Home Blog Page 414

Western Digital tag teams with WekaIO for HPC+ML workloads

Western Digital Capital has made a strategic investment in WekaIO. The amount is undisclosed.

WD aims to increase penetration of NVMe flash and object storage in enterprises and WekaIO’s Matrix software is a door-opener for high performance computing (HPC) and machine learning (ML) workloads.

WekaIO and HPE have worked together on HPC deals since at least 2017. Phil Bullinger, GM of WD’s data centre systems business unit, gave out a canned quote: “Western Digital and WekaIO have enjoyed significant customer wins in the machine learning and HPC markets over the last year, with our joint solution built on the Matrix software and Western Digital’s ActiveScale object storage system.”

Adding WD’s IntelliFlash to the mix and Western Digital Capital’s cash infusion will encourage more co-development with WekaIO.

Let’s work together

Western Digital supports WekaIO Matrix software running on its NVMe flash arrays with snapshotting to WD’s ActiveScale object storage system as a way of pushing into HPC and machine learning workloads.

WD’s ActiveScale object storage array.

Matrix is WekaIO’s performance benchmark-winning parallel filesystem software. WD’s NVMe arrays are the Tegile-originated IntelliFlash arrays.

IntelliFlash N-Series NVMe-based flash arrays.

WD’s OpenFlex composable systems connect to their components via NVMe over Fabrics, and the company also has an NVMe storage server in the UltraStar range. These could play a part in the WD-WekaIO partnership.

WD uses ActiveScale for big data and IntelliFlash for fast data. WekaIO and WD say some HPC and ML applications involve a concentration of rarely changed data such as image recognition, stock market trades, geospatial imaging and text. ActiveScale can store much of this data, and Matrix on IntelliFlash can serve it up quickly to HPC and ML compute cores.

Names, faces and storage people places

Without further ado…

Kirill Tatarinov has joined the Acronis board. He was most recently the president and CEO at Citrix Systems and is said to know effective product development strategies and possess a deep knowledge of enterprise IT. He also spent 13 years at Microsoft in several exec roles.

David Bennett becomes Axcient CEO this month, replacing Matt Nachtrab who quit in October last year.

Cohesity has hired James Warnette from Veeam to run EMEA communications – the role he had at Veeam.

NVMe over Fabrics startup Excelero has appointed high performance computing pioneer Sven Breuner as field chief technical officer. He joins Excelero from ThinkParQ, the company behind the BeeGFS software-defined parallel cluster file system and software-defined solution for scale-out storage, where he served for five years as CTO and CEO.

InifiniteIO has appointed Chris Galvin as chairman of its board. Galvin was chairman of object storage supplier Cleversafe which was bought by IBM. He previously served as the chairman and CEO of Motorola and chairman of NAVTEQ.

Chris Galvin.

Scale-out filer developer Qumulo has hired Mercer Rowe as VP of business development and Josh Harbert as VP of demand generation. Rowe, previously SVP and GM of Avaya’s cloud business,  will drum up partnerships with software application vendors, hardware vendors and public cloud companies. Harbert was with enterprise software company Apptio, where he was VP of marketing.

Scale Computing has appointed Dan Pierce as VP of strategic sales. He will run Scale’s OEM business in licensing its IP to strategic partners, and lead sales in APAC and LATAM.  Existing OEM deals include Lenovo, BCDVideo, Ingram Micro and APC by Schneider Electric.

As flagged in January, Scale has hired Marlena Fernandez as VP of marketing, running content marketing, demand generation, brand management and the go-to-market strategy. Pierce and Fernandez will work with Lenovo on worldwide go-to-market strategies and sales and partner enablement.

Marlena Fernandez

WD and Toshiba get ready for highest capacity 3D NAND

Western Digital and Toshiba have developed a 128-layer 3D NAND die with TLC (3bits/cell) cell formatting and 512Gbit capacity. 

A 128-layer die will be BiCS-5 in Toshiba flash generation naming terminology – BiCS-4 is 96-layer and BiCS-3 is 64-layer.

Adding 32 more layers should add a third more capacity – assuming the same process technology is used as in the two companies’ 96-layer 3D NAND. Accordingly, flash drives built with 128-layer dies could have a third more capacity than 96-layer drives. Alternatively they could be built to the same capacity at reduced cost. Product might appear in late 2020, with production ramping in 2021.

Wafer-thick

Senior Wells Fargo analyst Aaron Rakers has built a production model that assumes 66mm2 die size and a 7.8Gb/mm2 bit density rating. He thinks WD-Toshiba has the industry’s highest NAND density and models the suppliers at an 85 per cent wafer yield:

The die uses a Circuit-under-Array (CuA) design in which the logic circuits are built at the bottom of the chip with the data layers stacked above. Rakers says “his enables a 15 per cent smaller die size vs. non-CuA technology; we estimate a 23 per cent total die shrink vs. 96-layer.”

Partly through this freed-up space W. Digital was able to utilise four planes (vs. traditionally 2 planes) to increase performance by 2x. The die is divided into four planes or sections which be accessed independently and in parallel. Because of this it can reach a 132MB/sec program throughput level.

BL is bitline; a line of bits in a die. RC is Resistance-Capacitance.

The chart refers to bitlines and their place in the NAND world is explained here.

How 3D NAND is made conceptually by starting with 2D or planar NAND.

The 132MB/sec speed exceeds the 83MB/sec program throughput of a 110+ layer chip from Samsung which can operate at 1.2Gbit/s IO bandwidth and has a 45 microsecond read time.

WD is accessing data on the 128-layer die using 4KB pages and not the industry standard 16KB pages which limits electricity usage.

Lastly, this is a TLC die. A QLC (4bits/cell) version would have a third more capacity again – 682Gbit.

WD presented this technology last month at the International Solid State Circuits Conference in San Francisco.

Mid-range confusion crimps Dell EMC storage revenues

Dell EMC this week posted its fourth consecutive quarter of storage growth and anticipates a revenue boost from the upcoming unified mid-range array.

Revenues for Dell Technologies in the fourth fiscal 2019 quarter, ended February 1, were $23.84bn, up nine per cent y-o-y, with a loss of $287m (-$133m)

Two years of Dell Technologies revenues and profit/loss.

Storage is part of Dell’s Infrastructure Solutions Group (ISG) and brought in $4.6bn, up seven per cent y-o-y. That paled against the 14 per cent increase to $5.3bn for servers and networking. Two storage highlights: the hyperconverged VxRail appliance business has a $2bn annualised run rate and licensed bookings for VMware’s vSAN, which includes vSAN within VxRail, were up over 60 per cent y-o-y.

The storage line on the chart above shows a dip in revenues starting a year ago. There was growth using an annual compare but absolute numbers dipped overall from Q4 fy18 to Q3 fy18. Now they have accelerated.

The revenue decline shows that Dell EMC lost storage ground its multiple, overlapping products from the EMC and Dell sides of the house competed with each other and confused customers. The company is now sorting this out.

Earnings call

In the earnings call Vice Chairman Jeff Clarke said: “In storage we have now gained [market] share for three straight quarters and over 300 basis points more than anyone else in the marketplace through the first three quarters of 2018 according to IDC.”

He thinks Dell has gained share in the fourth quarter too and expects confirmation when IDC when brings out its storage supplier numbers later this month.

Clarke added: “We have plenty of work to do, but clearly we’ve stabilised the business. The investments that we’ve made in sales, capacity and coverage are yielding net new buyers, which is good to see. We continue to tune the sales compensation focus on storage, which is very important to us.”

He said Dell EMC has a leadership product in high-end storage [PowerMax] and is in good shape with unstructured data storage [Isilon.] But mid-range is a work-in-progress.

PowerTBD in mid-range

He declared: “While you’ve heard me talk about the need to simplify and consolidate the mid range, the products that we have in the mid-range today are more competitive than they were a year ago.” 

Dell EMC has settled on PowerXxx as its storage array branding scheme, with PowerMax at the top, PowerVault at the low-end and Power-name-to-be-decided (PowerTBD?) in the mid-range. Currently the Unity, XtremIO, SC and ScaleIO products occupy that mid-range

The company is to merge these products into a single platform and aims to provide upgrade and /or migration paths for all legacy lines.

In November 2018, it revealed its plan to add storage class memory to this new mid-range product, which is expected before the end of this year.

The unified mid-range product is Dell’s “plenty of work to do”. When completed, the company will give its salesforce a simplified, better positioned mid-range product set to sell to less-confused customers. The result? Profit! If all goes to plan…

Storage industry darlings Nutanix and Pure score revenue own goals

Nutanix and Pure Storage both identified revenue target issues in their most recent fiscal quarters; Pure with actual revenues and Nutanix with forecast revenues – and they have no-one to blame but themselves.

Nutanix revenues in Q2 fy2019 ended Jan 31 were $335.4m, up 17 per cent year-on-year, with a loss of $122.8m, near double the $62.6m loss a year ago. Next quarter’s revenue guidance is $290m -$300m.

This represents 1.9 per cent growth at the $295m mid-point – far less than Nutanix’s usual growth rate.

The main reason for the forecast shortfall is that Nutanix switched spending from lead generation to engineering and product development a couple of quarters ago. It thought it would enjoy continued lead-generation efficiencies. 

Nutanix quarterly revenues and profit/loss. The Q3 fy2019 number is Nutanix’ outlook

CEO Dheeraj Pandey said in the earnings call: “We recently identified some imbalances in our lead generation spending that were beginning to impact our sales pipeline. We recognised these imbalances in Q2 and have adjusted our lead generation spend accordingly. Despite these, these actions will take some time to take effect and therefore our Q3 guidance reflects the short-term impact of these imbalances.”

Nutanix co-founder, Chairman and CEO Dheeraj Pandey.

That was compounded by a shortage of sales reps in the first half of the fiscal year, CFO Duston Williams said.

A third problem was sales execution in the USA. The company aims to address this by promoting Chris Kaddaras, current head of EMEA sales, to lead both the Americas and EMEA sales organisations, Pandey said. Kaddaras has led Nutanix’s “EMEA sales engine to tremendous growth over the past 2.5 years turning around the business that was initially challenged and we expect that he will bring the same execution excellence to North America.”

Pure Storage

Pure’s Q4 fy2019 revenues were $422.2m, up 24 per  cent y-o-y, which looks heathy enough but undershoots revenues expectations of$442m. There was a loss of $25.7m. Full year revenues were $1.36bn, up 33 per cent y-o-y. 

Unfortunately Pure had guided analysts to expect $1.376bn- $1.384bn so that overshadowed the 33 per cent growth.

The inability of a contract manufacturer to build enough product was the main reason for the shortfall. Also, more customers than expected took out ES2 subscription deals instead of buying perpetual licenses. Pure wants them to do that but miscalculated the adoption rate.

Nutanix and Pure say their problems are short-term and will resume growth at historical rates. Nutanix’s sales pipeline will take two quarters to fix. Pure should resolve its product ship issues more quickly.



Western Digital wants to be the Everything Storage Company. How does its ambition stack up?

Western Digital is undertaking a huge project – to own and integrate the entire stack from silicon, glass and sand to arrays and composable systems software.

Originally a disk drive manufacturer, the company has made at least 14 acquisitions in recent years. These include the purchase of HGST in 2012, which helped consolidate the hard drive industry and took WD in to archiving. In 2015 the company got into object storage by buying Amplidata and in 2016 it bought its way into NAND manufacture and SSDs with the SanDisk takeover. In 2017 WD entered the drive array market by buying Tegile.

Let’s graphically relate the biggest enterprise storage players to each other:

WD covers more of the computing system stack than any other supplier.

In comparison to legacy disk drive competitor Seagate, WD’s business stretches further down the stack, into media and drive controller CPUs – RISC-V, and up the stack through drive arrays into system software.

WD now sells disk drives and SSDs to storage array and server-based hyperconverged infrastructure companies. It also competes with these companies.

This is a much more determined attempt to move up-stack than Seagate, which has also tried to build drive array revenues. For example, it bought Dot Hill in 2015, but that business has languished, and in 2017 it offloaded ClusterStor to Cray.

As well as competing with storage array customers WD has to manage sales at drive and systems levels. Drives mainly go to OEMs – storage array and server vendors – and to consumers. Storage systems go to enterprises and also hyperscale customers.

WD can, and does, use channels to sell to consumers, OEMs and enterprises. But hyperscalers need direct sales and custom products and the different sets of channel customers need different handling, marketing, product offers and so forth.

Channel flannel

OEMs could prefer to deal with Seagate and Toshiba for disk drives, which don’t compete with their customers for storage array systems business.

WD’s channel may also find selling storage systems to enterprises is an uphill struggle against established suppliers. WD is a relative newbie trying to enter an external drive array market with entrenched major players and facing competition from HCI suppliers and the public cloud.

Can WD win its great game? No-one else has done such a thing since IBM in its mainframe heyday. It’s a big ask – and probably no other company could do such a thing. It will be fascinating to see how Western Digital fares.

The irresistible rise of NVMe means SATA SSD days are numbered

SATA interfaces for SSDS are on their way out and the NVMe SSD interface will take over in 2020, according to Eyal Shani, Western Digital’s director, technical product marketing. Speaking at the A3 TechLive conference in London last week, he presented a chart showing price trends for 15K, 10K and 7.2K rpm disk drives

According to Shani, the 15K disk drive is already dead – that’s why the price is unchanging on the chart after 2020 – because of SSD cannibalisation.

Eyal Shani

Death-by-SSD is also affecting 10K drives and WD has withdrawn from their manufacture.

The disk drive future in enterprises and hyperscale data centres is 7.2k capacity-optimised drives. They have a substantial price advantage over SSDs and will maintain this differential through increased density.

Asked about the SAS SSD interface, Shani said it was “still here, not going overnight, but most of our effort is in NVMe.”

In other words SAS declining in popularity as well. Shani showed another chart showing data centre SSD revenues by interface from 2017 to 2022:

NVMe grows to domination, with SATA almost disappearing and SAS declining albeit more slowly. WD has even introduced a disk drive box with an NVMe interface – the D3000 in its OpenFlex composable systems line.

Infinidat girds up loins for NVMe fabrics and storage-class memory

Infinidat’s performance-boosting software upgrade, announced this week, will help the company add NVMe fabric links and storage-class memory to iInfinibox arrays.

InfiniBox 4.0.40 adds SMB support, better protection with snapshot changes, and delivers 30 per cent IOPS improvement and 22 per cent throughput increase.

Latency is reduced by 50 per cent as a result of many small software, according to our sources. For example, the last remaining kernel module – for Infiniband – was transplanted into user space. Also the company has implement a more sophisticated dispatch algorithm was.

This work is part of Vision 50, an engineering project to reduce Infinibox average global latency in the field to under 50 microseconds.

In November 2018 we reported Infinidat’s goal to add NVMe over Fabrics networking. Blocks & Files understands this latest work prepares the way. Also the company is tapping into storage-class or persistent memory – in other words it will support Optane. This will accelerate access to Infinibox data faster than NVMe-oF on its own.

Huawei bakes NVMe into OceanStor Dorado all-flash arrays

Huawei has launched the all-flash entry-level OceanStor Dorado3000 v3 array and added NVMe support to mid-range and high end all-flash Dorado storage.

The OceanStor Dorado is a classic SAN array. From high to low-end the range consists of the Dorado18000 v3, Dorado6000 v3, Dorado5000 v3 and Dorado3000 v3. With NVMe latency is reduced to as low as 300 microseconds,

NVMe support has been a long time coming. Huawei first talked about adding NVMe SSDs to Dorado arrays in 2015. The Dorado V3 all-flash storage products were launched in 2016.

The company claims OceanStor Dorado V3 mid-range and high-end all-flash storage are the industry’s first to fully support the NVMe architecture. This is unremarkable since NVMe drive support is now widespread throughout the industry.

No pricing and availability information was released at time of publication.

Western Digital adds shingles to MAMR drives

Western Digital will introduce shingling to its MAMR drives to build a path to 20TB capacity in 2020 and from there to 40TB and 100TB.

Write tracks are wider than read tracks and shingled disk drives take advantage of this to overlap write tracks as a way of cramming more read tracks on a platter. The downside is that blocks of write tracks need erasing and re-writing when new data replaces old data. This is time consuming.

Shingled Magnetic Recording (SMR) drives are slower than conventionally recorded disk drives and as such are not drop-in replacements. Rewriting block changes can be handled by the drive itself or by a host server. Western Digital prefers host-managed shingling.

The company is targeting 2020 to introduce 20TB and 20TB+ MAMR (Microwave-Assisted Magnetic Recording) disk drive technology to increase areal density beyond the current Perpendicular Magnetic Recording (PMR) technology’s limit. 

Eyal Shani, Director, Technical Product Marketing, Devices GTM, at WD, told a Technology Live briefing in London today that these drives will use shingling. He said SMR adds between 16 per cent and 25 per cent additional capacity beyond the increase due to MAMR.

At the 16 per cent level, a 20TB MAMR drive without SMR would have 17.2TB capacity. This is a small increase over current 16TB PMR drive technology. Hence the need for shingling.

Shani said cloud hyperscalers like Dropbox are the target market for such drives.

He revealed WD is developing dual-actuator technology to increase disk drive IO rates. The first actuator will handle the high 4 platters and the second looks after the lower 4 platters. Expect a statement on availability in June 2019.

WD Dual-actuator prototype.

If WD thinks it necessary to introduce shingling in its post-PMR drives Seagate and Toshiba are likely to follow the same route.

Update

March 8. A Western Digital spokesperson said: “Western Digital’s energy-assist PMR (ePMR) products (MAMR, for example) will not be limited to SMR, and we wanted to make sure that this is clear, as it may not have been clearly communicated during our presentation.

“Our plans are to continue to deliver both SMR and conventional magnetic recording (CMR) as they both offer distinctive benefits and value to the market. Western Digital also remains on schedule to ship energy-assist products during calendar 2019 – no specific date has been set.” 

Qumulo builds home-grown S3 bucket shop

Scale-out filer Qumulo is developing its own Amazon S3 bucket storage capability and has cloud archiving in mind.

Currently it uses Minio’s object storage S3 code to provide S3 storage capability. Datera does the same thing.

Molly Presley, Qumulo product marketing director, told a Technology Live briefing in London today that the company is developing its own S3 code. The first version should arrive in March 2019 and enables applications accessing the Qumulo store through NFS or SMB to read S3 object storage buckets.

Over time you’ll be able to write to that object, coming in from NFS or SMB again, and also influence the placement of the data object.

Presley said challenges include object storage being eventually consistent while file storage is instantly consistent. Another concerns the 1 to 1 mapping between files and objects. The S3 object in its AWS bucket needs to be a native S3 object.

Qumulo is also developing the ability to archive to S3 buckets, on top of the existing Archive-optimised nodes. This will see AWS S3 become a cloud tier behind the on-premises Qumulo archive box. Initially Amazon will move the objects to Glacier for lower cost archiving, but Presley said Qumulo is thinking about doing such things itself.

Joel Groen, Qumulo product manager, responding to an audience question, said multi-cloud access for file will emerge. Users will be able to move the data back and forth. Qumulo has replication today. Over time it will help develop ways to move data sets between clouds more efficiently.

Azure Blob storage support is coming sometime in 2019. Customers are running Qumulo in Google Cloud Platform but this capability has not been announced yet. Blocks & Files expects that to come this year as well.

Acronis preps HCI appliance for service providers

Is Acronis nuts? It’s developing a general hyperconverged infrastructure appliance for a market that has already matured and consolidated around the Dell EMC and Nutanix duopoly. But it just might have an edge with service providers.

The company has built a hyperconverged backup appliance and looking at a general HCI appliance, Ronan McCurtin, Acronis senior sales director, said in a briefing in London today. 

Ronan McCurtin

The thinking is that the appliance will use the Virtuozzo (KVM-based) hypervisor.

Acronis sells idata protection and cyber security products in two markets: on-premises business kit and cloud service providers. McCurtin says on-premises is Acronis’ core market, with five million or so users, while cloud is its newer and faster-growing market.

He said Acronis hasn’t really fulfilled its potential but is determined to do so in both markets, albeit with different growth rates; ”Growing 10 per cent in the on-premises market would be huge. Growing 10 per cent in the cloud market would be a disaster.”

McCurtin said Acronis was not going to go head-to-head with Nutanix or Dell EMC. But a software HCI appliance represents a new growth opportunity and Acronis thinks its channel partners can help it find a niche in a market owned Dell EMC and Nutanix. Service providers will sell services based on existing products, managed through a single Acronis facility and supported by Acronis. 

Comment

It ought to be relatively easy to graft an HCI product onto its existing service provider business, so long as it supports multi-tenancy, quality of service, and simple management.

The Acronis HCI represents an incremental upgrade for service providers.would be managed and can be supported through already existing arrangement.

An Acronis HCI would also represent an incremental extension of existing business for Acronis’ on-premises customers. 

McCurtin acknowledged the duopoly’s dominance but said it was early days in the HCI market and there was a long way to go. In other words, there is room for new entrants, despite suppliers like Maxta failing and exiting the market.

So just to be clear, Acronis is not nuts.