Representatives of the specialized Intel laboratory in Oregon, who are engaged in extreme overclocking of processors, do not believe in "horror stories" about the exhaustion of the overclocking potential of modern products manufactured using advanced lithographic technologies. If the operating frequencies of 7nm AMD processors are close to the limit, this does not mean that future Intel processors will not leave room for overclocking by users.
In recent months, Intel's top executives have been talking a lot about the prospects for mastering the 7nm process technology. Substantial funds are already allocated for this task, but Intel believes that the key to success is the reasonable setting of goals in the field of geometric scaling, since excessive ambitions have already tarnished Intel's reputation for mastering the 10nm process technology. After the transition to the 7nm process technology, Intel expects to return the so-called "Moore's law" to its previous course, and to change lithographic technologies every two or two and a half years. In addition, as part of the 7nm technology, Intel will begin to use lithography with ultra-hard ultraviolet radiation (EUV), albeit with a noticeable lag compared to its main competitors.
Intel's first 7nm product will be a GPU for the server segment, which will be part of the Ponte Vecchio computing accelerators. It will use Foveros' complex spatial layout and will enter production by the end of 2021. Next, server CPUs will have to switch to the 7nm process technology, but this will not happen until 2022. For consumer processors in this context, the prospects for an early transition to 7nm technology are still vague. To begin with, it would be nice to deal with the 10-nm process technology, which Intel is in no hurry to use in the desktop segment.
Run, processor, big and small!
Site representatives before the New Year, I managed to visit a specialized Intel laboratory in Oregon, where a team of eight people tests processors and motherboards compatible with them for overclocking potential. It is necessary to carry out such work not only with an eye on the needs of a small group of enthusiasts who are engaged in extreme overclocking. Limit operating modes allow us to understand the "margin of safety" of both the processors themselves and related components. In addition, such experiments allow us to estimate the residual frequency potential of each new generation of Intel processors.
By the way, the employees of this laboratory made it clear to journalists that they keep their finger on the pulse of the market and have an idea about the current capabilities of competitor products in the field of overclocking. In addition, they expect to work closely with the developers of Intel discrete graphics to provide users of the latter with familiar overclocking tools.
When representatives of Tom's Hardware asked the head of the laboratory Dan Ragland (Dan Ragland), whether overclocking can be considered an endangered craft against the backdrop of a reduction in the frequency margin of the competitor's 7nm processors, he resolutely objected to journalists. Phenomena observed when overclocking competing brand processors released by TSMC should not be transferred to future Intel products in advance.
Firstly, even within the framework of the 14-nm process technology, the company managed to significantly increase the frequency potential, and this is taking into account the trend towards an increase in the number of cores. Secondly, as we move to new stages of lithography, the frequency margin will always be preserved. Perhaps for some processors it will be less, for some it will be more, but representatives of the Intel profile laboratory are not going to say that overclocking will become obsolete over time. On the other hand, they acknowledge that as the transition to more “thin” technical processes, the overclocking potential of Intel products will decrease, albeit not always evenly.
Source: 3dnews.ru