Home Blog Page 433

Micron ejects Intel from flash JV. What will Intel do next?

Intel Optane

Micron last week said it would exercise its right to buy out Intel’s share of IM Flash Technologies, a joint venture between the two to fabricate NAND and 3D XPoint chips. Intel says this is a pre-announcement and nothing changes. For now.

At the same time Intel has strongly indicated that it intends to continue XPoint chip manufacture if Micron exercises the buy-out clause. That would enable it to continue to supply its Optane-branded devices and drives using 3D XPoint chips.

A new fab could cost more than $10bn and take two or more years to get into production. The $1.5bn cash that Micron will pay for Intel’s stake in IM Flash is little more than a down payment. Intel has a 3D NAND foundry at Dalian in China and could use space there to build XPoint chips, or seek to expand it.

In a statement Simon Read, Intel’s PR manager for memory and storage, implied that the company may already have plans in place to ensure continuity of supply.

He said Micron “can’t officially make the call until January 1, 2019. The operation of the IMFT factory would not change until after the close of the call, which is at Intel’s discretion for up to one year. There is no near-term change to Intel’s plans in the coming quarters—this has been part of our planning for some time now.

“Intel has a number of manufacturing options available to us within the time window. We’ve been shipping a broad portfolio of Intel Optane technology products for over a year with a continually expanding product line. We will continue to lead the industry with this exciting new technology.”

Veeam trumpets sales records, eyes $1bn run rate

Data protector Veeam has announced its 41st consecutive quarter of double-digit growth, and is edging towards $1bn annual revenues.

Veeam co-founder Ratmir Timashev

The company says it made its first $1m+ deal in the UK in the third 2018 quarter and gained almost 9,000 new customers in EMEA during the quarter.

Veeam provides backup to virtualized servers, and its marketing message majors on availability. The company was founded in 2006 by President Ratmir Timashev and co-CEO Andrei Baranv. Timashev previously founded Aelita Software, a systems management tool maker for Windows Server platforms that was acquired by Quest Software in 2004.

In 2013, Insight Venture Partners made an undisclosed investment in Veeam for an undisclosed, but minority, stake and a seat on the board.  The privately-held company appears to be self-funding since then. It does not disclose its profits.

Focus and hunger

Veeam does not do anything fancy. But over the years it has retained its focus on providing comprehensive and easy to use virtual server backup. The company has constantly updated and extended the core product’s capabilities as virtual server use grew and grew.

Revenues started climbing significantly in 2008 and accelerated through 2009 and 2010. The chart below shows customer growth in 1,000s and revenue growth in $1,000s too.

Veeam annual revenue and customer growth to 2014

Progress has been meteoric. For example, in March 2013 it had 64,000 customers and protected four million virtual machines. By August that year there were more than 73,000 customers. By then Veeam had a billion dollar revenue target as a goal. In April 2016 the customer count passed 193,000 and it said it hoped to achieve a billion dollar run rate in 2018.

It’s nearly there, having outgrown, as far as we know, every other backup company, from Acronis and Asigra, through Commvault, Datto, Dell EMC, Druva, HPE, IBM,  to Rubrik, Veritas and numerous others.

Its global customer base has passed 320,000. It seems fairly certain Veeam will hit the billion dollar run rate sometime in early 2019, perhaps before, if it has a knock out fourth quarter.

Can anything stop the Veeamster’s progress? Will it become a $2bn run rate company? We think this is achievable. But to do this it must continue to riding the virtual server backup wave in the hybrid cloud as the servers become containerised and cloud-instantiated. Data management provides upselling possibilities.

Lenovo gets edgy with Scale for hyperconverged retail love-in

Lenovo has teamed up with Scale Computing to jointly sell their edge computing wares to retailers.

And they have already bagged a marquee customer. This is the Dutch multinational supermarket giant Ahold Delhaize, which has deployed their kit in 800 stores in Belgium and is planning more installations next year.

The solution is a hyperconverged take on remote office/branch office (ROBO) IT, combining Lenovo servers and switches with Scale’s HC3 Edge Platform.

And it gets the thumbs up from Ahold head of infrastructure Rolf Vanden Eynde, who says the solution “delivered the stability, support and simplicity we needed. The solution outclassed competition on total cost of ownership and simplicity.”

HC3 is Scale Computing’s hyperconverged software system. It runs apps in KVM virtual machines running under the HyperCore OS and uses Scale’s Scribe software-defined storage.

Lenovo branches out

It looks as if Lenovo’s HCI strategy is to move into edge computing through a series of tactical relationships.

The company has another edge computing partnership with the hyperconverged systems vendor Pivot3. Together they provide a surveillance system for the city of Bogotá, Colombia.

It has also partnered with HCI software vendor Maxta on a hyperconverged appliance for the Chinese market. And it has teed up HCI-flavour deals with Nutanix and DataCore.


Caringo claims object storage speed record. Will show workings next month

Caringo is touting the industry’s fastest aggregate S3 throughput performance with the latest release of its Swarm OS software.

The object storage startup says Version 10 delivers 35GB/sec read and 12.5GB/sec write but it is light on the benchmark details. The company  reveals all next month in a white paper that will publishes the specific configuration used to achieve these results.

In the meantime, we have this soundbite from CEO Tony Barbagallo to share with you: “In this specific test [by a customer] we were compared to 5 other object storage solutions and performed better than everyone. We are confident that this is the fastest performance in the industry for S3 read/write throughput. To clarify this is complete throughput operations to the object store. Swarm does not use front-side cache. You will need to ask our competitors how they perform.”

The Swarm

The Swarm OS is a complete object storage stack that runs on bare metal servers. Swarm contains all the software needed in a parallel architecture running in RAM, according to Barbagallo.

He defines parallel in the “sense that all core storage nodes can perform all requests. We store the metadata with the object and the index is held in RAM across all the nodes in the cluster so the system does not need to rely on an external database.”

The company contrasts its software with that of other object-based storage solutions which require an initial Linux install on disk and rely on caching layers and load balancers to enable performance.

Barbagallo told us that such an approach is “expensive in the sense of cost and management effort. Our competitors often use a front side caching unit of some form relying on SSDs that eventually write to the object store.”

Yes, we have an appliance

To coincide with Swarm v10, Caringo has launched a single-server appliance targeted at media and entertainment organisations – specifically post-production houses, studios, broadcasters and small enterprises.

Caringo single server


This is a scale-out device and the server hardware has dual Xeon CPUs, 96TB of raw storage capacity (60TB usable), dual 10GbitE NICs, LSI 12Gbit/s disk controller, and dedicated SSDs for boot and search. The server runs Swarm with a core object store, a web-based storage and content management portal, Elasticsearch and VMware ESXi.

Caringo says this single node starter system is a quarter of the entry-level cost of competing object storage systems. The MSRP (manufacturer’s suggested retail price) is $49,995 and includes all software and three years’ support and maintenance.

FileFly v3.0 software now supports AWS, Azure and Google Cloud, providing file tiering from Windows and NetApp. Unlike  v2.0 the update is now not dependent on Swarm, according to Barbagallo. You can, he says, tier data “to/from NetApp and Windows File Servers directly to Amazon S3, Azure or Google Cloud. This is for remote site or disaster recovery purposes. The data can be rehydrated in the cloud by running FileFly 3.0 in a virtual appliance on any cloud and retrieving the data.”

Caringo said its SwarmNFS 2.1 software delivers parallel, petabyte-scale sustained streaming of NFS to object. In recent tests on standard hardware, a single instance of SwarmNFS sustained reads of 1.6GB/sec (3PB+ per month) with no caching or spooling.

We asked Barbagallo about this and its configuration. “The NFS files were read from an existing data set within the network. They were streamed into Swarm via NFSv4 going from POSIX compliant data into the object store,” he replied. “I say ‘streamed’ because SwarmNFS is not a gateway or file system emulator. It is a file to object conversion interface which is why we focus on sustained streaming. There is no cache to fill up. If you need more throughput simply add another instance.”

“[The configuration] is the same as the S3 configuration and will be covered in the white paper.“

Roll on November.

+Comment

With a mere $10m in venture funding Caringo is a startlingly different object storage startup from competitors Cloudian, with its $173m funding, and Scality with $152m raised to date. Compared to these large-scale object storage powerhouses Caringo is like an artisan microbrewery – but it could still make great products.

And now the company has turned up the marketing heat with its claim on the fastest aggregate S3 throughput performance and introduction of the single server starter system. This may reflect the influence of Barbagallo, previously VP for products, who became Caringo’s CEO in June  2018 when co-founder Jonathan Ring stepped back to take the CTO role.

Tech Data offers pay-for-use kit

Tech Data is taking a leaf out of the public cloud playbook with a financing program that bundles data centre hardware and software and services into one monthly utility bill.

The catchily named “Tech-as-a-Service – Advanced Solutions” enables customers to have “private, dedicated hardware either on premises or in a colocation and pay for it as if they were in the public cloud”.

All very good, but there is a disconnect between what Tech Data is pitching and what is on offer on launch day – we can discern no data centre networking, servers, hyperconverged infrastructure and shared storage devices in the program. The initial TaaS-AS launch vendors are HP (Notebooks) , Lenovo (ThinkPad X series) , Lexmark (printers), Microsoft Surface (Windows 10) and Samsung (Galaxy Note8 Enterprise Edition, DeX Station and the Galaxy Book 12).  Presumably, the managed services bit of the contract is consumption-based.

You are what you eat

TaaS-AS may seem like an unusual departure for Tech Data, but it is business as usual for the giant IT distributor which generates most of its revenues from shifting tin. The company is in effect offering a 21st century rental agreement in response to market appetite for consumption-based pricing.

In its launch announcement, Tech Data points to an finding by analyst firm IDC that consumption-based procurement in data centres will eclipse traditional purchase methods by 2020, accounting for up to  40 per cent of enterprise IT infrastructure spend.

“We’ve seen a critical shift in the way organizations want to pay for and use their data center technology,” Jolea Kidd, vice president of Financial Solutions, Americas, at Tech Data, said in a statement. “Many have chosen not to adopt an absolute public cloud strategy, and our new program gives them the best of both worlds—the ability to pay and consume as if it were public cloud, but with dedicated hardware that resides in the location of their choice.”

The Tech Data channel partners have tools that show end-user customers’ consumption as well as chargebacks to specific areas. They will be able to see when the end user is close to capacity limits and work with them to get an add-on hardware sale which they might not have made in the past.

We have asked Tech Data about the points we make above and when it gets back to us we’ll get back to you. 

Also fit for public consumption

We note that Zadara offers virtual private storage (file, block and object services, and upcoming all-flash arrays with compression and deduplication), similarly paid for by usage-based subscription and located on-premises, in a colocation facility or the public loud. Zadara picked up a $25m C-round in September.

Pure Storage introduced a pay-for-use scheme in May 2019 , using its Evergreen Storage Serve (ES2). Customers get storage at up to 50 per cent less than the public cloud with terms as short as 12 months and a base commitment starting at 100 effective TBs.

Wells Fargo senior analyst Aaron Rakers says Pure differentiates this from public cloud offerings by charging on the basis of storage used, not storage provisioned.

Pure Storage goes old school with Veritas alliance

Pure Storage has struck an alliance with old school data protection vendor Veritas to protect FlashBlade and use it as a backup target and also to protect FlashArray with snapshots sent to public clouds.

Pure Storage pitches FlashBlade, its flash array for unstructured data, as a data hub suitable for combining five storage workloads that might be in separate silos: data lake, backup, data warehouse, streaming analytics and AI cluster.

It says FlashBlade can scale out and provide fast access to stored data regardless of whether the IO is sequential or random, whether the data comes in small or large blocks, and irrespective of the access type, file (NFS) or object (S3).

Veritas NetBackup protects Flashblade’s entire data hub architecture  and can also use FlashBlade as a backup target. The  results is rapid restore when necessary.

Wells Fargo senior analyst Aaron Rakers tells clients: “We believe Pure is beginning to actively push the strategy of positioning FlashBlade as a solution for Rapid Restore adoption. In addition to the announcement with Veritas, we would note that our conversations with Pure at last week’s Commvault Go conference highlighted rapid restore as an increasingly visible growth opportunity for the company going forward.”

The CloudPoint product in Veritas’ portfolio offers snapshot-based enterprise cloud backup to multiple public clouds.  The company has integrated NetBackup and CloudPoint with Pure’s primary data array, FlashArray, to achieve this.

Veritas claims to be the worldwide market share leader in enterprise data protection, with Pure’s Katie Colbert, VP Alliances, saying it is the industry’s “undisputed market share leader.”

We have no independent numbers to validate this claim but by another metric, Veritas has some serious competition. The most recent Gartner enterprise data protection magic quadrant includes Veritas in the leaders’ quadrant,  in third place behind Commvault and IBM. Veeam is in fifth place behind Dell EMC.

Gartner Enterprise Data Protection Magic Quadrant

That was in June 2016, and the next enterprise data protection MQ has been delayed because vendors have recruited key Gartner analysts. Given Veeam’s growth rates and Veritas’ restructuring initiatives Veeam could overtake Veritas in the next release of enterprise data protection market share reports.

+Comment

This an astute marketing move by Veritas to strike up an alliance with fast-growing Pure, which also has partnerships with Commvault, Rubrik and Veeam.  For example, Veeam can back up FlashArray to FlashBlade, Commvault can protect Flash array, and Rubrik can protect both FlashArray and FlashBlade.

Typically no flash array vendor has exclusive deals with a data protection supplier, and no data protection supplier has exclusive deals with a flash array vendor. We expect Veritas to sign similar deals with all the main storage flash array vendors.  We also expect Commvault, Rubrik and Veeam to replicate Veritas’ FlashBlade support in their deals with Pure Storage  This is a defensive necessity.

Intel vet takes the helm at Toshiba Memory Corp

There’s a new exec chairman at Toshiba Memory Corporation (TMC).

Step forward, Stacy Smith, who most recently worked at Intel as group president of manufacturing operations and sales. Other senior positions at the company include stints as CFO and CIO. In August last year Intel announced he was retiring, effective from the end of January 2018.

Stacy Smith

Toshiba said he will work closely with CEO Yasuo Naruke to provide global leadership and help steer TMC to the next phase of growth.

TMC is the company that owned Toshiba’s interest in the flash foundry joint venture with Western Digital. It was sold in June 2018 to a Bain Capital-led consortium for $18bn, with Toshiba retaining a 40 per cent interest.

It looks as if TMC has got itself a respected, steady and experienced exec chair. Let’s hope Smith gets on with Western Digital CEO Steve Milligan, who struggled with Toshiba’s Japanese executives, taking the joint venture to the brink while contesting the TMC sale to the Bain-led group, with court actions against the sale and Toshiba locking out WD employees.

Cohesity has yen for Japanese business

Cohesity has recruited Networld, a local IT distributor, to take it into the Japanese market.

The hyperconverged secondary storage vendor talks of an enormous opportunity in the world’s third largest economy and it can easily afford the expansion. This comes fresh on the back of a $250m funding round in July. At the time it said it would use the funds for large-scale global expansion.

Networld will recruit and train resellers to sell and support Cohesity’s products with help from Cohesity marketing and support teams in Japan. 

Cohesity says the Networld partnership has already bagged a new customer, namely SoftBank Corporation.  Softbank hired Networld to use Cohesity for VMware backup and to support previously unprotected applications without affecting production workloads.

Cohesity’s July funding round was led by SoftBank Vision Fund, a part of Softbank. This may also helped to jumpstart the new customer relationship.

Barracuda praises cloud firewall growth. But what about the other stuff in its kit bag?

Barracuda
Barracuda

Barracuda Networks wants us to know its firewall business is growing but it is more cagey about the performance of other product lines.

For its fiscal 2019 year, the backup and security vendor said it exceeded 200 net new public cloud customer deployments in each consecutive quarter.  These include deployments of Barracuda CloudGen Firewall and CloudGen WAF (Web Application Firewall).

Customer wins include Scandinavian Airlines, NTT Data Corporation, global transport company H. Essers, and Rose Metal Systems.

Barracuda says its customers consumed more than three million pay-as-you-go (PAYG) hours on Microsoft Azure and AWS combined. Which sounds great –  but we have nothing to compare this number to.

Barracuda was acquired by private equity biz Thoma Bravo about a year ago, when revenues were an annual run rate of $400m. Now the good news is that its cloud firewall product business is growing. But what about the rest of the portfolio?  Barracuda provides email security and data protection too but is silent about their performance, praising only the cloud firewall products.

BJ Jenkins, Barracuda CEO, said in a statement: “We’ve made a deep investment in public cloud security solutions that fit cloud-connected customers’ evolving needs and are optimistic about the continued growth in our network and application security business.”

Could this presage a re-shaping of the business to focus more on network security?

Together we can build a better SSD controller – CNEX bags $23m funding

CNEX Labs, a startup that enables big cloud computing vendors to customise SSDs, has picked up $23m in Series D funding.

Typically, SSD controllers are one size fits all – apart from relatively crude differentiation into enterprise and consumer SSDs. And typically they are supplied by SSD manufacturers such as Micron, Western Digital and Samsung.

Using CNEX controller technology, hyperscale vendors can build their own controllers and squeeze substantially better price/performance, according to CNEX. The company claims customers gain higher drive throughput, lower power consumption, lower and predictable latency and better cost control.

Or as CNEX Labs CEO Alan Armstrong puts it: “CNEX Labs technology relieves customers from the mercy of a commoditised market and puts them back in control of their own destinies.”

So how does it work?

We first came across CNEX in March 2018, in connection with Microsoft’s Project Denali. The idea is to make solid state drives dumber by executing upper-level controller software functions, such as wear-levelling and garbage collection, in a host server or in an ASIC/FPGA/SoC device front-ending the drives.

That’s where CNEX comes in, at the ASIC/FPGA/SoC level. The pitch is cheaper drives and operations performed more efficiently at systems level instead of drive level.

The company offers its own controllers and turnkey SSD design assistance. Its technology includes a programmable interface to NAND flash memory, allowing the same controller to work with multiple types of NAND.

It has flexible drive- or host-based Flash Translation Layer (FTL) control supporting optimisation for different workloads. Proprietary hardware acceleration supports key functions that typically run today on slower firmware.

Talking about money

CNEX began life in 2013 with $2.5m seed funding. It picked up a $21.2m A-round in 2014, $25m B-round in 2015 and $17m C-round last year. Today, total funding is around $88m. The latest infusion is led by early investor Dell Technologies Capital, which also led CNEX’s A-round.

D round funders are a roll-call of the great, the good and the filthy rich. They include M12, Microsoft’s venture fund, which led CNEX’s Series C round, major semiconductor foundries, and large storage and networking semiconductor companies. VC backers included Sierra Ventures, Walden Venture Investments, and Brightstone Venture Capital.

Let’s hope we see a product next year and customer validation. If this technology idea flies, we think CNEX Labs is likely to be acquired by an SSD vendor. Micron, for example, already offers a  degree of SSD customisation and looks to be a good fit with CNEX Labs’ technology.

Kaminario CEO completes Epic Israel bikeathon

Once upon a time Kaminario CEO Dani Golan streaked above the Israel landscape in fighter jets. Now he uses self-propelled wheeled transport and has just completed an Epic Israel mountain bike marathon, a three-day stage race.

Dani Golan looking composed while punching the pedals

Day one was 98km, day two 105km and day three 62km with between 1500m and 2400m climbing each day, and a time limit for completion. There were more than 130 competitors, men and women, from 25 countries and 84 UCI riders, competing in various categories.

After three days in the saddle, Golan is possibly the fittest and the most exhausted storage company CEO in the business.

What’s next for the challenge junkie? He is eyeing an eight-day event in South Africa next year …

Are there any other fitness freaking storage CEOs and execs out there? Let us know!

Snowflake is backed to the tune of $923m. Why are VCs investing so heavily?

Snowflake Computing has taken in $450m in a second 2018 funding round to advance its data warehouse in the cloud business. Total funding now stands at $923m.

I interviewed  CEO Bob Muglia in my news story for our sister title The Register about the funding roundYou can see what he has to say about what Snowflake will do with all the money

Here we explore  a little deeper why the VCs are so keen to  stump up some much cash.

Snowflake is competing against two whales: AWS with RedShift and Microsoft Azure with SQL Data Warehouse.  These are both cloud-generation data warehouses and not on-premises product pigs with cloud lipstick rubbed on their faces.

AWS and Azure have huge resources at their disposal and Snowflake has to meet and beat their features, out-perform them with better code, and out-market and out-price them as well, while selling cloud style with on-demand / subscription pricing and not having customers buy perpetual licenses.

This does not seem like a promising battleground for any startup, but, so far, the VCs see Snowflake is doing well, tripling its customer count in 12 months to pass the 1,000 mark, for example. They now value the company at $3.9bn but it needs to be bigger, to scale more.

In other words,  the company has a huge need for dollars and could well have a prodigious cash burn rate, which is sustainable as long as the company’s business keeps on growing such that reaching critical mass and break-even can be forecast with a, hopefully, massive exit coming the VCs’ way.

Snowflake wants to recruit more and more customers, blizzards of them, to its multi-cloud data warehouse, such that it ushers in a cloud data warehouse ice age, with a Snowflake ice-cap dominating the market, and burying its competition.

And why are the VCs bankrolling such grand designs?

Blitzscaling

We think the answer could be blitzscaling. Jedidiah Yueh’s “Disrupt or Die” book on Silicon Valley startups has a chapter that looks at blitzscaling; Execution, page 187.

Yueh founded and ran Avamar, which was bought by EMC for $165m in 2006, and also funded and is the executive chairman of database virtualizer Delphix.

He says the blitzscaling idea, according to Reid Hoffman, a Greylock partner and chairman and founder of LinkedIn, is “the science and art of rapidly building out a company to serve a large and usually global market, with the goal of becoming the first mover at scale.”

In a section called Pay to Grow ( page 228), Yueh says a VC is driven by getting a larger payoff when a funded company is acquired or goes public. So VCs have a force-feeding tendency; blitzfeeding.

Yueh suggests VC-funded blitzscaling can lead to undisciplined financial management and poor capital efficiency.

He gives an example of two exit scenarios by a startup company that nets one billion dollars. Scenario A has the company funded by the VC to the tune of $20m, and the VC gets a $200m payoff –  20 per cent of the $1b. Not bad.

In scenario B the VC invests $100m to build the $1bn exit; “netting the VCs 51 per cent or $501m.” Yueh says VCs want scenario B. The more money they put in, the larger the percentage of the company they own and the larger their payoff.

He says; “The highest returns for VCs come from the early rounds (Series Seed, A and B) but the safest places to put large amounts to work are big, later rounds, especially when the VCs already have clear view on how well a company is performing.”

Which brings us back to Snowflake and its enormous 2018 funding rounds.