Selective memory: why focusing on price over quality will cost you in the long term

This article originally appeared in the July'23 magazine issue of Electronic Specifier Design – see ES's Magazine Archives for more featured publications.

However, not every memory chip is born equal. The disparity between cheap alternatives and quality memory storage units can make or break an operation. This is true whether that’s security forces’ bodycam footage or a video drone used to assess a mining operation. These two examples, where high-definition audio and visual data are essential, shows just how important an industrial memory chip is: if it fails, it could risk peoples’ safety and your business’ time and capital.

Nicholas Small of Transcend says: “The lesson for industry is clear: price should not be your guiding priority. Going cheap where it matters is a false economy, as operability, reliability, and longevity will be sacrificed. This may be fine for a piece of throw away consumer tech, but in an industrial context, letting the price decide what you buy inherently leads to an inferior product which can put everything at risk.”

Why memory matters

It doesn’t need to be stated how important video and audio footage is in the modern world. However, there are stories that really highlight how essential video evidence is. One such story is that of Yuliia Paievska, or Taira as she is known in Ukraine; a medic who recorded 256gb of footage documenting her team’s efforts to treat and save injured or dying Ukrainian and Russian soldiers in May 2022.

Taira managed to get the footage to one of the last international journalists working in Mariupol, for the Associated Press (AP), who smuggled the tiny memory chip out of Russia inside a tampon. This video documented her valiant efforts and helped shine a light on the severity of the conflict.

Due to her arrest by Russian security forces, the footage also exemplifies Russian war crimes – as arresting or harming medics or civilians is against the Geneva Convention. If this memory chip had failed, we would never have heard about Taira or her story.

Why industrial memory chips are different

The above example may be particularly unique, however there are lessons for industries and businesses here too. The most important being to buy industrial grade memory components. Consumer grade chips are not built for industrial purposes.

John Fitzpatrick, of Solid State Supplies, said: “If memory should fail in low cost toys, people do not consider this as a problem. If used in police or security cameras that are used to store evidence of crime, it could have a devastating consequence.”

There are several reasons for this. Firstly, it simply comes down to build quality. For example, consumer grade chips are made from the outer core of chip ‘wafers’, which are of poorer quality and will more commonly have defects in memory size or the number of read and write cycles that are possible. Industrial memory components are selected from the inner wafer which will support an exact tolerance as an industry standard.

This increases the chip’s lifetime, compared to consumer grade components which typically have a purchase availability of 6-12 months before being replaced or superseded. Many industrial grade products can exist for more than ten years with only one or two Product Change Notifications during that time – and PCNs often have a six-month notice period. This is an essential lead time; businesses can then replace, upgrade, or otherwise ensure their end product stays up to date, reliable, and working in the field.

Secondly, the testing process for consumer chips is not as thorough. Even if a chip passes a test once, it does not guarantee reliable consistency, and different chips made from the same wafer, as said earlier, will not necessarily be of the same quality. The result is inconsistency in your end product – with chips failing to stand the test of time or use.

The stark difference in product testing impacts how and where these chips should be used. Industrial grade products are tested for extended periods at a wider range of temperatures – usually between -40°C to +85°C. Industrial manufacturers, knowing the context in which these products will be used, can also increase this resilience to extreme temperatures through screening or conformal coating options.

Consumer grade products, on the other hand, are tested at only 0°C to +70°C, limiting their use in much of the world’s winter season. If we consider Taira’s work in Mariupol, the memory chip and therefore the footage would not be guaranteed in their winter temperatures.

These differences create a more resilient and reliable chip, which translates directly into the number of programme erase (p/e) cycles that is possible. Although any component can come in a wide range of qualities, the best examples of an ‘industrial standard’ can be hundreds of times more durable than the most basic consumer alternative. In context, this means that if a typical, consumer-use 32GB consumer USB drive has around 500 p/e cycles, the highest quality, industrial equivalent could be anything up to 100,000 cycles.

These facts will only become more important

“This fact is only becoming more salient. Products are becoming thirstier for data – needing larger amounts of storage, for higher quality audio, or video footage and data for example. This means that a chip will need a larger re-write area; yet another reason to focus on quality over price,” said Fitzpatrick.

As stated in the reasons before, using consumer grade chips – or letting their price point influence which chips you use in your products for that matter – can only lead to an inferior product.

As Nicholas of Transcend puts it: “This will impact your company’s reputation and may put your customers operations or even their safety at risk.”