Soldered RAM has great advantages. But it continues to complicate expansions and the right to repair

Soldered RAM has great advantages. But it continues to complicate expansions and the right to repair
Soldered RAM has great advantages. But it continues to complicate expansions and the right to repair
  • Criticized for preventing expansions and making potential repairs very difficult, these modules have compelling reasons to prevail in the market.

  • The option to traditional SO-DIMM modules is already underway: CAMM2 modules are promising

Technology is increasingly at odds with the right to repair. Manufacturers make it very difficult for users, and although sometimes their reasons for doing so are debatable, there is at least one area in which the arguments are powerful. Specifically, that of laptop RAM memories.

Soldered RAM, increasingly popular. In recent years we have seen how more and more manufacturers have been betting on RAM memories soldered to the motherboard. The option, which at first seemed to become popular among laptops dedicated to productivity, has been gaining strength among gaming laptops and it is now very normal to see a laptop have this type of memory option.

critics. Criticism was frequent from the beginning for an obvious reason: having the RAM soldered prevents users from expanding it later, and repairing the component in case of problems is also almost impossible. However, as they point out in DigitalTrends, there are several arguments that explain this decision among manufacturers.

Argument 1: space. These types of modules allow something crucial for many manufacturers: they save space, allowing our laptops to be thinner. Not only that, but they also contribute to better cooling of the interior of these devices and to be able to include larger capacity batteries.

Argument 2: faster. Manufacturers have been taking advantage of LPDDR (Low-Power DDR) modules for some time, which are more efficient and help extend the autonomy of our equipment. The problem is that these memories are always soldered, but not only are they more efficient, but above all they are faster. LPDDR5/X modules can reach 7467 MHz, while DDR5 modules reach 5600 maximum for Meteor Lake chips. It is true that the final impact for users is not extreme, but together with efficiency it is still a compelling reason for manufacturers to choose it.

It’s not because of the price. HP and Dell executives commented in that article how there are practically no differences in the cost of components or the manufacturing process. By welding the components, labor can be saved because this work can be done by robots in the assembly phase, but both commented how the final difference in cost is not appreciable and this is not a factor when deciding whether to opt for soldered memories or traditional SO-DIMM modules for these devices.

But the disadvantages are there. What is clear is that in equipment with soldered RAM, repairability is impaired: if the component ends up damaged for any reason, the repair is very complicated, if not impossible, and having the equipment still under warranty is the lifeline. . The other big argument against is the lack of options to expand that RAM in the future. If we buy a device with 8 GB now because it was cheaper and we want to expand it to 16 GB in the future, we will not be able to do so, for example.

You have to decide well when buying, be careful with the cost. It also often happens that although manufacturers argue that price is not a determining factor, the difference between 16 and 32 GB configurations can easily range from $100 to $300. The impossibility of expanding this memory later makes it necessary to make a decision about the memory configuration when purchasing the equipment, something that is difficult since we will not know what our future needs will be.

Apple has gone even further. The introduction of the Apple M1 chips made Apple make an even more radical decision in this regard. Memory ended up being part of the die in the SoC along with the rest of the components such as the CPU, GPU or NPU. The decision favors transfers between all elements of the system, but maximizes those arguments against: (practically) zero repairability, and enormously high cost if we want to make the leap to configurations with more memory.

Future buyers, a little note. If you are thinking about purchasing a new laptop, it wouldn’t hurt to keep in mind that a foreseeable avalanche of artificial intelligence options is coming for both Windows and (probably) macOS computers. In these tasks the NPU is essential, but be careful, because the faster the memory, the better. And certainly, the more memory the better, although it is true that for most scenarios today 16 GB is a highly recommended amount.

CAMM2 modules to the rescue. In April 2022 Dell launched its DDR5 modules with a new format called CAMM that mixed the best of soldered memories (faster and more efficient) and traditional SO-DIMM modules (repairability, expansions). The idea ended up resulting in an even better CAMM2 universal standard that represents a small revolution in the sector—it will be possible to have LPDDR5X memories without soldering them—if they end up being adopted by manufacturers. Samsung already has plans to do so, and other manufacturers are expected to follow in their wake.

Image | Artin Bakhan

In Xataka | How to check the status of RAM in Windows and if you have problems

 
For Latest Updates Follow us on Google News
 

-

PREV “Cultural appropriation”: Japanese criticize Assassin’s Creed Shadows for giving prominence to an African samurai
NEXT Apple releases operating system update on iPhone and iPad, what does it bring new?