Nvidia Max-Q & WhisperMode: Information & Initial Benchmarks

The big topic here is to change old priorities, which is certainly not what all gamers want. Max-Q will have a hard time with users looking for the maximum performance, which is certainly the case for many people planning to buy a gaming laptop. Nvidia sacrifices a part of the potential performance to ensure the previously mentioned quiet and cool operation in slim and compact cases. It is therefore not about the maximum performance, but a reasonable performance in consideration of power consumption, emissions, and thermals.

We have first data for a benchmark comparison available from our review of the HP Omen 15-ce002ng with Nvidia GeForce GTX 1060 Max-Q. We compared it in three notebooks with standard GTX 1060 (laptop): The EVGA SC15 has a particularly high 3DMark 11 graphics score, while the Asus Strix GL502VM-DS74 represents the average and the Razer Blade (2017) trails the pack with 10% lower performance than the EVGA.

Our HP Omen cannot even keep up with the relatively slow Razer Blade and is 4% and 5% slower in 3DMark 11 (GPU) and the Fire Strike test of the newer 3DMark (2013), respectively. It even performs 14 and 10%, respectively, worse than the especially fast EVGA. We used the  Schenker XMG P507 (GTX 1060) for comparison of the Time Spy and Ice Storm (3DMark), Luxmark, and Cinebench Open GL benchmarks, since we do not have results from the devices mentioned above.

It is suspicious that the HP Omen with Max-Q performs better than the Schenker P507 in the (less and less important) Ice Storm benchmark. In all other benchmarks that we have run, the Omen comes out last. We cannot explain this anomaly. From the overview below, you might get the impression that the Max Q GPU's performance loss in comparison to standard 1060 graphics cards is higher under high graphics load than under moderate load (Cinebench for example).

In the gaming comparison, the GTX 1060 Max-Q in the Omen 15 also almost always trails the pack. It only got 2nd of three in the terribly PC ported “Batman Arkham Knight” (High Preset) and the “Witcher 3” (High). In the listed titles, the Max-Q variant performs worse than the fastest GTX 1060 by 8% to quite high 24 % in the very demanding “Rise of the Tomb Raider”.

In the Omen 15's defense, we want to point out that none of the games listed below that could be smoothly played with a standard GTX 1060 became unplayable in maximum 1080p setting. However, the FPS reserves, which might be important for features such as V-Sync, got smaller. Those who want to enjoy games such as “Tomb Raider” and “The Witcher 3” in Full HD and maximum finery will appreciate every additional FPS. The GTX 1060 is not suitable for gaming in 4K resolution anyway, but 10% higher performance can make the difference here: sink or swim?

In the meanwhile, we had the chance to benchmark the MSI GS63VR 7RG with Nvidia GeForce GTX 1070 Max-Q. Unfortunately, we did not have a test model with the GTX 1070 Max-Q in time for this article, so we have to rely on unofficial scores from two manufacturers that do not want to be mentioned. We do no have pure GPU scores either, so we have to use 3DMark11 overall scores instead. We took two laptops with the standard version of the GTX 1070 and the i7-7700HQ from our database to evaluate the results. One of the laptops is at the top, and one at the bottom of the small performance range.

The performance loss by Max-Q is not that big since the regular GTX 1070 GPUs are only 7 to 12% faster. Contrary to initial expectations, the Max-Q-1070 in the MSI GS63VR performs significantly worse than its conventional counterparts in the Asus and Acer. It is 16% and 8% slower than the fast Asus G752VS and the relatively slow Acer Predator 17 in the graphically demanding Fire Strike Graphics benchmark. In 3DMark 11, it performs 20% and 8%, respectively worse. We can (with reservations) still say that Max-Q is much more efficient in combination with a GTX 1070 compared to a GTX 1080, which – as we will see in the next section – seems to be working a lot in the range above the Sweet Spot clock by default anyway.

Except for Prey, we can repeat what we said about the GTX 1060 Max-Q here. The performance loss through Max-Q does usually not cause limitations in 1080p resolution. However, if a standard GTX 1070 can deliver just sufficient FPS at certain settings, the frame rate becomes critically low for the MSI with Max-Q GeForce graphics card.

We currently have to rely on the new Asus Zephyrus GX501 for our performance analysis of the GTX 1080 Max-Q, which will be reviewed shortly. Besides our own scores, we can also use reference scores from Nvidia. Acer kindly provided a 3DMark 11 score for its upcoming Triton 700 as well. We have once again included a notebook with a particularly fast and one comparatively slow regular GTX 1080 (Alienware 17 R4) in the comparison.

It seems that the Sweet Spot clock of the 1080 is further away from the TDP related maximum clock compared to the GTX 1070 Max-Q. Our 3DMark scores (GPU) of the Zephyrus (perhaps due to the newer driver) are slightly higher than the reference scores from Nvidia, so we have used them and can see that the Alienware with the comparatively (!) slow standard GTX 1080 is 12% faster; the powerful Asus even 33%. This will probably be an issue for many potential GTX 1080 buyers. They will want the highest possible performance from a very expensive GPU.

This is only a first impression, a comprehensive analysis does not make much sense at this point. The results of “Tomb Raider” in particular, however, suggest that Max-Q GPUs are more competitive in games compared to synthetic benchmarks.

Note: We do not have any information about the GTX 1060 so far. However, notebooks with this GPU suffer less from the mentioned issues in the first place and the performance headroom for Max-Q is lower, so we do not expect a lot.

It is actually too early for a verdict since we only have a limited number of results. We would have liked to give you detailed information about the power consumption, heat development, and system noise at this point. First results will be  available in review of the Asus Zephyrus GX501.

What can we say so far? The performance drop is very limited in combination with a G-Sync display and the GeForce GTX 1070, and current games are usually still playable at the maximum graphics settings, although this obviously depends on the resolution as well. There are no news to the GTX 1080 so far.

The GeForce GTX 1060 Max-Q performs up to 25% worse than the standard notebook GPU with the same name. Hence, it might have it hard in practice. In our opinion, the main reason is, among others, that it is not too difficult for manufacturers to produce sufficiently quiet and cool laptops with GTX 1060 without Max-Q. Hence, the Max-Q variant appears obsolete.

Below our verdict, there are several diagrams which simply compare the Max-Q graphics cards 1060, 1070, and 1080 reviewed by us.