Can flash storage bring certainty to the chaotic world of tech?
Kaminario CEO Dani Golan: "2016 was the worst year financially since 2002, but we raised the money because we could demonstrate for most of our customers how cost efficient flash could be."
All-flash storage company Kaminario recently secured $75m in a new round of financing, bringing the company's total funding to $218m. ZDNet spoke with company CEO and founder, Dani Golan, to find out where it's going next.
ZDNet: Why is flash proving so popular?
Golan: What's attracting investors to this area -- and our latest round of funding was heavily oversubscribed -- is that companies like ours now sit in the middle of the two most disruptive innovations in IT: cloud and flash.
One is a totally different conception model and the other is the means to get there.
The reason to start there is because while flash by itself has been around for decades, flash is everywhere now, and that's because it is now reliable enough and cost efficient enough to penetrate the enterprise in large quantities.
But, by itself, flash is not interesting enough so that is why I always talk about the three ways of disruption in the flash world.
Now in the beginning, flash was used for one thing and one thing only: speed. If you want to run your trading application a bit faster, you just go out there and put a bunch of flash modules inside the machine and immediately it would run faster.
Then you had a few companies who said: "You know what? You can take flash to the next stage and do some things interesting with it." So, they started to look at small portions of the market, at the infrastructure and a few things such as storage for SMBs and so on. Every area could be different and you could help it do things faster and better.
We looked at this and we recast the company in late 2010/11 so that it was dedicated to the next generation infrastructure, and we felt that that would be cloud based, 100 percent cloud.
What we realised back then -- when flash was $1,000 a gigabyte -- was that the day was coming when it would cost $10c a gigabyte and flash would be doing everything.
No longer would it be penetrating a certain type of application or a certain type of workload, flash would do everything.
But that would not just be because it was so much faster but also because it was much more scalable, easier to use, and, at the end of the day, much more cost efficient.
Now that $75m [which we recently] raised was done in a very tough market, and in a very tough year. 2016 was the worst year financially since 2002, but we raised the money because we could demonstrate for most of our customers how cost efficient flash could be.
That's why we could raise in private finance the sort of amount that really resembled in size a public offering.
How does your flash work?
At the heart of what we do is our very elegant media flash. With that we looked at the difference between what IT has traditionally done and why the cloud is growing so fast.
Businesses are being asked by the customer to run faster and be much more adaptable to the changes. IT, if you think about it, was the complete opposite: slow to move, slow to grow. And every new product could take months, if not years, so you had this disconnect between the business and its IT capability.
And the cloud really emerged more as a concept: what if the different stacks could grow very fast, could shrink very fast, could adapt to changes in load, changes in performance needs? When you look at the heart of cloud adoption, that was the key.
Now, if you look at 85 percent of the storage market over the last 30 years, it has been dominated by half a dozen companies that have monolithic products that can't be transferred to the modern cloud architecture. They're not agile, they're not scalable. That's the difference between the mainframe and a modern cloud infrastructure.
So, that's when we decided that the future of IT rested on flash and that 100 percent of our datacentres would be flash, and at the end of the day flash was going to get cheaper and cheaper.
Now in that way, last year was critical because we truly got to a cost advantage, at both the capex and op-ex levels. Now that is very powerful when you can step into your customers and show them that no matter what legacy solution they have, we can do the same thing far more efficiently.
Once customers are deploying Kaminario as part of their operation, they are very fast at adopting it in other areas. To me, that is the strongest proof that it's addictive. Once you go down this route it is very hard to go back.
In the long term, do you see flash overtaking the conventional solutions?
Software and services is by far the most successful part of the public cloud. This year, the public cloud will be $130bn and by 2020 it is going to be close to $200bn. By far the biggest part of that is SaaS (software-as-a-service).
The companies that are really building their software growth are the SaaS companies that are doing it in a very mobile way, a very agile way. The largest and fastest growing companies are SaaS companies.
The big guys are fighting old battles. They are fighting a shrinking IT budget, they're fighting legacy architecture, and so they are moving now into a SaaS deployment model. And that's our strength.
In your own systems, what proportion of your customers are going for 100 percent flash against 60 percent or some other number?
Among our customers, around 25 percent are going all flash for everything. There is no spinning disk in their environment, period.
Then approximately 50 percent are doing their main, business critical applications with flash and they are doing it with their important, money generating applications.
And another 25 percent are keeping theirs for critical processes. It could be real-time analytics, it could be a data warehouse, and it could be some modules that are in their first year of adoption and they are not jumping in head first.
I can give you an example: Telefonica. Telefonica is running its worldwide Wi-Fi operations on Kominario. Priceline is running their whole real-time analytics on Kominario. Nuance, the largest Siri voice recognition company in the world, is a user.
That company and all the others are really using us because of three things they can get from us that they can't get anywhere else.
Number one is the ability to scale one system as much as they want. You can grow and mix and match different capacities and performance as much as you want.
Number two is cost efficiency. We believe that we can offer better economics than any of the others
Thirdly, if you ask any CIO what their challenges are and how far into the future they are looking, they will say that we are living in a chaotic world. We can't predict anything. They want an infrastructure that can handle anything and that is where we step in. We can work with any combination of application and that is exactly what they need.
What's new, then?
We're expanding and using the proceeds of this round of financing in three major areas.
One is a very fast international expansion. Two is that we will continue to work very closely with our ecosystem.
When you look at who adopts our solutions very quickly, it is the financial institutions, healthcare, and high-tech in general.
Finally, if you look at how our product is structured you will see ours is the only infrastructure that offers complete separation of computer and storage. That allows us to adopt new technologies when they come in. Things like NVMe and NVMe Fabrics are things that we predicted. Protocols are shifting in the storage world and we believe we have the most comprehensive vision in that area.