A lot of engineering resources are tied up in updates, while not adding much to existing platforms.
Intel is currently undergoing a major reorganization, which will certainly include cutting the research and development department. The company has practically acknowledged this, but how is still in question. According to a recent rumor from manufacturing partners, the company has chosen a very conservative path for the time being, as the development of the intermediate generations has largely been stopped. These are the platforms where the original design receives a very minimal redesign, almost only chip variants based on new stepping are coming, while numerically they are set for a generational change.
Although updates are an integral part of product portfolios, they actually have little added value, as they mostly copy the basic platform in terms of capabilities, but they tie up a lot of engineering resources due to the necessary validation background. Intel would withdraw this resource, so the workforce could be devoted entirely to true generational changes, which these days tend to come in two-year cycles.
A change in strategy in this direction would not actually harm the company’s competitiveness much, since an updated platform usually brings a maximum of 10% extra performance, which in itself is not bad, but if the financial framework for implementation is very limited, it is more worthwhile to give it up and use all resources regroup for a real generational change. Only the OEMs would be at a loss because it is of utmost importance for them to get something new every year, but the latter can be solved simply by renaming the current generation, the financial costs of which are negligible, since the validation procedure associated with the update can be saved.
Although all of this sounds strange, it can basically be considered a very thoughtful decision, because Intel thus withdraws engineering resources from a place where they suffer the least disadvantage. To some extent, AMD already works this way, because for example there was no intermediate generation of Zen 4 (Ryzen 7000) processors on the desktop market, but the predecessor was followed two cycles later by the Zen 5 design (Ryzen 9000). And the fact is that it pays off, because even AMD could bring an extra 10% at the level of the intermediate generation, but with the kind of investment it entails, it’s simply not worth it. Intel tries to pick up the pace of its competitor with a similar strategy, which works spectacularly better, and this will ultimately only be good for the competition.
Source: prohardver.hu
