The outcry was great when Nvidia expanded the TDP’s influence on performance with the GeForce RTX 3000 laptop GPUs and initially remained silent about it. With mobile processors this is already old hat. Nevertheless, one should continue to talk about it, as the following duel between the Core i9-11980HK and the i7-11800H shows.
The biggest, fastest and best is just good enough for some customers. But does the customer get exactly that if they base their choice on the type designation of the CPU or GPU installed in the notebook? Yes, it does as long as it compares different CPUs or GPUs in the same notebook. If, on the other hand, he compares the supposedly same CPU or GPU in two different notebooks, the type identifier is often no longer of any help.
The TDP influence is enormous in the notebook
The reason: The CPU and GPU manufacturers meanwhile give each model the option of being individually adapted by the OEM – within a seemingly endless range. As a rule, this is done via the TDP, i.e. the maximum permanently available power consumption. Every now and then the cooling system is added as a limiting factor.
An outcry among modern mobile GPUs
This topic came into focus at the beginning of 2021 with the publication of the Nvidia GeForce RTX 3000 laptop GPUs (test), the individual representatives of which could be configured very differently, which was less visible to the customer than ever before. The buyer can hardly see in advance that a mobile GeForce RTX 3060 could definitely beat a 3080 if it was plugged into notebook A and not B. Only the publication by ComputerBase of the TDP configurations allowed by Nvidia, which should actually only be seen by OEMs, brought more light into the dark and set the ball rolling, which now at least provides for the specification of the TDP for every notebook.
Always an issue with mobile CPUs
The problem with mobile CPUs has been around for much longer and therefore less “sensational”. There, too, the official ranking of the CPUs is only worth something as long as they sit in the same notebook with the official standard configuration. Because there has always been a broad TDP framework here. It ensures that supposedly slower processors in Notebook A are much faster than the supposedly better model in Notebook B.
Because in desktop operation, PL2, the maximum permitted power consumption for short load peaks, can also become a more or less enduring endurance runner (PL1) – and at a very different level.
This problem also needs to be examined again from time to time, which the editors did in this test using the golden mean processor Core i7-11800H in the XMG Neo 15 (M21) and the flagship Core i9-11980HK in the MSI GE76 Raider U11 Has.
Intel’s official Tiger Lake ranking
At the start of the test, there is a tabular overview of the Tiger Lake H45 in the notebook, which are most relevant here. This does not claim to be complete, because Intel has massively expanded the Tiger Lake range in notebooks over the past few months and continues to do so on a regular basis. As of the status of the test in mid-September 55 mobile CPUs (including Embedded and Xeon) in the Tiger Lake database deposited.
Important Intel Tiger Lake-H45 for notebooks
Test model 1: MSI GE76 Raider with full equipment
Intel did with the MSI GE76 Raider 11 UH Not just any top model provided, with the Core i9-11980HK and a GeForce RTX 3080 (150 + 15 watts) it is also equipped with the most powerful components on the market. With the gamer in focus, there is only Full HD on 17 inches, but there with a brisk 350 Hz. That has its price: the notebook costs around 4,000 euros.
The new MSI GE76 Raider (Image: MSI)
Image 1 of 3
Even before the first benchmark it was noticed: Neither MSI’s software nor the BIOS allow the HK to be overclocked, even if this is possible with this – and only this – CPU from Intel. The CPU can only be accelerated via the performance profile, although no precise information can be found as to what is hidden behind the respective settings.
The MSI Center for configuration
Image 1 of 3
Test model 2: Schenker XMG Neo 15 (Mid 2021) at full throttle
Schenker shows once again that there is another way with the current one XMG Neo 15 (M21) in the Intel version, which, according to the manufacturer, sells very well. It starts with a Core i7-11800H and an RTX 3060 (115 + 15 watts), which can be exchanged for an RTX 3080 (150 + 15 watts) as in the test sample. Then the price is included just under 2,550 euros. In this case there are 15.6 inches with 2,560 × 1440 pixels at 165 Hz.
XMG Neo 15 (M21) (Image: Schenker)
Image 1 of 2
In this niche, Schenker attaches great importance to freedom for the experienced user, but also to exploring the specified maximum. No other notebook vendor is currently so open to the outside world when it comes to TDP, power limits and more for processors and graphics cards – also in the ComputerBase forum. Where many companies are silent and sometimes waste performance through bizarre constellations, Schenker at XMG proceeds exactly differently if the technical implementation of the ODM allows it. And the XMG Neo 15 (M21) allows everything.
The control center of the XMG Neo 15 (M21)
Image 1 of 4
Not only can the GPU in the XMG Neo 15, as in the XMG Core 17 with Ryzen (test), be set completely freely at the beginning of the year, the CPU can also endure it – be it via regular performance profiles or manually.
For example, the Core i7-11800H can be set to power limits of 120 watts for short and long-term loads and the fan can be turned up to match. This limit is set two restarts later.
The test objective: maximum performance for the maximum possible time
The aim of the test is to get the maximum possible CPU performance out of the notebook. A look at the paper suggests that the result is clear.
The Intel Core i9-11980HK in the MSI model has a clear lead with turbo clock rates of 4.5 GHz for load on all cores up to 5 GHz for load on two or one core. The Core i7-11800H from the Schenker Neo 15 has to be satisfied with a slightly lower 4.2 GHz on all cores, and it can only clock up to 4.6 GHz for two cores. Here the requirement clearly comes from above, i.e. Intel: Only flagships are allowed to go all the way up in single-core clocks, although almost every other CPU in this portfolio could probably do the same.
In practice, MSI allows the CPU to operate at 95 watts with a configTDP ex works of 65 watts with the maximum possible performance profile and the highest fan level. This is not new per se; previous generations of Intel were also often on the move in these areas. Intel itself even allows 107 watts for Tiger Lake according to the specification. Schenker goes one step further. Here the setting of the power limit can be clearly defined for PL1 and PL2. These can each be set to 120 watts – and without any time limit.
Intel Core i7-11800H in the Schenker Neo 15 at 120 watts
The interesting thing about the Schenker XMG Neo 15 (M21) is that this game can also be driven in the other direction if desired. The built-in Core i7-11800H is also officially available with a configuration of only 35 watts, the so-called configTDP down. A manufacturer can therefore decide to build a very thin notebook with a small cooling system, but then nail the CPU directly to 35 watts – for now and forever. That would be the worst-case scenario for processor performance, but it is by no means excluded these days, as test models have repeatedly shown in recent years. In this respect, exactly this extreme is also taken into account in the test.
On the next page: benchmarks and experiences
Post a Comment
Post a Comment