Building Hardware During a Pandemic

The global pandemic has affected the technology industry in a number of ways. While it has led to more growth in cloud and data center services as many of our routine daily activities such as work, school and entertainment move online, it has brought about new challenges in hardware design, manufacture and distribution. This is impacting both current and new products alike.

At TORmem, we are focused on realizing our vision of One Memory for All. A key part of this strategy is designing, manufacturing and distributing our disaggregated memory appliances. Our entire product development cycle is taking place during the pandemic, which creates a unique set of challenges for our team as we navigate this unprecedented industry situation.

Designing for Availability

Although we design our products to operate in a variety of failure conditions with features such as redundant power supplies and different memory failover modes, the type of availability that we are talking about here is not operational availability; it is component availability, something that technology consumers don’t often get a lot of visibility into. The global supply chain for everything from complex processors to the right kind of screw is complex and ever-changing.

The pandemic, and other contributing factors such as rare weather events and environmental change, have led to shortages of many key components across the world. As fabrication and packaging plants for processors, memory modules and other vital system ingredients have encountered a mix of production interruptions and large-scale demand, acquiring these parts can be challenging. When a fabrication plant is shut down, even for a very short period due to power availability or a freak weather event such as happened in Texas earlier this year, there is no quick fix to get production back on schedule and make up for lost time. Like stopping the presses for a newspaper, there is a significant period of time required to resume production.

Being a startup with significant hardware design, manufacture and distribution experience has many advantages in this situation. TORmem is able to redesign our products around many key component shortages while preserving the original functionality and quality of our designs, in a way that many larger companies with existing product lines in mass production cannot. As our team is so close to the design and manufacture of our products, we are able to quickly redirect around supply issues as they become apparent to avoid delaying bringing our products to life.

To date, our team has created several versions of each of our upcoming disaggregated memory appliance products, each with many iterations, to design around component availability issues. Every key component from machine screws to memory controllers is a target for multi sourcing, and is able to be designed out if it is simply unavailable within the timeframe that we need it.

Sourcing and Transportation

Once a suitable set of components has been identified for our hardware design, our team faces the dual challenges of sourcing them economically, and transporting them efficiently to where they are needed for our products to be assembled. With our strategy of multi-sourcing as many components as possible, we are well-positioned to circumvent any single vendor who is having issues supplying us on time, and with our global partners, we can avoid many shipping troubles.

Much of our team is located in the United States, but we work with global manufacturing and component sourcing partners. For any of our products, components may originate from Taiwan, Germany, the United States and several other countries to build our finished product. This is not unusual, as many products utilize global component sourcing; but where TORmem stands out is our ability to adjust where specific tasks for component integration take place, as we can shift where our boards are manufactured or our chassis are prepared based on our needs. We are not limited to a rigid production pipeline, where a single hiccup can cause production to back up.

By designing for availability and working with our global partners to ensure the efficient supply and transportation of key components, TORmem is able to circumvent many of the issues that are impacting the global supply chain. This is crucial as many of these concerns, such as the availability of many silicon products such as processors and memory modules, look to carry on into 2022. Companies who are not able to take this flexible approach are likely to have to pause product production during key times of high demand, whereas TORmem will be ready to deliver.

Experience and Flexibility

Even though we are a startup, our team has many years of experience at both the engineering and executive levels to work around any design or supply issues that we encounter. Both our CEO, Thao Nguyen, and our CTO, Steven White, have extensive backgrounds in hardware engineering which gives us a unique perspective into how component issues will impact us.

Experience at both of these levels is vital. Without the relationships we enjoy with our global partners, we would be limited in our ability to ensure we have access to key components, and without our entire team’s understanding of the intricacies of our hardware and software designs, we would run the risk of sourcing parts which are not fit for purpose in our products. This type of tight integration between hardware, software and our supply chain is an example of TORmem’s team philosophy; to work quickly as a small team, making sure that concerns from one area of expertise are properly understood across the others, without creating silos or slowing us down.

Although there are plenty of challenges ahead in bringing our vision to market, we believe that TORmem and our global partners are well-positioned to succeed in our ever-changing world.

About TORmem

At TORmem, we believe in One Memory for All, our vision of high-speed disaggregated memory at data center scale, for enterprise, cloud and HPC use cases. Decouple your memory from your servers, speeding up today’s applications and enabling tomorrow’s, while optimizing costs. 

Find out more at