SPECIAL REPORT: All-Flash Array Market Growing But Is it Lucrative Business?
Not sure, up to now.This is a Press Release edited by StorageNewsletter.com on June 2, 2016 at 3:12 pm
Who is profitable in all-flash arrays (AFAs)? We cannot answer precisely to this question but what is certain is that the only three independent public companies in this field are losing a lot of money and all of them saw their revenue decreasing quarterly for the last three month of their financial results.
Nimble Storage, Pure Storage and Violin Memory recently announced their figures for the period ended April 30, 2016 for all of them. The three firms are in AFAs, more recently for Nimble.
Note that generally the first months of the calendar year are not generally the best ones to sign customers for storage subsystems. But the results are alarming.
Nimble reported quarterly revenue up 21% Y/Y at $86.4 million but down 4% Q/Q, with huge loss ($42.7 million) continuing to increase.
For the first time, Pure recorded a quarterly decline in revenue, down 7% Q/Q (but up 89% Y/Y).
Violin is in financial big trouble, continuing to fall with quarterly sales for first time under $10 million, 11% lower sequentially and 20% yearly, with $22 million net loss.
|Nimble Storage||Pure Storage||Violin Memory|
|Total of financial rounds*||99||470||186|
|Amount received for IPO*||168||425||162|
|Net income (loss)*||(43)||(63)||(22)|
|Revenue expected next quarter*||93-96||153-157||11-13|
|Number of new customers during quarter||580||300||7|
|Total customer base||8,160||1,950||6,000|
* In $ million
Among other start-ups, Kaminario saw revenue tripling over the last five quarters and hopes to hit $100 million in revenue in next couple of years. Tegile and Tintri are also growing very fast
AFAs were not invented by big storage companies but introduced by start-ups especially Pure XtremIO, acquired by EMC, Nimble and Violin.
Dell, EMC, HDS, HPE, Huawei, IBM, NetApp, Oracle rarely reveals their operating profit in AFAs. What’s sure is that all these huge firms saw their revenue on their traditional storage business drastically decreasing and not compensated by AFAs’ sales.
There is an intense competition now in AFAs with 71 actors in this field (see the list below) with new entrants arriving regularly, especially array vendors moving from HDDs to SSDs, and as usual in a new promising businesses, their goal is more to try to get market share than profit. Consequently it’s a fierce price war.
AFA’s largely leader in a global market of approximately $2.5 billion in 2015 in front of IBM, HPE and Pure and NetApp, EMC said to have recorded XtremIO revenue of $1 billion (without EMC DSSD D5, all-flash VNX, and Unity) in 2015 with now thousand customers. Run rate for booking was $2 billion in 4CQ15. Demand for XtremIO plus the VMAX all flash were up more than 100% for its more recent quarter.
NetApp all-flash FAS, EF and SolidFire product and services grew 35% from 3FQ16 to 3FQ16 for an annualized net revenue run rate of $700 million.
The only official IBM’s comment on storage on its most recent financial quarter: “(it) continues to reflect weakness in the overall storage market. We are well positioned to capture the strategic growth areas of flash, software-defined-storage and object storage – which together grew in the quarter.”
There are two many actors in AFAs. The biggest will survive, the smallest will disappear.
All Companies in All-Flash Subsystems by Countries
71, up to now
Beijing Memblaze Technology
Shannon Systems (acquired by Silicon Motion Technology)
The Netherlands (1)
EchoStream Innovative Solutions
eRacks Open Source Systems
Hitachi Data Systems
IBM/Texas Memory Systems
JBOD Storage Solutions
Nimbus Data Systems
One Stop Systems
Sihua Technologies (also in China)
Silicon Graphics International/Starboard Storage Systems
Violin Memory/GridIron Systems
Here is a comment by Thomas Z. Isakovich, CEO and founder, Nimbus Data, also in all-flash arrays: "A key metric to also consider: gross margin. Pure and Nimble still earn 65%+ gross margin, yet as cloud and SDS gain hold, margins will go down, putting even more pressure on their financial performance. The conventional storage go-to-market business model is broken; it’s far too expensive and fundamentally unprofitable."
Another one from Mohammad Moshref: "I was wondering if Hyper-Converged implementation has affected the slow down? Hyper converged booked $895 million on 2015, and looks like it will do about $1.3-1.5 billion this year? As your list indicates there are still too many players, and not enough customers to go around, more consolidation will take place, it have to."