HardwareFreaks

Tag: Nova Lake-S

  • 28-core Intel Xeon CPUs for Dunlow workstation LGA1954

    Featured image 28core Intel Xeon CPUs for Dunlow workstation LGA1954

    Intel is signaling a significant shift in its processor strategy with the upcoming Dunlow platform, a project codenamed to redefine entry-level server and workstation applications. This new architecture promises to introduce processors based on the Xeon 6300P-series, setting up a competitive landscape that bridges the gap between high-end desktop performance and enterprise requirements.

    The Dunlow platform is engineered around a specific set of specifications designed for efficiency and capacity. It will feature up to 28 cores and support a dual-channel memory subsystem, housed in an LGA1954 form factor with a modest processor base power of 95W. This design focuses squarely on delivering robust performance without the extreme demands seen in enthusiast desktop systems.

    Data by NBD, screenshot by Tom’s Hardware

    This move creates an intriguing contrast when placed next to Intel’s consumer-focused desktop line, codenamed Nova Lake-S. The Nova Lake-S CPUs are reportedly aimed at enthusiasts and promise a much more aggressive core count, potentially featuring up to 52 cores, including high-performance Coyote Cove cores alongside energy-efficient Arctic Wolf cores within the compute tile.

    The distinction in design philosophy is stark. While desktop chips chase raw power, aiming for peak performance in demanding applications and pushing TDPs up to 474W, the Dunlow platform opts for a more deliberate core count of 28 P-cores with a lower 95W base power. This difference suggests Intel is tailoring silicon specifically for the server and workstation segment rather than simply replicating desktop prowess.

    Experts are analyzing why this specific configuration was chosen. Building on prior design trends, incorporating 28 cores does not align naturally with standard desktop or even notebook-grade designs, which typically feature a mix of P-cores and E-cores. Furthermore, Intel generally avoids fusing nearly half a desktop die into server products, suggesting the Dunlow architecture is an entirely unique approach.

    This tailored design may stem from a necessity to bridge a widening performance gap. As high-end Xeon processors with complex memory subsystems become increasingly expensive, and as options like the ‘Diamond Rapids‘ octa-channel solutions are canceled, there is a critical need for a viable alternative in 2028.

    The Dunlow platform, with its 28 P-cores and dual-channel memory, offers a potential sweet spot. It provides substantial core count necessary for demanding server tasks while avoiding the complexity and cost associated with over-specifying memory bandwidth that many standard applications do not require. It represents Intel’s strategic attempt to fill the chasm between enthusiast desktop hardware and high-end enterprise solutions.

    Credit: Tom’s Hardware

  • Intel Nova Lake-S midrange CPUs could be bringing AMD’s X3D cache trick to more affordable chips

    Intel’s New Midrange CPUs Master the Art of Balance

    The latest wave of mid-range processors from Intel is signaling a shift in design philosophy, moving away from simply maximizing raw performance and toward a smarter balance of efficiency and throughput. This architectural pivot, seen in the new Nova Lake-S lineup, promises a more effective way to handle the diverse demands placed on modern computing systems.

    At the heart of this evolution is a sophisticated core configuration. Each processor in this new generation intelligently combines various types of cores designed for different workloads. Specifically, these processors are equipped with 6 Coyote Cove P-cores, 12 Arctic Wolf E-cores, and 4 LP-E cores.

    What makes this combination truly interesting is not just the raw numbers, but the deliberate strategy behind them. This configuration demonstrates that Intel is prioritizing balanced compute throughput—the ability to handle heavy processing tasks efficiently—rather than simply stacking more performance cores.

    By incorporating a robust number of efficiency-focused E-cores alongside the high-performance P-cores, the new architecture excels at managing complex, real-world scenarios. This design is highly adept at juggling intensive foreground applications while smoothly handling necessary background processes and low-power tasks without compromising overall system responsiveness.

    This balanced approach means that users can expect greater efficiency and better performance consistency across a wider range of operations. It’s a testament to modern processor design: achieving peak speed by intelligently allocating resources, ensuring that every component is working in harmony for the benefit of the user experience.

    Credit: TechSpot

  • Intel reportedly preparing high-cache CPUs to combat Ryzen X3D

    Intel is gearing up for a major push in the desktop processor market, reportedly preparing two high-performance mid-range chips designed specifically to dominate gaming environments. These upcoming processors, labeled the “Nova Lake-S”, promise a significant performance boost through the integration of a massive 108MB of specialized cache.

    These new models are positioned within Intel’s Core Ultra 5 400S processor family and are built around a single compute die design, optimizing them specifically for mid-range gaming systems. The architectural strategy aims to tackle the latency challenges inherent in high-performance gaming by combining updated core microarchitectures with Intel’s innovative new Last Level Cache layer.

    Under the hood, the 22-core architecture is a carefully balanced blend of efficiency and speed. It features 6 Coyote Cove”>“Coyote Cove” performance cores, supported by 12 Arctic Wolf”>“Arctic Wolf” efficiency cores, alongside 4 low-power efficiency (LPE) cores. This configuration is designed to mimic the performance experience offered by AMD’s Ryzen X3D chips.

    The goal of this design is clear: mitigate latency-sensitive workloads like gaming, effectively aiming to reclaim the crown in performance against competitors.

    Not all configurations are created equal. The two rumored models boast distinct power profiles and multiplier support. One variant is set up as an unlocked “K” series processor, offering a powerful 125W TDP for enthusiasts who crave overclocking potential. The second model operates with a more controlled, locked 65W base TDP, catering to efficiency-focused users.

    The supporting desktop platform for these CPUs, the Z9x0, also demonstrates serious scalability. This new LGA-1954 socket is capable of accommodating up to 52 cores, suggesting room for future expansion in high-core count systems.

    For those looking at extreme performance potential, rumors suggest that high-end dual-tile configurations on this platform could push power consumption up to 474W in PL2 mode, underscoring the potential horsepower packed into these next-generation Intel processors.

    Credit: KitGuru

  • Intel reportedly adding two new 22-core SKUs with game-boosting cache to Nova Lake-S lineup — 125W unlocked and 65W locked part rumored to be part of single-tile Core Ultra 5 tier

    Featured image Intel reportedly adding two new 22core SKUs with gameboosting cache to Nova LakeS lineup 125W unlocked and 65W locked part rumored to

    Intel is gearing up for what promises to be a massive performance leap with its next-generation desktop family, Nova Lake. But the real excitement isn’t just in the architecture; it’s in the rumored introduction of bLLC—Big Last Level Cache—a potential game-changer that could redefine how CPUs handle data and responsiveness.

    This innovation positions Intel to tackle the performance challenges set by competitors, particularly AMD’s X3D chips. It’s essentially Intel’s highly aggressive answer to high-speed cache technology, aiming to deliver significant gains right on the desktop.

    The latest leaks suggest this caching strategy is being integrated across the lineup. A recent disclosure indicated that two new Core Ultra 5 SKUs are slated to feature bLLC, bringing an impressive total of 22 cores into the mix for these chips. This isn’t just a marginal update; it signals a fundamental shift in how Intel structures its high-end processors.

    Breaking down the structure, each of these new chips is rumored to utilize a configuration of 6 P-Cores (Performance cores), 12 E-Cores (Efficiency cores), and 4 LP-E cores (Low Power Efficiency cores) on a single tile. This arrangement allows the CPU to access up to 144MB of bLLC, offering a substantial boost in immediate data access.

    The physical implementation also shows promise for versatility. Dual-tile variants of Nova Lake-S could even push the cache count up to a staggering 288MB, though these high-end options are likely reserved for the absolute pinnacle of the product line.

    Under the hood, this new generation will leverage established yet highly advanced architectural components. We anticipate the P-cores will utilize the Coyote Cove architecture, while the E-cores will be built upon Intel’s Arctic Wolf design, ensuring both efficiency and blazing speed.

    Beyond cache and cores, the thermal management remains critical. The leaked details suggest a mix of power consumption profiles, with one SKU potentially operating at a 125W TDP (Thermal Design Power) for unlocked variants, while another locked version targets a more restrained 65W TDP. This flexibility caters to different use cases.

    This expansive Nova Lake family is set to offer several compelling options, some of which might not even make it to the market before its launch next year as part of the Core Ultra 400 series. While the exact naming conventions are still pending, the roadmap points toward a diverse range of chips.

    As we look ahead, rumors point toward an imminent CES 2027 announcement for Intel’s next-generation family. Despite ongoing component uncertainties in the industry, the trajectory for Nova Lake is clearly set to deliver a dynamic and cache-rich experience for PC builders and gamers alike.

    SKU

    Core Config (P+E+LP-E)

    bLLC

    TDP (Unlocked/Locked)

    52 Cores (dual-tile)

    (8+16)+(8+16)+4

    288MB

    175W

    44 Cores (dual-tile)

    (8+12)+(8+12)+4

    264MB

    175W

    28 Cores

    8+16+4

    144MB

    125W

    28 Cores

    8+16+4

    		–

    N/A

    125W / 65W

    22 Cores

    6+12+4

    108MB

    125W / 65W

    Credit: Tom’s Hardware

  • Intel’s next-gen 52-core Nova Lake CPU could pull up to 474W — high-end LGA1954 motherboards may need three 8-pin power connectors to feed the monster

    Intel is gearing up for a massive hardware overhaul with its upcoming Nova Lake series processors, promising to push the boundaries of power draw and multi-core performance. Based on recent leaks, this new generation aims squarely at dethroning the current leaders in the CPU market.

    The flagship desktop variant is anticipated to be a beast, boasting up to 52 cores and utilizing a sophisticated dual-compute tile architecture. To handle this immense processing power, the design features a demanding maximum short boost power limit (PL2) target of 474W. For enthusiasts pushing the limits, rumors suggest that top-end models might even accommodate an emergency PL4 power draw exceeding 700W.

    Beyond raw compute muscle, Intel is redefining the platform itself. The new lineup is tied to a future platform known as Nova Lake-S, which will introduce the highly anticipated LGA1954 socket. This transition signals that motherboard vendors will need to adapt, classifying boards by sustained power limits—ranging from 35W up to 175W—to accommodate various CPU configurations.

    For the hardcore overclockers who thrive on extreme settings, there are whispers of enhanced connectivity. Some enthusiast-grade motherboards are rumored to feature three EPS 8-pin CPU power connectors, providing the necessary infrastructure for pushing CPUs into new realms of performance.

    Under the hood, the Nova Lake-S chips are expected to arrive with specifications that put them in direct competition with AMD’s leading innovations. The flagship model is projected to feature 16 performance cores and 32 efficiency cores, alongside a novel Big Last Level Cache (bLLC) design aimed at capturing the dominance currently held by technologies like 3D V-Cache.

    Memory support will also be a major focus. The new processors are slated to support blazing fast DDR5-8000 RAM, ensuring that these powerful chips can communicate with high-speed memory without bottlenecking performance. Furthermore, the platform integrates next-generation connectivity features, including Thunderbolt 5 and PCIe 5.0 interfaces.

    The future of Intel’s architecture also incorporates specialized components, including integrated Xe3 graphics, an upgraded Neural Processing Unit (NPU) optimized for AI workloads, and significant leaps in communication speed.

    In short, Intel’s next generation is not just an iterative upgrade; it’s a bold step toward creating a new standard for desktop computing. While these exciting specifications are still circulating as leaks, the trajectory suggests substantial gains in gaming performance, multi-threaded throughput, and overall platform capability.

    Credit: Tom’s Hardware