It has been two generations of Max-Q graphics cards and now we have to say goodbye for some quite strange reasons on the part of NVIDIA. As we already dealt with this in the corresponding news item, we are going to present other (possibly more accurate) reasons for making the decision.
NVIDIA ditches its Max-Q laptop GPUs
Why not stick with the Max-Q and Max-P line of graphics cards at the same time? Above all, when two ranges are perfectly segmented in performance and price, giving more users the opportunity to choose.
The first reason has to do precisely with performance and secondly with the manufacturers, who curiously retouched these GPUs. As we well know, the Max-Q models were intended to offer a lower consumption than the Max-P, so the TGP of the former was always lower than in the latter.
This was done so that slimmer and more stylized laptops could have an NVIDIA gaming GPU without compromising too much on the cooling of the equipment and instead, the buyer would get good performance in their games.
But as time went by, the AIBs realized that these capped GPUs had great potential for optimizing TFLOP per watt consumed, as NVIDIA had better BIOS SKUs to improve their efficiency. Therefore, raising the frequencies with lower voltages resulted in it being possible to gain performance and even exceed a Max-P model without problems, with a similar consumption and without generating more heat than its superior model, but for a lower price.
What was the point of buying a laptop model with GPU Max-P if certain models with Max-Q performed the same or better? It didn’t make sense, so NVIDIA decided to cut its losses …
Simplifying models is always a good option for the consumer, or maybe not?
The second reason would be somewhat controversial, not to say extremely controversial. Removing the Max-Q models, only Max-P models remain, that is, the models without the tagline, such as the RTX 3070 Mobile to dry and not the RTX 3070 Max-Q.
But because the Max-P models are now going to be segmented by speed, NVIDIA has really reduced the number of GPUs per name, but has boosted its distinction into a single series by speeds.
To the point that there will be, at least initially, up to 38 GPU models encompassed from the RTX 3060 to the RTX 3080, something that does not make sense from a practical point of view, and much less if as you are seeing it is not They specify the characteristics on the box to be able to discern each of the speeds.
Faster production speed, lower costs, fewer quality controls, more models on the market
NVIDIA released Max-Q and Max-P as standards for manufacturers and for their own chips, but while this was wonderful for identifying models, it was actually wasting time and money for the entire industry, including them.
The Max-Q models have been limited in terms of maximum loudness for almost two years, although finally a large number of laptops passed the sound test, the reality is that at home they exceeded the thresholds, in some cases by much. Therefore, the test ISO-7779 It was of little value as such, since, although the maximum thickness was defined, although the voltage ranges were minimal, the tests were finally skipped without too many problems and costing time and money, since each laptop had to be tested.
Therefore, adjusting specific frequencies for a single model and segmenting them as the manufacturer wants is something that is part of it and must be specified, since now with the Max-P models these have more freedom of configuration and TGP, as well as tolerances of noise or chassis design.
Undoubtedly, this is going to be an exciting topic to deal with based on the experiences and problems that users encounter, since access is not being facilitated, manufacturers are being given freedom to make and undo the policy with their models of names and segments.