Building chips is becoming more and more expensive, margins are getting lower and lower, and the union of the two will make many computer manufacturers move towards using components from the PostPC device market such as tablets, smartphones and others. types of devices that use similar hardware in order to build new types of computers.
One of the peculiarities of the different ranges of each product is not only the ability to reach different purchasing power levels, but each range is presented in a different way, with the high range being the one with a lower level of demand, but with the highest profit level and the low end of the market which has a greater volume but with extremely low margins.
The increase in manufacturing costs in the chips together with the push of PostPC devices may end in the creation of a new type of PC in the low-end of the market, completely replacing PCs as far as internal hardware is concerned, not so in functionality.
The “End” of Moore’s Law as a driver of changes in the low-end range
Moore’s Law tells us that every so often the number of transistors in a given area doubles, allowing for more complex processors and systems.
In recent years we have heard about the death of Moore’s Law, this statement is due to the product of aligning the statement of Moore’s Law with the cost per transistor, which has been increasing in such a way that the cost per The area has not remained intact or has fallen but has risen.
This means that if a manufacturer has a budget per chip, it finds that the area that it can use at the same cost is less, unless it decides to raise prices, which is counterproductive for the low-end market that is already moving. with margins
The consequences in the design of processors are already being felt
In order to use processors with a smaller size, both Intel and AMD, which have decided to take different paths to alleviate the increase in cost per area in their new processors.
At AMD, the solution has been separating the CPU cores and their caches from what is the rest of the processor, Northbridge and Southbridge, placing all the elements in an MCM configuration.
In Intel, the solution has come from separating only the Southbridge, but in a 3DIC configuration, as seen in Lakefield, in which we have all the I / O interfaces on a separate chip.
The birth of a new type of computer
To understand why the amount of PostPC devices disguised as computers will appear in the following years, we must understand a phenomenon that is none other than the increase in the size of the chips not only removes fewer chips per wafer, but also each of these chips has a higher percentage of failure than the smaller ones.
PostPC devices use SoCs with an area that usually does not exceed 120 mm2, which is a much smaller size than PC SoCs, as well as the CPUs and GPUs that we usually use in our PCs, which makes them extremely attractive in expensive to want to keep costs low.
With the increase in the cost per mm2 in wafers, this leads to the PC in the lower part of the market being doomed to disappear, since the margins are going to be much lower, but the SoCs for smartphones cannot cover the needs in what to the connectivity capacity of peripherals that PCs have.
The future based on Chiplets is not only for the PC
In an interview with Alex Katouzian of Qualcomm, he discussed the separation of their SoCs into multiple chips, which is known as a chiplet-based configuration, which is interesting given the fact that Qualcomm does not make processors for PCs.
It’s a very good question, look, I think it will still take a few years. And let me give you an analogy, you know, but you can make a comparison to solar panels. Let’s say you are 100% efficient or close to what you are spending. You could add more panels, but then you would enter diminishing returns. As for the technology (of making chips) I think we still have a few years left until we can integrate everything in one SoC.
Obviously the chip area starts to grow, but the downsizing of the transistors cannot keep up with the growth in functionality. But, as I said, the efficiency of reusing a process (or manufacturing node) or a modified version of it is giving us some capacity, so I think it will still be a few years before we can think about separating the chip in such a way it makes sense for us to do it in parts.
The other thing to keep in mind is that you are talking about a heterogeneous system on a chip, which means that there are parts of the chip that depend on others to operate. It’s not as simple as saying, I’m going to take the graphics apart and put them somewhere else, or remove the CPU and put it somewhere else. We are in the process of studying it, but I would say that in three or four years, our architecture will change, but we still have a few years before this happens.
These statements from Qualcomm are important because we can think from the outset that due to the smaller size of the SoCs for PostPC devices then there is scope to create larger SoCs, but it does not seem that this is going to be the path they will take since the The approach will be to separate part of the SoCs functionality for PostPC devices on a separate chip.
Post-PC devices with the expandability of a PC
If we were to choose the differences between a PC device and a Post-PC device outside of the use of the ARM register and instruction set on one side and x86 on the other, the key answer would be the expandability of the different types of devices. .
For example, a laptop PC may need several USB ports, some video output and even ports to expand the storage capacity. This would mean having to build a new SoC, which has those interfaces. On the other hand, the Southbridge is separated from the rest of the chip, this allows manufacturers of SoCs for PostPC devices to rethink their SoCs for use in very low-power PCs without having to create another chip model.
This opens the door to a new type of device, which are devices with PostPC processors but with the expandability of a laptop PC or an AIO, all thanks to the separation of the Southbridge from the SoC. These devices will be placed in the lowest part of the market and will be a result of the increase in the cost of wafers that has been in the making for some time.
But we can also see computers with a complex Southbridge that allow several PCI Express ports, for example, the safest under CXL protocol, this makes it possible to use dedicated graphics cards in this new type of computer.
The end of x86 computers in the low-end of the PC?
With everything explained above, it is clear that the low-end in the case of the PC has the sword of Damocles on top. One of the things that we will most likely see in a few years from both Intel and AMD, are hybrid processors between x86 and ARM, these will be processors that will take the code written for x86 and reinterpret it in ARM code using of special units for it.
These processors will not do the reinterpretation continuously, but during the installation of an application what the coprocessor in charge of the interpreter will do is translate the code to ARM, it will not do it as efficiently as a direct compilation.
The key is that the coprocessor in charge of the interpreter can be added not as part of the main unit but as part of the separate Southbridge.
