Intel had no intention of being beaten by AMD. Barely a few weeks after the launch of the new generation of SoC Ryzen 4000 Mobile by its competitor, Intel is launching its H series of mobile processors of 10e generation.
Known by the code name of Comet Lake H, the H is there to highlight the high performance, these first six chips are now the flagship of Intel's mobile offer. SoCs that dissipate 45W and that embed up to 8 physical cores and 16 logical cores. A response under the sign of pure performance with two very precise targets in mind: gamers and creators.
If you follow the processors a bit, you probably know that Intel is now behind AMD in terms of pure engraving finesse, the latter using the 7nm process of the Taiwanese TSMC. Intel for its part only industrializes 10nm on certain mobile chips such as the Ice Lake ultraportables. But curiously, not on these new Comet Lake H which are engraved in 14 nm ++.
Why the hell not to use 10 nm? If there are undoubtedly reasons for production – all factories at Intel may not be ready – Intel has communicated that their “Mastering the 14nm ++ process is the only one that allows us to reach such high frequencies”.
However, high frequencies are the weighty argument of this new generation of chips. More precisely, the high frequencies in single core: basic, the multicore frequencies of the different models range from 2.4 GHz to 2.7 GHz. But for little or no multi-threaded applications, the counter explodes. The highest-end model pushes up to 5.3 GHz – and it's an “unlocked” model that can potentially go even higher.
Even the smallest Core i5 in the family displays a single-core 4.5 GHz boost. Our regret in Intel's communication being that it does not speak of the multi-core boost, very important for creation (we just know that the Core i9 grows at 4.4 GHz on all hearts).
According to Intel, few video games take advantage of more than one or two cores. A statement that sticks for the moment with the performance measures that always give (or almost) the advantage to chips of Intel, even less provided in cores than those of AMD. Thanks, you will understand, at their highest frequency.
If we have not yet seen or performed performance measurements, it seems likely that the 5.3 GHz Turbo of the Core i9 10980HK dominates the 4.4 GHz of AMD’s flagship, the Ryzen 9 4900H – 900 MHz from difference in single-thread, it starts to count!
The heart of the wheel
To reach these frequencies, this new batch of chips incorporates Turbo Max 3.0 technology. A device invisible to the user which consists in identifying the “preferred CPU cores”, or, more precisely, the fastest cores. Because even if their production is the result of a cutting-edge industry, all CPU cores are not equal and do not respond as well to the increase in frequency.
By coupling information from low-level drivers in Windows and a memory area of the CPU, the chip knows which are the two fastest cores of the group and relies on them almost instantly when it is necessary to increase rapidly in frequency.
To add some performance, Intel has integrated a functional overclocking system on several models, the Easy Optimizer. Same perfs clustering on the memory side that goes from DDR-4 2666 to 9e generation to DDR-4 2933. PCIe 4.0 and DDR 5 will be for the next time.
For creators, whether they are video editors, graphic designers, 3D modelers, streamers, etc. the Comet Lake H generation obviously offers a horde of hearts to face AMD. Above all, it benefits from massive software support from Intel, which “lends” many engineers and trains numerous software development studios to ensure excellent support for its chips. A software asset often forgotten, and yet capital when the applications in question are your livelihood.
This Core H family of 10e generation isn’t there to revolutionize the market like AMD’s Ryzen 4000s. Logic: AMD had everything to prove and had to make it happen in an attempt to break into Intel’s virtual monopoly on high performance mobile chips.
Intel is therefore content here to ensure its leadership in terms of pure power and guarantees the same number of cores as AMD so as not to leave it a field in the field of “marketing figures”. But the fact remains that the fine engraving of these new chips leaves the risk of greater energy consumption at rest and that the integrated graphic part – an old generation UHD chip – looks a little pale in front of the camera. offer Ryzen 4000 equipped with small Radeon Vega – even if the target remains machines with an additional graphics card.
We have to wait for performance measurements to decide, but for target markets (creation and gaming) where users are still on the sector, it's less the battery life that counts than the number of frames per second or software rendering time. Points where Intel has always shone. But even if Intel keeps the performance advantage on this generation and even if it would align its prices on AMD, the giant must pass the second quickly because the generational jumps in performance of its competitor are of an entirely different magnitude.