Intel's next-generation client CPU "Nova Lake" may not adopt a large-capacity cache in its mobile version. On August 18, 2026, Jaykihn, who shares hardware information, posted on X regarding the unannounced "Razor Lake," stating that "mobile SKUs with bLLC will appear." Previous information originating from the same source had suggested that Nova Lake's bLLC would be limited to desktop variants. If this latest post is also accurate, Intel's large-capacity mobile cache would not debut with Nova Lake, but rather begin in earnest with the subsequent Razor Lake.
That said, Intel has not announced Razor Lake, nor has it confirmed whether Nova Lake Mobile will feature bLLC. The post only mentions "mobile SKUs" in the plural. Specifications including SKU count and cache capacity, as well as launch timing, remain unknown. Even so, in the current high-end laptop market, AMD has already placed the Ryzen 9 9955HX3D with 128MB L3 cache. While cache figures alone cannot be treated as an equivalent performance metric, which generation Intel chooses to counter with matters for anyone selecting a gaming laptop.
Is Nova Lake Mobile Deferring Large-Capacity Cache?
This latest piece of information amounts to a brief statement: a bLLC mobile SKU is coming with Razor Lake. Neither Intel nor the poster has explained what bLLC stands for, where it sits relative to a conventional last-level cache, or whether it involves 3D stacking. Therefore, it cannot yet be written that Intel has adopted bLLC as an official product designation, or that it was designed specifically to counter AMD's X3D.
However, combined with the previous assumption that "Nova Lake bLLC is desktop-only," a certain reading becomes possible: the mobile version of the Nova Lake generation would not carry a large-capacity cache, with mobile SKUs instead arriving with Razor Lake. This is not a confirmed roadmap but unverified information relying on the same source. Until Intel releases an official SKU chart, it cannot be definitively stated that "Nova Lake Mobile has no bLLC" either.
The timeline Intel has officially disclosed pertains to the Nova Lake product lineup as a whole. During its Q4 2025 earnings call in January 2026, the company described Nova Lake as its next-generation client product launching at the end of 2026. This statement refers to the launch timing for the client product lineup and does not guarantee the mobile version's retail release date, branding such as HX, or cache capacity. Care must be taken not to conflate discussion of an unannounced codename with officially confirmed timing.
36MB vs. 128MB: Laptops Already Show a Cache Gap
Comparing current top-tier mobile CPUs by official specifications, the Intel Core Ultra 9 285HX offers 24 cores, 36MB Intel Smart Cache, a Processor Base Power of 55W, and Maximum Turbo Power of 160W. Meanwhile, the AMD Ryzen 9 9955HX3D features Zen 5 architecture with 16 cores/32 threads, 128MB L3 cache, a Default TDP of 55W, and cTDP of 55–75W, and AMD markets it as a second-generation AMD 3D V-Cache product.
| Product | Cores/Threads | Rated Last-Level Cache | Rated Power Specifications |
|---|---|---|---|
| Intel Core Ultra 9 285HX | 24 cores | 36MB Intel Smart Cache | Processor Base Power 55W, Maximum Turbo Power 160W |
| AMD Ryzen 9 9955HX3D | 16 cores/32 threads | 128MB L3 | Default TDP 55W, cTDP 55–75W |
This 36MB versus 128MB comparison is valid as a comparison of rated last-level cache capacity for top-tier mobile CPUs. However, Intel Smart Cache and AMD 3D V-Cache differ not only in implementation details such as configuration and shared scope, but also in latency and power behavior. Furthermore, the 55W figure refers to Processor Base Power for Intel and Default TDP for AMD, so they cannot be placed side by side as identical power measurements. It would be premature to draw conclusions about gaming performance superiority from capacity differences alone.
Even so, what AMD's product demonstrates is that large-capacity mobile cache has already become a commercially available option. If Nova Lake Mobile does not incorporate bLLC, Intel would need to compete through a combination of core performance, conventional LLC, memory, and power management. This does not imply that Intel's products are inferior—rather, it means the axis of comparison differs from AMD's products, which foreground cache capacity.
Mobile bLLC Cannot Be Evaluated by Capacity Alone
A large last-level cache can reduce DRAM accesses in scenarios where the working set during processing fits within it. In workloads like gaming, where data-wait time affects performance, being able to place data closer to the cores can influence average frame rates and frame times. However, the same effect does not manifest across all applications; results vary depending on memory access characteristics and how the game engine itself is built.
Nor is it simply a matter of adding more. Increasing cache capacity consumes silicon area and places additional burden on leakage power, design, and yield. In laptops, where cooling capacity and power budgets vary by chassis, how this burden is allocated toward performance shapes a product's character even more than in desktops. These are general design trade-offs, not facts that Intel has explained as reasons for Nova Lake Mobile to forgo such a feature.
Even if bLLC does arrive with Razor Lake, the material needed for evaluation is not yet in place. How far will it extend—across HX, AX, or another mobile lineup? The original post suggests multiple SKUs but does not indicate a specific number. Without details on cache capacity, visibility from all cores, latency, sustained performance, and supported memory configurations, it remains impossible to measure how it differs from AMD's 3D V-Cache products.
Correction from N2X to N2P V2
Regarding Razor Lake, a correction has also been made concerning the manufacturing process name. Tom's Hardware has added a note that Jaykihn corrected the initial information from N2X to "N2P V2." This correction record also serves as a caution when using process names as grounds for performance or launch timing claims.
According to TSMC's public description of its advanced HPC-oriented technology, N2P uses the same design rules as N2 while improving performance by 5%. N2X is a separate category that, by selectively using ultra-high-performance standard cells and high-speed devices, claims a 5% improvement from each, for a combined 10% speed increase. The publicly available page contains no name or definition for "N2P V2."
Therefore, the corrected designation cannot be equated with an officially published node name from TSMC. To begin with, Intel has not disclosed either the existence of Razor Lake or the manufacturing process it will adopt. One should avoid both retaining the initial N2X report as a current claim and preemptively assessing Razor Lake's performance based on the undefined name "N2P V2."
Nova Lake's Year-End Launch Is Visible; Judgment Should Come from Spec Sheets and Real-World Testing
What Intel has officially committed to is only the timeline for the Nova Lake product lineup arriving by the end of 2026. The mobile version's cache strategy remains unannounced, and information regarding Razor Lake remains limited to social media posts. This story is best read not as confirmed product specifications, but as a hypothesis concerning when Intel's large-capacity cache will reach laptops.
What directly informs purchasing decisions will be spec sheets and real-world measurements for individual products, not future codenames. Once Intel discloses Nova Lake Mobile's SKUs, cache configuration, and power envelope, it will finally become possible to compare, against the AMD Ryzen 9 9955HX3D's 128MB L3, what Intel's design prioritizes at which price point. Even if Razor Lake's bLLC proves true, its value in laptops will be determined not by capacity figures, but by sustained performance across different chassis and real-world gaming benchmarks.