Jeff Clarke On Dell's New Midrange Storage And AI Strategy
Dell's Jeff Clarke Isn't Satisfied Yet
The overhauling of Dell's product strategy and portfolio by Jeff Clarke is driving massive market share gains for the global infrastructure leader in storage and servers. Clarke is looking to build on Dell's widely successful 2018 by launching a new flagship midrange storage product next year while also injecting artificial intelligence and machine learning capabilities "across all of the portfolio."
Clarke, vice chairman of Products & Operations at Dell, has worked side by side with Dell Technologies founder and CEO Michael Dell for decades and is the driving force behind the company's storage sales charge. "I just removed the barriers for us to be successful," said the 32-year Dell veteran Clarke, in an interview with CRN. "It's working. I'm pleased. But we're not done. It's not all fixed."
Clarke talks to CRN about the company's storage strategy in 2019 and Dell's artificial intelligence and machine learning future.
Can you give us any details on Dell's new midrange product coming out next year?
It is going to be a modern file system and modern software stack – better described as, it's going to be a micro-service-based architecture and a container-based architecture. Once we get that foundation in, my ability to iterate new features and new capability is going to be absolutely faster than we've ever done. I committed to the end of next 2019 that we'll have the product ready. I have a lot of work to do, it's roughly a year away, but that foundation and that lift of a new file system and a new modern software stack is going to be a platform for us to go from NVMe to NVMe over fabric to storage-class memory to memory persistent architecture over time. I like my chances there.
What's your message to Dell EMC partners selling midrange storage right now?
Two things we have been doing. One, we have kept the current products competitive. In fact, increasingly more competitive – whether that's data reduction rates, deduplications, new cloud-like software capabilities that we've been adding to the portfolio, particularly the midrange, because we need to. We're closing any perceived competitive gaps which we've largely already closed. My promise is that my new product -- that we call 'Destination New Product' -- for every existing midrange product, I will have a data migration capability to the new platform that is seamless.
How big of a leap will that new midrange product be over the competition?
I think it's pretty good. It's going to be a competitive product. We think we understand what the requirements of the market will be, what our customers need and the technology available at that time. We've built a compelling product. … For our midrange storage, we put our very best minds in storage working on this. We need to deliver it and deliver it timely.
What's the storage strategy for 2019?
We will spend more in storage R&D next year than we spent this year. We're having two-times as many developers working on the new things than we had just 15 months ago. They're working on new innovative things and looking down the pipeline. We now have an incredible and capable product management team looking at the customer and market requirements; a product development team and a CTO team. Our ability now to be able to have a conversation about what a storage subsystem should look like [for] 2023 requirements, 2025 requirements – we hadn't been doing that in awhile.
We're not done with storage [growth]. We've stabilized the business. We now have a trend of share gain. We have a lot of share gain to claw back that we gave back [to competitors] over the past four years. We took 580-basis points of share gain in first quarter 2018, 100-basis points of share gain in second quarter, 260-basis points in third quarter. I just removed the barriers for us to be successful. I used my experiences of taking a highly capable and talented team to focus on fewer things. It's working. I'm pleased. But we're not done. It's not all fixed.
How much artificial intelligence and machine learning is inside Dell's storage portfolio?
Our PowerMax makes six billion decisions a day. It’s got artificial intelligence (AI) built into it ahead of anybody else. It has the greatest AI engine in storage. For a storage administrator, it does the tasks on its own. You have the world of artificial intelligence workloads to help customers and do their analytics, then there's the role of AI and machine learning that makes their way into the products. If you start with, for example our Precision Workstations, we have AI in those products so we can understand the characteristics of the application being run and we can recommend multiple configurations to optimize performance. You go to the other extreme, our high-end PowerMax leading product in the market, it has 30 years of Dell EMC knowledge in its AI engine of how to run, build and administer storage. It makes 6 billion decisions a day. The AI engine is incredible.
What is Dell's AI strategy going forward?
[AI] reduces the tasks a storage administrator needs to do because it's making the decisions, which is how I look at artificial intelligence and machine learning as it makes its way into our products. Many of the tasks that humans do today, we're going to automate so those humans can do other higher-value tasks. The data migration work that we're doing is going to have an AI engine in there so we know how to pull off our legacy arrays to be able to pull it into the new array. You'll see us apply artificial intelligence and machine learning across all of the portfolio, as well as build AI-specific workloads servers like we did with the R940xa, for example.
Talk about how you successfully rebuilt Dell EMC's storage business.
One of the changes I'm most pleased with and we're witnessing the early dividends of is, we've become one team. Our storage organization was siloed and compartmentalized. It was to the point where you had fierce supporters of a brand or a product inside the organization and we were not necessarily stepping back and looking at market needs, customer needs and technology trends. My view 15 months ago, I had XtremIO, ScaleIO, VMAX, Unity, VNX, SC – all competing almost internally with one another and dividing up engineers and not sharing IP across the boundaries. It was highly fragmented and we weren't all rowing in the same direction, some weren't even in the same boat. Fast forward to today, I only have one storage team and one portfolio that we look at. We have PowerMax, a new midrange product by the end of next year and we have an entry level product with the PowerVault ME4. We've become a single team and now we share IP, engineers, algorithms and methods that quite honestly we couldn't before.
With more data being generated thanks to the Internet of Things, will there be an explosion in the available market around storage?
There's a ten-fold increase over the next six years of how much data is going to be created. Increasingly, more of that data is created on the edge and that data on the edge has to be handled in real time. I don’t think we have looked through the entire pipeline and can accurately know how much storage is going to be needed, but a ten-fold increase is a lot. Even if some of it will be thrown away after it's processed, you still have to do short-term storage to be able to run the analytics and that data has to be protected and secure because it's going to be at the edge of the network. To me, that bodes well for us.
What storage form does that actually take?
Is it going to be more traditional arrays, or a more hyper-converged infrastructure (HCI) appliance-like implementation out on the edge, or is it going to be more storage in a server or another form of software-defined storage – I don’t know the exact answer. I do know that we have the best assets in the software-defined world and we'll have the best assets in the traditional array world. We're positioned uniquely and we also have servers. So whatever way it’s going to happen, we have an advantage position of a leader in servers, a leader in HCI and a leader in primary storage.
So what are the key things you're working on in 2019?
Prior to 15 months ago, we weren't necessarily as aggressive in our cloud positioning as perhaps you thought we would be. We launched our Cloud Marketplace that puts all our cloud assets in one place. We talk about key elements of our cloud architecture, we talk about a multi-cloud world and the role that we play. You'll continue to see us build out more cloud assets through next year.
The operational challenges that we had a year ago, they're largely behind us today. We focused and simplified the product development teams on fewer things and doing them better. We have a lot of work to do to complete the work, but I'm pleased we've become one team that shares resources, IP and common methods, and we're reducing the amount of replication and duplication in the product development community. There's more people working on newer things that will bode well in time. That's not changing in 2019.
Will Dell Technologies becoming a public company change your strategy at all?
No. No changes. We're playing for the long run.
How should partners prepare themselves for next year?
My advice to a channel partner is that you want to be able to partner with someone who has a full solution set across how architectures will be deployed and how customers will want to drive outcomes. We're uniquely positioned there. We have our Pivotal cloud-native application for new application methodologies. We have Boomi for data integration. We have VMware as this operation hub or control plane. Then our world class infrastructure sits underneath that – from servers to data protection to primary arrays to converged infrastructure and HCI. It's going to clearly be a multi-cloud, hybrid cloud world. We have unique properties to help customers navigate that. Partners are going to want to provide that value and help customers implement the digitization of their business. More of the customer conversations tend to be about, 'How do I use this technology to drive better business outcomes. Help me be more competitive against my competitors.' Then you throw new workloads like artificial intelligence and machine learning and the outcomes it can drive, it's a pretty exciting time for partners.