While specialist end users typically have workflows for genomic sequencing, frequently they don't know how to relate it to processors or the amount of storage that they need. Michael McManus, director of precision medicine at Intel, has been an end user himself, and brings to the table the ability to translate what end users need in terms of compute storage capability to the actual system. Dell technologies has taken Intel’s methodology and used it to create a modular and really scalable approach.
What does Dell technology’s approach provide for customers?
You can buy a smaller system based on needs today, and if you get more sequences in the future, or when more genome sequencing is needed, you can expand that system. Again, it is based on these building blocks for compute and storage and based on what the needs are. So, you can linearly grow the system and get more capability along the way.
What support is available to partners?
In addition to the methodology, customers sometimes also ask for benchmarking and want a hands-on approach. Dell offers access to its Solution Centres globally, including the Innovation Lab, its main facility in Austin, Texas which is available to all of its customers free of charge and other large systems dedicated just to HPC. These genomics pipelines and systems are available via remote access for customers to try out for themselves so that they can get an idea of how the systems will perform. Similarly, Tech Data will soon have a facility available at its Bracknell location for customers to visit.
Dell also offers access to its own experts, who are users themselves, familiar with the pain-points researchers face and can discuss common issues in depth, in the vernacular that researchers themselves use.
A typical customer workflow
In the typical customer workflow, businesses are doing the actual sequencing. This includes going from tissue or blood to extract the DNA, all the way to raw data to be processed on these Dell systems. So, their concerns are whether they have the right compute capacity to manage the volumes they are dealing with. If they are successful in their business, and they get more samples to sequence, they need to know if they are going to run out of compute or whether they can stay ahead. So, they are looking for a partner like Dell to advise them on how to best design their system and grow with it.
Who should partners be talking to regarding an opportunity?
It is actually the researchers, not IT managers that partners should be talking to about the systems at customer sites. One of the things that Dell has done really well is to take what the researchers are talking about, in terms of their requirements, including genomics processing, coverage, how many genomes they would like to process on a weekly or monthly basis, then translate that to system size.
Different users have different experiences and different requirements sometimes, so you should talk to as many users as you can. Ultimately, you do want to be able to provide a system that meets everybody's needs or as many people's needs as possible.
What pain points should our business partners be looking for when they're speaking to their customers for opportunities?
Discussions frequently focus on the processing of whole genomes as DNA, but customers often have other workloads that they want to be confident could also run on that system. So, the systems are designed to encompass a wide variety of workloads. Customers want to know if one of their investigators or one of the people in the lab wants to try something different, they run it on the hardware. The answer is usually yes, unless it’s something very esoteric.
Intel systems are designed to do both general genome assembly as well as ‘de novo’ assembly. The standard kind of clinical use case is a reference genome, which is a human reference and involves processing human data. Some places using these systems aren't like clinical labs, such as agricultural companies (ag bio companies) which may be looking for the next version of some sort of genetically modified crop that they're producing. Most of those kinds of activities are done using no reference genome. So, that requires a new method, or a different method called de novo assembly. Dell systems are equipped with the ability to handle these kinds of ‘de novo’ assembly workloads, in addition to the human reference genome workloads.
Once they have all of the information, how do partners get started?
Working with Tech Data and Dell, partners can be trained on how to approach customers. They need to identify targets and qualify them quickly. Dell acts as a resource for sellers to reach out to, as well as Tech Data. Once these customer conversations have happened a few times, sellers then have the confidence to have them on their own with customers.
Remember that researchers want to tell you about their challenges, so being inquisitive is the best approach. You don't have to be a rocket scientist; you just have to ask the right questions. Then, Tech Data, Dell and Intel can offer support on the significant opportunities and challenges these partnerships present.