Danny Allan, the now-former chief technology officer at Veeam, is moving to developer code security business Snyk.
Snyk’s CEO, Peter Mackay, posted a comment on Allan’s LinkedIn post about his departure from Veeam, talking up the “next chapter” of Allan’s career “with his friends at Snyk”.
Mackay was co-CEO and President at Veeam from July 2016 to November 2018 and worked at VMware, Desktone, IBM and Watchfire before that. Allan was also at Veeam during the same period, and has spent time at VMware, Desktone, IBM and Watchfire too. The pair of execs appear to go way back.
Danny Allan.
Adi Sharabani, who also used to work at Watchfire and IBM, became a Snyk investor and advisor in November 2018 and remains in these roles, according to his LinkedIn profile. He also was appointed as Snyk’s CTO in May 2022 and left that role in June last year. So Snyk has a CTO vacancy and Allan is moving to the position.
A Snyk spokesperson told B&F: ”Danny is joining Snyk as its new CTO on Tuesday.” That will be February 6 and Snyk will make a formal announcement then.
Snyk was founded in Tel-Aviv and London in 2015 by Assaf Hefetz, Danny Grander, Guy Podjarny, and Jacob Tarango. Podjarny was the founding CEO and he gave way to Peter Mackay, who was an investor in the company, in June 2019. The headquarters are located in Boston.
The business has raised $1.2 billion in funding, according to Crunchbase, with the last G-round in 2022 pulling in $196.5 million to give Snyk a $7.4 billion valuation. However an earlier $304 million F-round in 2021 was at a higher valuation of $8.5 billion. There was a $25 million corporate round last year with ServiceNow the investor.
Privately-owned Snyk reportedly achieved a 153 percent revenue increase from 2021 to 2022 to $147 million, when it had in excess of 2,300 customers. It had 1,135 employees at the end of 2022 but laid off 128 of them in April 2023.
A researcher at the Rochester Institute of Technology (RIT) has devised a microfluidic lab on a chip that can perform artificial neural network (ANN) computation on data stored in DNA.
DNA storage relies on data being stored as specific combinations of the four nucleobases – cytosine (C), guanine (G), adenine (A), and thymine (T) – found in the double helix formation in the DNA biopolymer molecule. One method has pairs of these nucleobases symbolizing binary ones and zeros. Amlan Ganguly, computer engineering department head in RIT’s Kate Gleason College of Engineering, co-authored a scientific paper in which he envisions “a computing platform using DNA molecules that are capable of computation in-situ without the need for domain conversion of information from DNA to electronics.”
Ganguly states: “DNA is excellent at storing information, in fact, it is much better than the electronic modes of memory because it is about 3-to-6 orders of magnitude more compact than most memory hardware that we have; it is also much more reliable and durable.”
He adds: “We proposed to represent numbers through concentrations of solutions containing specifically manipulated DNA molecules and computing operations as manipulation of DNA molecules – operations like addition and multiplication and other non-linear functions necessary for network computations can be performed. That is the bridge from storage to computation and using DNA as a vehicle to do the computation.”
Diagram from Ganguly’s paper
The paper states: “While biochemical reaction representing computations using DNA molecules are several orders of magnitude slower than electronic gates, their data density is 3 orders of magnitude higher and 8 orders of magnitude lower in energy consumption than solid state memory.”
The RIT integrated circuit (IC) based on microfluidics is suited for “highly dense, throughput-demanding bio-compatible applications such as an intelligent Organ-on-Chip or other biomedical applications that may not be latency-critical.”
The research paper abstract says: ”It computes entirely in the molecular domain without converting data to electrical form, making it a form of in-memory computing on DNA. The computation is achieved by topologically modifying DNA strands through the use of enzymes called nickases.”
Data is represented stochastically through the concentration of the DNA molecules that are nicked at specific sites. A stochastic process has random or probabilistic dynamics over time, and the randomness is modeled mathematically to study the statistical properties of the process. The probabilities of different outcomes can be analyzed.
Ganguly says DNA computation and storage uses less energy than electronic storage and computation: “We are in the age of big data that needs to be stored somewhere. We don’t think that more datacenters are the answer, or even the best answer. Each datacenter requires the equivalent of a city block of power. Building, maintaining, and operating more traditional datacenters is not sustainable.”
Comment
DNA storage has little chance of replacing electronic storage media such as disk and solid state drives, which have millisecond-class data access speeds, and DRAM and CPU/GPU computation operate at microsecond-class speeds or faster. This is because reading and writing with DNA, meaning sequencing and synthesizing DNA molecules, is slow. It needs hours. Ganguly admits that computation using biochemical reactions is several orders of magnitude slower than using a CPU or GPU.
Startups that are over ten years old often disappoint their early venture capitalists (VCs) due to prolonged timelines in yielding returns. This locks up their invested cash, making it unavailable for other ventures, like startups with quicker exit potentials.
Silicon Valley VCs are taking a leaf out of the private equity playbook by setting up continuation or secondary funds. These funds purchase investments in mature startups that have not yet exited through acquisition or IPO. The cash is used by the VC to return money to its investment partners, the LPs (Limited Partners), while the continuation fund retains the VC’s holdings in the startup.
According to the UK’s Financial Times, a VC startup investment can have a ten-year run time with a possible two-year extension. After that, and if there is no return prospect, they can shut the startup down or sell their holding at a discount.
Here is a table, not an exhaustive one, listing some long-life storage startups:
Many appear to be growing, such as ExaGrid, which claims to achieve record revenues virtually every quarter, and yet do not IPO.
Ditto the two object storage suppliers, Cloudian and Scality. They’re growing, but competing with ten-year-old MinIO and its $126 million in funding. Meanwhile, all the main IT storage suppliers, except HPE, have their own object storage tech, meaning an acquisition looks unlikely.
Many storage startups are thus delaying an IPO. Some, due to high valuations, face problematic acquisition exits – as in, who can afford them? Others face an unattractive IPO landscape as investor interests have shifted from the areas popular during their founding days to newer fields like big data analytics and AI. The same can be true of acquisitive businesses that may, these days, also want to invest in generative AI technology businesses and not technology that was a good idea ten years ago.
The continuation fund concept provides a way for worn-out VC investors to get some money back from their holdings in these companies, and thus provide liquidity for new investments.
Insight Partners has set up a continuation fund and moved 32 of the companies in which it had investments into the new fund. Insight’s LPs received $1.3 billion as a result. Potential LPs for a VC will look at the capital it has distributed to LPs in the past and, if this has shrunk, they’ll look elsewhere to for a dependable return on their cash.
In effect, the continuation fund is a way for VCs to transition towards becoming private equity players.
The FT also reports that Lightspeed Venture Partners is talking to investors about setting up a $1 billion continuation fund for ten of its holdings. It has $25 billion in invested assets, including holdings in Nest, Snap, and, of particular relevance to our storage focus, Rubrik. We are not saying that any one of these is heading for continuation fund status. Rubrik seems more likely to IPO than not.
The continuation fund notion raises two questions in our mind. First, will the valuation of VC-held companies change if they are moved into continuation funds?
Secondly, will the continuation fund VCs, adopting more of a private equity mindset, start involving themselves assertively in the management of such startups, and restructure them to become profitable and acquisition-ready? Long-life startups cannot be VC-funded businesses forever. It’s an IPO, acquisition exit, or private equity management via the halfway continuation fund house for them.
Bootnote
The Institutional Limited Partners Association has various continuation fund resources for investors, both limited and general partners in VC funds.
Hitachi Vantara and Cisco are partnering to sell hybrid cloud managed services based on Cisco compute and networking, and Hitachi Vantara’s storage and management software.
Update. Hitachi Vantara and Cisco have changed the partnership’s original and temprary Hitachi | Cisco Hybrid Cloud name to the Hitachi EverFlex with Cisco Powered Hybrid Cloud. 3 February 2024.
The two suppliers set up global partnership agreements last summer, with Hitachi Vantara entering Cisco’s Service Provider and Solution Technology Integrator (STI) partner programs. The intent was to integrate Cisco UCS servers and its Nexus networking products with Hitachi Vantara storage such as the VSP arrays. That way Hitachi Vantara could represent itself as a more of a whole datacenter IT systems supplier and Cisco gets a more effective partner pushing its gear into enterprises.
They have announced the Hitachi EverFlex with Cisco Powered Hybrid Cloud with a portfolio of offerings such as infrastructure, IT, storage, and containers as a service, as well as data protection, to be sold on a consumption basis through Hitachi Vantara and Cisco’s partner channels. These offerings can be used on-premises or in the public cloud.
Kimberly King, SVP of strategic partners and alliances at Hitachi Vantara, said in a statement: ”Hitachi EverFlex with Cisco Powered Hybrid Cloud is a strategic response to modern enterprise needs. Our ability to deliver both through our partners and Cisco’s network directly tackles the complexity challenge by providing strong IT operational capabilities that can be scaled based on customer data management needs.”
Cisco’s Alexandra Zagury, VP of partner managed services, added: “Our joint efforts with Hitachi Vantara signify a great opportunity where we’re going beyond cloud capabilities to enable a holistic approach to business transformation.”
Hitachi Vantara is providing a managed services capability, hybrid cloud management, and predictive analytics. The two companies say their combined offering provides hybrid cloud acceleration, automation to increase efficiency, and added security and compliance.
Comment
The background here is that enterprises and public sector organizations tend to use a combination of on-premises and public cloud IT, with consumption-based billing applied. Hitachi Vantara competitors have announced branded hybrid cloud programs such as Dell’s APEX, HPE’s GreenLake, NetApp’s Keystone, and Pure’s EverGreen.
Hitachi Vantara has been busy setting up partnership deals such as one with CTERA for cloud file services, Model9 for mainframe data access, and SQream for massive data set access.
It wants to sell its own products and services as part of whole systems to its enterprise-class customers. This type of deal is a continuation and reinvention of the old converged infrastructure (CI) deals Cisco participated in with Dell and others back in the day.
Cisco devised its CI-based Cisco+ hybrid cloud service based on existing FlexPod and FlashStack deals with NetApp and Pure Storage in June 2022. The Hitachi EverFlex with Cisco Powered Hybrid Cloud is another example of this kind of deal and shows how long-lived the whole concept can be.
Australia-based data manager and orchestrator Arcitecta has hired Eric Polet as Director of Product Marketing. Previously, he served as Product Marketing Manager at Spectra Logic, where he developed strategic product positioning and messaging to differentiate the company’s solutions in the market. Arcitecta has a partnership with SpectraLogic. Polet is based in Longmont, Colorado.
…
Cohesity research found that 97 percent of UK businesses have paid a ransom after a ransomware attack in the last two years, despite almost all companies having a “do not pay” policy. Also 83 percent had been the “victim of a ransomware attack” between June and December, and 73 percent said their company would be willing to pay over £2.4 million ($3 million) to recover data and restore business processes. The research polled more than 900 IT and Security leaders in the US, UK, and Australia.
…
Commvault announced the members of its Cyber Resilience Council chaired by Melissa Hathaway. They are:
Roland Cloutier, Principal, The Business Protection Group, former Chief Security Officer, Tik Tok
Shawn Henry, Chief Security Officer, CrowdStrike, former Executive Assistant Director, FBI’s Criminal, Cyber, Response and Services Branch
Mark Hughes, President, Security, DXC Technology
Nancy Wang, Cybersecurity Investor, former GM of AWS Data Protection and Data Security
John Zangardi, CEO, Redhorse Corporation, former CIO, US Department of Homeland Security
…
Lakehouse supplier Databricks has acquired Einblick, a German startup comprised of experts in machine learning, human-computer interaction, and natural language processing. The Einblick team has spent the last four years pioneering techniques for translating natural language questions into the code, charts, and models needed to generate insights. Einblick built a sophisticated multi-step architecture that processes raw user input, enhances it with relevant contextual information, and breaks it into smaller solvable chunks using SQL, Python, and higher-level logical operators. By joining forces with Databricks, the ideas behind Einblick can be more powerfully extended through integration with the underlying data catalog and Databricks’ Data Intelligence Platform. Data intelligence helps users simply ask questions of their data platform and get detailed, high-quality answers.
…
UK-based Ad Signal is partnering DataCore’s Perifery business unit to provide media customers with seamless deduplication workflows. Ad Signal’s SaaS Match software, which deduplicates copies of video, image, and audio files, can now integrate with Perifery’s Object Matrix to help media teams automatically identify duplicate assets. The news follows Ad Signal’s recent growth spike, with revenues topping £1 million ($1.27 million) as the team has nearly doubled in size.
…
Kioxia SSDs are on their way to the International Space Station courtesy of HPE. They were carried on the SpaceX Falcon 9, Cygnus NG20 resupply rocket, which delivered an updated HPE Spaceborne Computer-2, based on EdgeLine and ProLiant servers. This provides edge computing and AI capabilities on the ISS including real-time image processing, deep learning, and scientific simulations. Kioxia provided 8x 1 TB XG NVMe SSDs, 4 x 960 GB SAS and 4 x 30.72 TB enterprise SAS drives for a total of 134.72 TB. This is the most data storage to travel to the space station on a single mission. SSD health will be monitored daily throughout the mission, with daily log files transmitted from the ISS to be tracked and analyzed by Kioxia in order to better understand how flash memory storage operates in space.
The Spaceborne Computer-2 enables data to be evaluated in low Earth orbit in near-real time, making it possible to achieve a 30,000x reduction in download size by only transmitting the data output, or insight, to Earth instead of the raw data, thereby drastically reducing download times.
…
The UK arm of distributor TD SYNNEX has added Lenovo TruScale pay-as-you-go infrastructure to its portfolio, giving partners the ability to provide customers a cloud-like experience along with the reassurance of having systems deployed on premises. TruScale provides hybrid and multi-cloud environments, virtualization, hosted desktop, scalable storage, HPC, and other technologies.
…
The Memory Fabric Forum is a MemVerge initiative designed to accelerate CXL education, solution development, and co-marketing. It is running aQ1 Memory Fabric Forum Webinar CXL for the Enterprise on February 8. Here’s the Zoom link to register to attend.
…
Quantum announced that Australia’s Amidata has implemented Quantum ActiveScale object storage as the foundation for its new Amidata Secure Cloud Storage Service for active and cold data. Amidata has already built Backup-as-a-Service and File Sharing Service offerings on Quantum’s DXi and StorNext products.
…
Quantum announced that the Listing Qualifications Department of the Nasdaq stock market has approved its plan to regain compliance and Quantum has been granted an extension to file its Form 10-Q for the second fiscal quarter on or before May 7.
…
Cloud-based search and analytics supplier Rockset announced a new instance class, resulting in 30 percent reduction in compute costs. The class is general purpose with different memory to CPU ratios optimized for low-cost search and AI applications. It has a lower entry price to enable developers to start building real-time analytics and AI-powered search applications starting at $232 per month. There are autoscaling compute capabilities to scale compute based on the workload. Rockset has had a strategic investment fromHewlett Packard Pathfinder, the venture capital program of HPE.
…
Scale Computing has launched a VMware Rip & Replace Promotion for partners transitioning their business from VMware in the wake of Broadcom’s acquisition. Partners bringing customers looking to migrate from VMware to Scale Computing Platform (SC//Platform) will receive a 25 percent discount on Scale Computing software and services, as well as free migration tool access, complimentary Scale Computing Advanced Training Certification, and a free registration pass to the Scale Computing customer and partner event, Platform 2024.
…
SIOS Technology, which supplies application high availability (HA) and disaster recovery (DR), promoted Sajid Shaikh to Vice President of Engineering, reporting directly to COO Masahiro Arai. Previously, Sajid served as SIOS’s Director of Engineering.
…
SoftIron has added facilities to its private, on-premises HyperCloud, with new nodes that allow customers to tackle tougher workloads:
Veeam has launched a Cyber Secure Program combining Veeam data protection software with a team of experts to help enterprises prepare for, protect against, and recover from ransomware. It starts with pre-incident support including architecture planning, implementation assistance, and quarterly security assessments. When there is an attack, customers are connected with Veeam’s dedicated Ransomware Response Team and the program offers post-incident support to enable rapid recovery. There are three main components:
Design and implementation assistance to ensure Veeam best practices in implementing Veeam solutions to the highest security standards. Customers receive advanced seven-phase onboarding support and rigorous quarterly security assessments.
The Veeam Ransomware SWAT team (ransomware recovery “black belts”) is available 24/7 when a ransomware attack or cyber incident occurs. There are prioritized 30-minute SLAs and customers have a dedicated Support Account Manager (SAM) for assistance and escalation, plus access to specialized senior support engineers.
The Veeam Ransomware Recovery Warranty – up to $5 million in data recovery expense reimbursement for a verified attack.
Commvault earned $216.8 million in its third fiscal quarter ended December 31, a record amount for the biz, yet a quarter-on-quarter downturn has been forecast.
Revenues for the data protector were $216.8 million – up 11 percent year-on-year driven by a 31 percent increase in subscription revenue. There was a $17.1 million profit – a world away from the year-ago $310,000 loss, and representing 7.9 percent of revenues.
CEO and president Sanjay Mirchandani noted in prepared remarks: “Our Q3 results exceeded expectations, including double-digit year-over-year growth across our most important KPIs. By our own metrics, this was an exceptional quarter. We also set the stage for the future, by introducing market-leading innovation with Commvault Cloud, our revolutionary platform for cyber resilience.”
He added: “We’re two months away from closing our fiscal year, and I couldn’t be more excited about our momentum as we approach FY 25.”
Wells Fargo analyst Aaron Rakers observed that Commvault “sees an ‘amazing opportunity to accelerate growth’ in FY 2025.”
Financial summary
Operating cash flow: $44.4 million
Free cash flow: $42.6 million
Gross Margin: 82.9 percent up 0.1 percent on the year
Cash and cash equivalents: $284.3 million and no debt
Total ARR grew 17 percent year on year to $752 million
Subscription ARR grew 29 percent year on year to $571 million
SaaS ARR increased 77 percent year on year to $152 million
CFO Gary Merrill declared in the earnings call: “Our execution was strong as large software deal close rates improved sequentially and we delivered against our largest term subscription renewal quarter of the fiscal year.”
Commvault introduced Commvault Cloud powered by Metallic AI in November to unify all its SaaS and software offerings on one platform, with single pane of glass management, threat prediction and recovery features, and an Arlie – short for “Autonomous Resilience” – generative AI copilot using Azure OpenAI.
Mirchandani claimed this was “the most important pivot in our 27-year history,” and Commvault reckons it now has a strong SaaS offering in a dynamic growth market. Subscription revenues were $114 million in the quarter – up 31 percent year on year. It added 500 new subscription customers in the quarter, taking its customer count total to 8,800.
The outlook for the fourth quarter is for revenues of $212 million give or take $2 million – growth of just 4.2 percent at the midpoint and a sequential decline on the latest quarter. The reasons for this anticipated slowdown in growth are not immediately clear. It doesn’t appear to be seasonal as there has been a third-to-fourth quarter revenue increase in six of the last eight financial years.
Merrill discussed the outlook: “For fiscal Q4, we expect subscription revenue, which includes both the software portion of term-based licenses and SaaS, to be $111 to $115 million. This represents 20 percent year-over-year growth at the midpoint. This Q4 subscription revenue outlook reflects continued momentum in our new customer and expansion business, but a smaller renewal pull in fiscal Q4 relative to Q3.”
William Blair analyst Jason Ader told subscribers: “Revenue guidance for the fourth fiscal quarter was above consensus by $1.8 million at the midpoint (up 4.2 percent year-over-year). Though the renewals base is sequentially lower in the fourth quarter, subscription revenue was guided above consensus by $5.8 million at the midpoint (up 20 percent year-over-year), driven by strong SaaS momentum and an improving pipeline.” He mentioned that “fiscal Q3 had the largest renewal pool we’ve ever had in our history.”
In other words, the annual comparison is more important than the sequential comparison and it’s the lower renewals number that drags the outlook down.
The full 2024 outlook is for revenues of $828 million up or down $2 million – a 5.5 percent increase at the midpoint on fiscal 2023. There was no outlook for fiscal 2025, but Merrill opined: “We have an amazing opportunity to accelerate as we move forward. Obviously we have not given guidance for FY 2025 yet. Clearly our expectation [is] that FY 2025 growth will be higher than FY 2024.”
Samsung revealed it is planning 280-layer QLC 3D NAND SSDs via an International Solid-State Circuits Conference (ISSCC) agenda.
The ISSCC conference will be held in San Francisco and a February 20 session presented by Samsung staff is entitled “A 280-Layer 1Tb 4b/cell 3D-NAND Flash Memory with a 28.5Gb/mm² Area Density and a 3.2GB/s High-Speed IO Rate.”
ISSCC 2024 agenda item
As reported by Germany’s ComputerBase, Samsung’s new V-NAND gen 9 flash has 280 layers, not the 300 previously associated with this generation. It’s a QLC (4bits/cell) technology and has, Samsung claims, industry-leading density of 28.5 gigabits per square millimeter as well as a fast IO rate for QLC flash.
UFS fingernail cards and M.2 gumstick cards will benefit from the higher density of these flash chips, given their limited physical space. YMTC has a 1 Tb QLC chip with a density of 20.62 Gb/mm² at 232 layers, while Micron’s density is 19.6 Gb/mm² using 232 layers, and WD/Kioxia is at the 13.86 Gb/mm² level with 162 layers. SK hynix has a 512 Gbit QLC chip with a 14.40 Gbit/mm² density at 176 layers.
Increasing the layer count is the key to increasing a chip’s density in Gb/mm² terms. At 280 layers, Samsung has the highest layer count flash, with SK hynix next at 238 layers, YMTC and Micron with 232, followed by WD and Kioxia with 218. SK hynix says it has a 321-layer chip coming next year. That could well give it a density edge until the other suppliers catch up.
Samsung chart.
A Tom’s Hardware report suggests that this Samsung flash could even enable 16 TB M.2 SSDs. That seems fanciful as current M.2 drives top out at the 2-4 TB level e.g. Sk hynix Platinum P41 at 2 TB, Seagate FireCuda 530 at 2 TB, Samsung 990 Pro at 4TB, and Micron 7400 Max M.2 at 3.2 TB. A jump to 8 TB might be feasible but 16 TB seems like a stretch too far.
It’s a year since IBM integrated Red Hat’s Ceph storage product roadmaps and wants us to know that it’s making progress in this increasingly AI-dominated environment.
Gerald Sternagl
A blog going live today written by Gerald Sternagl, manager of Technical Product at IBM Storage Ceph, says: “This self-healing and self-managing platform is designed to deliver unified file, block, and object storage services at scale on industry standard hardware. Unified storage helps provide clients a bridge from legacy applications running on independent file or block storage to a common platform that includes those and object storage in a single appliance.
”Software-defined storage has emerged as a transformative force when it comes to data management, offering a host of advantages over traditional legacy storage arrays including extreme flexibility and scalability that are well-suited to handle modern uses cases like generative AI.”
Sternagl is critical of IBM’s legacy storage array hardware, such as its mainframe DS8000 and x86 server FlashSystem arrays, but he was a Red Hat vet with more than 10 years service before IBM acquired the company.
In his view: “Ceph is optimized for large single and multisite deployments and can efficiently scale to support hundreds of petabytes of data and tens of billions of objects, which is key for traditional and newer generative AI workloads.” It can support data lakehouse and AI/ML open source frameworks and “more traditional workloads such as MySQL and MongoDB on Red Hat OpenShift or RedHat OpenStack.”
There is “a feedback loop where generative AI thrives on the abundance of unstructured data, and the continuous generation of realistic data by AI further enriches and refines your understanding of unstructured datasets, fostering innovation and advancements.”
Some 768 TiB of raw Storage Ceph capacity is included in watsonx.data, IBM’s data lakehouse architecture for data, analytics, and AI workloads.
Sternagl says: “Organizations … need a storage management solution capable of accelerated data ingest, data cleansing and classification, metadata management and augmentation, and cloud-scale capacity management and deployment, such as software-defined storage.” It also needs to support both the on-premises and public cloud environments.
By software-defined storage, he means Ceph of course. His company is not promoting MinIO, Cloudian, Scality, DataCore, or WekaIO here.
In December, IBM updated Ceph with object lock immutability for ransomware protection and previews of NVMe-oF and NFS support for data ingest into the underlying Ceph object store.
Comment
An issue with any three-way combination of block, file, and object storage is that each access protocol implementation has to be aware of the others and this can delay and possibly limit the adoption of new features, such as NVMe-oF and NFS support. If you need all three protocols supported in a single software package then Ceph is a good choice but you may find that block-only, block and file combined, file-only, or object and file combined may support new features faster and also provide speedier data access.
Data protector HYCU is using Anthropic’s Claude generative AI model to produce connector code for a customer’s SaaS app data to be protected by HYCU’s Protégé service.
Customer data in SaaS applications is not protected by the app provider; it’s the customer’s responsibility. There are tens of thousands of such apps and no backup provider can write connector code to backup customer data in all of them. HYCU is hoping customers take up its R-Cloud SDK so the app providers themselves could add a HYCU connector to their apps, and more than 60 such connectors have been built.
HYCU founder and CEO Simon Taylor said in a statement: “This development with Anthropic’s frontier generative AI model Claude is more than an integration; it’s a leap forward in the future of data protection.”
The Claude model has been trained with HYCU information and uses a SaaS app’s API to build a connector between it and HYCU such that HYCU can backup and restore customer data. This can cut development time from three weeks or so to a matter of hours.
R-Cloud module generation process.
Taylor told an IT Press Tour briefing that, with Claude: “Customers can button-hit to get R-Cloud to automatically generate an application to protect a SaaS app. … We trained the model. … This bot writes a complete and extensible app and provides testing for it.”
R-Cloud module generation success.
HYCU says the training ensured that Claude “fully understand the nuances of compliance, encryption, and recovery protocols, custom-tailored for HYCU’s platform.”
The launch of HYCU Claude means that a customer could generate code for a SaaS app protection which could duplicate what other customers have done. HYCU is considering setting up a marketplace for the distribution of such customer-sourced, Claude-built apps. It had a target of having 100 SaaS app connectors built by the end of 2023 which did not come to pass, and anticipates Claude-built connectors could send the connector count sky-rocketing. It will stop publicly counting after 100 connectors are available.
R-Scout protection module development query.
HYCU Claude is a significant development from HYCU’s earlier introduction of its R-Scout, an AI -powered assistant for customers and partners to learn about the details of HYCU’s offerings, and how best to use them.
The company has also added the capability to protect data in PineCone and Redis Cloud databases which are both used as vector databases in AI work.
Taylor said: “By harnessing AI, we’re not only accelerating our development processes but also reinforcing our commitment to security and operational efficiency. We’re excited to take R-Cloud to the next level and pioneer this space and set new standards for the industry.”
Comment
This is a novel and significant idea which could be adopted by other industry players. There are potential licensing and pricing aspects of such customer-generated but supplier-built extensions to a software product or service. It is a step on from co-pilot-based assistants created by Commvault, Druva, Rubrik and others and gives them a development direction.
Customers are clamoring for cloud-based file data services, and both CTERA and Nasuni are growing strongly on the back of this.
Such services have a cloud-based object storage vault storing files, which are served to accessing customers’ distributed datacenters, remote sites and end-points – typically with local caching to speed access. The central cloud repository synchronizes any file state changes. Startups CTERA and Nasuni are two such suppliers, with Egnyte and Panzura also active.
CTERA announced record results for 2023, driven by strong demand for its AI-based CTERA Ransom Protect. It recorded over 30 percent growth in annual recurring revenue for 2002 to 2023. It revealed it won major deals with state agencies, leading banking institutions, and global media groups. CTERA also expanded its presence in Asia, winning customers in India, Australia, and Singapore. It manages hundreds of billions of files comprising more than 200PB of data.
CEO Oded Nagel issued a statement: “We are thrilled by the staggering market reaction to our recent product releases, which combined with the expansion of our strategic partnerships, have played a pivotal role in increasing our reach and impact.”
Competitor Nasuni also had a good 2023, with a 30 percent growth in ARR. CEO Paul Flanagan stated: “Over the last five years, we have averaged a steady annual recurring revenue growth rate of 30 percent, and we are proud that we’ve done it again in 2023. We have maintained top decile key performance indicators in our business allowing us to be cash flow positive for the year.”
Nasuni has, he claimed: “Over 800 customers, 500 employees, and a very strong balance sheet that has enabled us to avoid any reduction in workforce.” It enjoyed 46 percent growth in new customer Annual Contract Value bookings and won more than 120 large enterprise customers in 2023, with its first eight-figure customer contract for the Nasuni File Data Platform product.
Like CTERA, ransomware protection was a focus – with an attach rate of 71 percent for its ransomware detection on new customer sales in 2023.
Competitor Panzura has just appointed a new CEO to improve its growth rate.
Kioxia and Phison are both developing UFS 4.0 fingernail flash cards and controllers, with Kioxia currently sampling its device.
UFS (Universal Flash Standard) cards are used for storage and data transfer in smartphones, digital cameras, tablets, smart TVs, augmented reality and virtual reality headsets, and automotive video and logging applications. The standard was developed to provide faster and higher capacity storage than eMMC (embedded MultiMediaCard) flash. The standard has progressed through several generations, with UFS 4.0 the latest:
UFS gen 2 and 3 cards are available now with gen 4 cards coming on stream.
Phison is launching a range of UFS controllers to support entry-level, middle, premium, and flagship smartphone devices. The line looks like this:
UFS 2.2 PS8327 with 64 GB capacity and the smallest UFS 2.2 controller.
UFS 3.1 PS8325 forhigh-end smartphones, supporting the latest 1 Tb NAND from various manufacturers plus Phison’s 4KB LDPC (Low Density Parity Check) error correction engine. It has secured design wins from many international leading smartphone manufacturers and is expected to start mass production in the first quarter of this year.
UFS 3.1 PS8329 for mid-range smartphones. It is the smallest UFS 3.1 controller on the market and is planned to begin sampling in June of this year.
UFS 4.0 PS8361 for high-end smartphones with a four-channel design. The highest read and write speeds of PS8361 will exceed 4,000 MB/s. Improved read performance per mA to extend the battery usage efficiency of mobile devices. PS8361 has been designated for use by NAND manufacturers and is expected to begin shipping in the second half of this year.
Samsung UFS card
Kioxia is sampling a UFS 4.0 embedded flash card for automotive applications including telematics, infotainment systems, and ADAS. It says the devices support High Speed Link Startup Sequence (HS-LSS) features, enabling Link Startup (M-PHY and UniPro initialization sequence) between device and host to be performed at a faster HS-G1 Rate A (1,248 Mbps) than that of conventional UFS.
The Kioxia card is available in capacities of 128, 256, and 512 GB, supports a wide temperature range, and meets AEC-Q100 Grade 2 automotive stress test standard requirements.
Samsung, SK hynix, and Micron have all developed UFS 4.0 products. SK hynix is pushing ahead with a UFS 5.0 initiative to double bandwidth to 46.4 Gbps which could mean a maximum 11,600 MBps bandwidth.
Bootnote
UFS implements a full-duplex serial LVDS interface. It supports SCSI Tagged Command Queuing. The standard is developed by and available from the JEDEC Solid State Technology Association. UFS can support higher storage capacities with 1 TB than other storage standards, such as eMMC with 256 GB. eMMC is only able to read and write as separate, asynchronous tasks due to its legacy parallel or half-duplex interface.
GigaOm analytical gurus have made large changes in their latest Primary Storage for Large Enterprises report, with only three leaders – NetApp, Pure and Infinidat – compared to six last year. Dell, Hitachi Vantara and IBM have been demoted to challenger status.
The GigaOm Radar for Primary Storage for Large Enterprises looks at 11 of the top primary storage products for large enterprises in the market and compares offerings against the capabilities (table stakes, key features, and emerging features) and non-functional requirements (business criteria) outlined in a companion Key Criteria report. Here is the Radar diagram:
(Its construction is described in a bootnote below.)
There were eight evaluated vendors in February 2023: Dell, Hitachi Vantara, HPE, IBM, Infinidat, NetApp, Pure Storage and Zadara. This year there are ten.
Storage-as-a-Service (STaaS) supplier Zadara has disappeared and is now evaluated by GigaOm in its Sonar for STaaS report. Fujitsu, Seagate and StorPool have entered.
Assessing the leadership circle rankings last year was complicated by the suppliers all being close to the outside edge of the leadership quadrant. Now we have NetApp clearly in front of Pure Storage, which is in front of Infinidat. All three have a nice balance between innovation and platform product plays, and are classed as outperformers.
There are three outliers. Newcomer Seagate is classed as an entrant based on its Exos X appliances. StorPool, another newcomer, is classed as a challenger with more of a feature play than a platform focus.
IBM is up on its own in the Maturity-Platform Play quadrant and, like Seagate, judged to be a relatively slow mover. GigaOm looked at its established FlashSystem 9500 all-flash array and Spectrum Virtualize architecture.
We would point out that IBM also sells its DS8000 high-end enterprise arrays to its mainframe customers, and these have not been considered by the report’s authors.
There is a group of four challenger vendors following the three leaders in the Innovation-Platform Play quadrant: Dell, newcomer Fujitsu with its ETERNUS and OEMed NetApp arrays, Hitachi Vantara (VSP 5000), and HPE (Alletra 9000). This group is led by Dell (PowerMax), with HPE close behind and then Hitachi Vantara, Dell again (PowerFlex), and finally Fujitsu.
The GigaOm analysts describe and measure each supplier’s offerings in detail in the full report, which is available to GigaOm subscribers.
Bootnote
The circular Radar diagram has four quarter-circle segments defined by a feature-vs-platform play horizontal axis and an innovation-vs-maturity vertical axis. Three concentric rings are overlaid on this with the outer one for new entrants, the middle one for challengers and the inner one for leaders. Vendor placements have direction of progress arrows indicating if they are a forward, fast or out-performing mover towards the center of the circle.
