other elements to be combined into a single compact package, improving both performance and power efficiency. By employing this type of packaging, Intel can fit more elements into a smaller area, resulting in more powerful processors without significantly increasing power consumption or heat generated.
This process also allows for greater flexibility in chip design, by allow different components to be manufactured using different production nodes and are integrated into a single package. This is especially relevant for the Intel Core Ultra 200S, since different parts of the processor (such as the high-performance and efficient cores, the integrated GPU and the NPU) can be optimized independently, which improves overall performance without increasing production costs significantly. Additionally, the optimized use of Foveros 3D allows for better heat dissipation, helping to maintain lower operating temperatures, even under intensive workloads.
Performance
The Intel Core Ultra 200S feature significant performance improvements compared to their predecessors. According to data provided by Intel, the new Lion Cove and Skymont cores provide a 32% increase in performancesingle-threaded operation and up to 72% in floating-point operations, as a result of improvements in the microarchitecture, as well as a larger L2 and L3 cache size, and optimizations in out-of-order execution (OoOE), which which allows more demanding tasks to be handled more efficiently and quickly, especially in scientific and processing-intensive applications.
In terms of energy efficiency, these chips They are capable of reducing energy consumption by up to 45%making them significantly more efficient under long workloads and in multitasking applications. Added to this is the new Xe-LPG GPU, which doubles the graphics performance of the previous generation, allowing faster processing of real-time graphics and high-demand games with lower power consumption. These advances are complemented by the integration of the NPU, which offloads artificial intelligence tasks, optimizing both the processing speed and the overall energy efficiency of the system.
Last year, with the presentation of Meteor Lake, Intel made an important difference between its integrated laptops and those aimed at desktop systems, a difference that now, with the presentation of Intel Core Ultra 200S, almost completely blurred. Will it be enough to give Intel a boost in low times? It is not at all clear if this will finally be the case, but in the absence of being able to test them, it seems that Intel has done a good job.
Source: www.muycomputer.com
