Storj Acquires PetaGene, Now Offers Distributed Object And File Storage With GPU

‘Rather than building out data centers ourselves or building, buying and running equipment, we instead leverage equipment that is already being run, powered and cooled around the world,’ says Storj CEO Ben Golub.

Distributed cloud object storage technology developer Storj Tuesday unveiled the acquisition of PetaGene, which offers distributed file storage technology.

The acquisition of Cambridge, England-based PetaGene follows Storj’s July acquisition of Valdi, which provides on-demand GPU computing. With the two acquisitions, Storj is in a position to provide distributed object and file storage and GPU processing in a combined, on-demand fashion, said Ben Golub, CEO of the Atlanta-based company.

Golub described Storj as a distributed cloud services company for distributed workloads offering all the same kinds of services people might get from Amazon Web Services, Google Cloud or Microsoft Azure, but in a different fashion.

“Rather than building out data centers ourselves or building, buying and running equipment, we instead leverage equipment that is already being run, powered and controlled around the world,” Golub told CRN. “But most drives are less than 30 percent full. Most GPUs are less than 40 percent utilized. And so basically, we take all this spare storage and compute capacity and combine it in a way that gives customers the same kind of services that they might get from [cloud hyperscalers], but in a way that’s completely more secure, closer to the edge, faster, and at a fraction of the price.”

Storj has infrastructure in 100 countries in the world and currently does about 40 percent of its business via indirect channel partners, Golub said. “But ultimately we want to be more channel-driven,” he said.

PetaGene is the developer of the cunoFS file system, which Golub said will help his company offer distributed file storage as well as object storage.

“And in particular, they’re incredibly fast,” he said. “They end up being a really good solution for our two main markets, which are the media video marketplace and the emerging AI marketplace.”

That AI marketplace stems from Storj’s Valdi acquisition, which brought it into the GPU compute space, Golub said.

“Valdi had a very similar model to us in that we were using spare capacity on drives all around the world and they were using spare cycles on GPUs and CPUs all around the world,” he said. “We tied them all together. And that business has just been taking off like gangbusters. The combination is really good, and half of our customers are now using both.”

For Storj, the acquisition of PetaGene gives it a fast track to combining its object storage technology with file storage, Golub said.

“They spent upward of eight years building this,” he said. “It was the best that we saw out there by far. We partnered with them first. We have a pretty extensive technical and go-to-market partnership program, and we saw that what they had built was miles ahead of anything else. And what we really liked about it was that basically any object storage infrastructure, whether it was ours or something a customer might be running on-premises or even in Amazon or Microsoft, could be helped with one comprehensive file system on top of it. And frankly, we also got great feedback on them from our go-to-market partners.”

Storj currently has no plans to add block storage to its mix, Golub said.

“Block tends to be best for databases and things like that that are inside a single data center,” he said. “But we're really focusing on workloads that happen at the edge where customers create, collect, consume and analyze data. And that’s really where the power of our model works.”

Golub declined to discuss terms of the PetaGene acquisition but said he expects PetaGene will bring his company several million dollars of revenue in the first year.

“PetaGene wasn’t yet profitable, but they built up a really large customer base,” he said. “I believe they have customers in the bioinformatics and pharmaceutical space, which is great because we do a lot of scientific work, and it is really a nice complement to our customer base.”

Storj is a private company that has no venture capital or private equity investors, Golub said. “We’re not yet cash-flow-positive, but we do have great unit margins,” he said. “And unlike most cloud companies, we don’t have to spend billions of dollars in capital building out data centers or revenue.”

Storj is in no hurry to do more acquisitions, Golub said.

“I think we need to digest this,” he said. “It gives us a nice combination of solutions. And as I mentioned, we are a very partner-friendly organization. We have a big network of technical partners who help us adapt for different use cases, and a large set of go-to-market partners as well. We think we don’t need to buy much more. We need to be a solid platform and then be good at working with others who can take us into different markets and who know their particular markets and their customers.”