StorageSearch.com - leading the way to the new storage frontier since 1998

Time 2020-09-29 23:08:59

Web Name: StorageSearch.com - leading the way to the new storage frontier since 1998

WebSite: http://www.storagesearch.com

ID:64542

Keywords:

the,way,leading,

Description:

The SSD market isn't scared ofmice. Butmice aren't the only animals you can find in SSD brands. There aremany other examplesof animal brands in SSD as you can see in this collected article.Andbefore the SSD market became the most important factor in the storage marketthere were also manyanimals to befound in other types of storage too.StorageSearch.com is published by ACSL founded in1991. 1992 to 2018 all rights reserved. Editor'snote:- I currently talk to more than 600 makers of SSDs and another 100or so companies which are closely enmeshed around the SSD ecosphere. Mostof these SSD companies (but by no means all) are profiled here on the mousesite. I still learn about new SSD companies every week, includingmany in stealth mode. If you're interested in the growing big picture of the SSD market canvass - StorageSearch will help you along the way. Many SSD company CEOs read our site too - and say they value our thought leadingSSD content - even when we say something that's not always comfortable to hear. I hope you'll find it it useful too.Privacy policies. Wenever compile email lists from this web site, not for our own use nor anyoneelse's, and we never ask you to log-in to read any of our own content on thisweb site. We don't do pop-ups or pop-unders nor blocker ads and wedon't place cookies in your computer. We've been publishing on the web since1996 and these have always been the principles we adhere to.at this time ofyear Megabyte sets aside more time for reading and re-reading articles about SSDsIs there an easy way to summarizethe main developments in the SSD market of 2018 already? Inpast yearssuch articles have been publishable from around the start of November.And with a few cosmetic edits those past summaries have generally withstoodthe test of time. But the closing months of the calendar aren't asnon eventful as that suggests - and every now and again - vendors have used thenews snooze of the holidays to perform acquisitions.Here are some examples.December 2011- Apple acquired Anobit (a pioneer inDSP ECC flashcontroller IP).December 2012- Samsung acquired NVELO (a flash caching software company).December 2013- LSI agreed to be acquired by Avago Technologies. December 2014- Western Digital acquired Skyera.December 2015- Network Appliance agreed to acquire SolidFire.December 2018 - this hasn't happened yet - and is the 1 which Ireferred to in my headlineSo on the subject of writing year endsummaries of the SSD market - it's more accurate to say that - generally inrecent years - it has been possible to capture all the majortrends in such articles before all 12 months had elapsed. And looking ahead - if there isn't an acquisition story in the December 2018 news - then you shouldwatch out for a note of one having been quietly done in January 2019.So- what can we say did happen in SSD Year 2018?For the essential month by month details you can see the SSD news archive pages below.January FebruaryMarchApril MayJuneJulyAugustSeptemberOctoberNovemberDecemberBut I didn't need to revisit any those pages towrite my shortlist - of 3 things - which goes likes this.3 things in SSD year 20181 - international trade barriersIn 2018 - the free flow of memory andSSD products between the US and China which had characterized business trendssince the start of the modern era of SSDs - came to an end. Agrowing set of patent disputes, government embargoes and the background worldwide babble of trade wars (related to other countries - not justinvolving the US and China - and related to other types of products too)created market conditions for memory and SSD companies in which the mood of doing business in 2019/2020 is expected to be different to how things had workedbefore.Will this affect prices? We saw the SSD marketsurvive and thrive when the costs of memory doubled during the2017memory shortages. Set against that market robustness and expectations offalling pricesagain (the way they used to be) - then I think tariff barriers won't havea damaging effect on prices.Will trade barriers affect SSD companyprospects?Yes - but it's a mixed picture. Sme companies (whose business models leveraged the freedom of micro managinginternational differences in the costs of assembling products) will seeinitial setbacks and will have to redraw their supply chain maps.Onthe other hand - there will be new opportunities for smaller companiesarising from a trading climate in which there is more local protection fromlocked out dominant international competitors.. And companieswhich operate outside the disputed hot spots - will see opportunities toservice markets in which they would have been uncompetitive before. 2 - memory cost outlookIn the 4th quarter of 2018 - the imbalancebetween supply and demand of traditional memory products (nand flash and DRAM)which had for 2 years flipped the price per bit curve outlook upwards fromits historicallyestablished downwards direction - looked like it was getting ready to flip backagain - according to price trends being reported by leadingmarket reporters.Lookedat in isolation - it would be reasonable to expect that the memory industry'sability to ship more products in 2019 - than it could when still dealing withthe overhang of yield issues related to the structural switch away from thelast significant generation of 2D scaling - would inevitably lead towardsexpectations of lower prices.However - the impact of trade warsintroduces new variables into the cost outlook for memory systems. Theinterplay between the industry's ability to manufacture traditional memoryproducts - compared to the growing difficulties of legally shipping them intotraditional geographic markets will create a rich vein of source material forSSD bloggers in 2019.3 - memory defined processingBecausestandards organizationshave traditionally taken so long to say anything meaningful about the pastdisruptive trends in the modern era of SSDs (such as the gap between the proprietary PCIe SSD market invasion of the server market and its taming by the unifying language of NVMe) you know that when a standards ORGintervenes in this market - then something has really happened.So in2018 when SNIA said it was interested in getting involved in talks about tamingthe computational storage market - then that was another one of those things.Buta problem for the industry - which hasn't been solved yet in 2018 - is what tocall this thing. I 've touched upon it each time a news story hasappeared on these pages and theSSD jargon pageincludes some (but by no means all) the examples which have been used by variouscompanies which have implemented products.in-situ processingprocessing in memorycomputational storagein-memory computing (historically means something different - but isstarting to sound like its meaning should change)This is one of thosesubjects - like the original adoption of SSDs - about which we already know alot - except what to call it. It intersects with many other top levelviews of memoryfication architecture. You can describe it as one ofthe eleven SSDdesign symmetries - specifically - adaptive intelligence flowsymmetry . Or you can say it's a variation of the data industry's creative use ofcontrollernomics- whereby any data whose latency is beyond the here and now in this chip -whether due to media speeds or distance and the speed of light - requires alocal intelligent agent which can do useful things at our remote bidding.Butthe concept still lacks a universally agreed name. And by universal I mean a term which is usable whether the memory is an SSD or RAM array - andwhether the local intelligence is an SSD controller, FPGA, ASIC or otheracceleration engine.In my headline above - I used the words memorydefined processing . I wasn't being provocative in offering that. I don'tthink it will stick. But real concepts need acceptible words. And just asmemorydefined software was a temporary placeholder in a blog for a real marketconcept - even if the words seem as if they are in the wrong places. I thinkthat 2018 is the year that the concept of memory defined processing was cryingout for a better name than it had received so far.See also:-2018 timeline andpermalink for this articlestrategic bifurcations in SSD market history4 ways to split SSDhistory into before and after to understand nowZsolt Kerekes,editor - StorageSearch.com - October 29, 2018we're going toadd the comments to the source code laterWhenI published the first version of my popular article Charting theRise of the SSD Market in October 2004itwasn't the long rambling messily formatted historical narrative which you seetoday.In 2004 I and my readers in the SSD market had already lit thefuse for the market explosion which I predicted would inevitably follow - basedon a market adoption and why buy SSDs user value propositions whichwere completely different to anything which the industry had thought aboutbefore. I was too busy then changing all my plans to reduce myeditorial coverage of all other storage subjects to focus all my energies onthe SSD market. I was more interested in thefuture than thepast. And I wondered if I'd be able to keep up with tracking market and makinga difference to it.So in 2006 when when I was looking ahead at mypredictions about the growing number of SSD companies which no one had heard of (and which would need anew series calledthe Top SSD Companies) I realized that SSDs was a market without a reliablywritten history. How would newcomers understand it? And we neededthose newcomers to make the market work. And they as early adopters werealready facing a big barrier of challenging intertwined technical issues inprocessors, memory, storage and software which was did not have any clear centralized ownership.Due to my past experience with another growingmarket before that - SPARC systems - the quickest way to write the first draft of history is to repurposenews storiesyou've already written into a simple timeline. So in October 2006 mycharting the rise of the SSD market article was repurposed to look a bit likethe history article today. You can see that versionhere.Andafter that quick and dirty first version of SSD market history (based on what Ihad experienced first hand) I continued to be busy writing about currentand forward looking stuff. So my history article just became a dumping groundfor adding more stories each year. It also became a fertile sourcefrom which other publications extracted timelines and sometimes entire clumpsof text. Irritating for me that so few acknowledged what their source was -but hey that's the internet.interpeting SSD history on a rollingbasisSomething which every market analyst and editor does at manytimes in their working lives (if they're any good is to try and interpret fortheir readers how news stories relate to the emerging market context. That'show we get all those end of year looking back and predictive lookingfoward articles which seasonally become increasingly common as Decemberdraws nearer. I was doing exactly the same kind of thing and in someyears I would confidently assert this will be or has been the year of suchand such important emerging trend . So I added those into myhistory article too.And because I care about authenticity (and because I was simply shortof time) I just cut and pasted those present tense analyses into the growingnarrative - whether they were right or wrong - was another matter.Whichis how we get to this point here.As I'm retiring from suchactive involvement in the SSD market in 2019 (which may already have happenedby the time you read this) I was looking back at SSD history and asking myself- is there a story which I can wrap around the past 40 years of ramblinganecdotes and an interpretation which might help a modern reader to appreciatewhat happened - without having to know all the details?That was myeureka moment - when I realized - I've been doing the same kind of thing toexplain SSD controllerarchitectures, andmemory cachingratios and designsymmetries etc for many years. And technology experts who know far morethan me about what goes on inside semiconductors and systems have told me theyfound my simple splits of architecture into different sets have been handy waysof thinking about stuff. So how to split SSD history?Hereare my 4 proposed strategic splits.before and after the Modern Era of SSDsBefore the Modern Era - theenterprise SSD market wasn't a sticky market. After the Modern Era -it was. And the only thing likely to replace an enterprise SSD wasanother SSD or memory system.In my articlecelebrating thefirst decade of the Modern Era of SSDs I placed the tipping point at around2003.Before the Modern Era - even if an end user had deployedSSD accelerators to fix performance problems in one key application the balanceof probability was that for future applications the users would turn back againto legacy solutions such as faster clocked servers and storage. But thoselegacy options stopped getting faster in a series of steps:-hard drivesstopped getting faster random access times after 2000 - when RPM stopped at 15KRPM - seezero RPM SSDs killed the 20K HDDserver DRAM access times stopped getting faster in 2000 - due partly to thegrowing complexity andindeterminatelatency of memory controllers in complex multi-core processors and apreference indicated by server designers for higher capacity RAM and caps on power consumption.server processor makers had begun telling me in 2000 that future clockspeeds couldn't keep growing in the same kind of way which had been anassociated assumption of semiconductor design shrinks and Moore's Law since thestart of the microprocessor era. For processors this aspect ofMoore's Law (incrementally faster clock speeds) broke in 2000 to 2003. Iwrote about this problem in a spoof article April 1, 2002 -HotServerTechnologies Announces the 3GHz hotSPARC 9000 and more strogly later -Why Sun ShouldAcquire an SSD Company (May 2004) in the SPARC Product Directory.Ididn't realize in 2000 that clock speeds would stay at the same kind of speedlimits for the next 20 years. But the good thing about the failureof these legacy markets to deliver ever faster solutions was it forced morepeople to look at SSDs - despite the steep initial learning curves involved.before and after Fusion-ioInSeptember 2007 -Fusion-io launchedthe ioDrive - a PCIe form factor flash SSD with upto 640GB capacity and 100KIOPS performance. That event created thePCIe SSD market whichsoon after became a key focal point for innovations in enterprise SSDarchitecture and SSD systems software due to the combined efforts of the manycompetitors which entered that market.There are many strands ofSSD market history associated with Fusion-io. In this article about strategicbifurcations I'm just going to mention 2 of them.SSDs becamemust-have options in server product linesBefore 2009 -when Fusion-io began announcing a series of design wins for its ioDrives in enterprise servers made by HP, IBM and Dell - big server manufacturers didn'toffer SSD accelerators as standard options in their product ranges. Afterthese initial design wins - it becameimperativefor all server manufacturers to offer an SSD solution embedded in future products if they wanted them to look competitive. (This tipping point - if onedoes it they'll all have to do it - was predicted in my 2003 article -could enterprise SSDs becomea $10 Billion Market?)the strategic importance of SSD softwareBeforeFusion-io there wasn't an SSD software ecosystem. And the lack of automaticsoftware tools for integrating SSDs into caching and other acceleration roles meant that deploying SSDs as accelerators required expensive andskilled manual hotspot tuning.One of the spin offs from the standardization of SSDintegration in servers was that it inspired a wave ofSSD software startupswhich had never been viable before. Most of the early SSD softwarepioneers were acquiredby SSD manufacturers as the market learned that having compatible software forvirtualization and caching made it much easier to sell their SSDs. Although it's hard tobelieve it now - this was still a time when there was avacuum in thespace where the SSD software brain might have been expected to be in thevision of legacy giants in the enterprise software, processor chip and arraystorage markets - who were all taken completely by surprise by this newindustry.split 3before and after Diablo Technologies' Memory1 Diablo Technologies was apioneer in shipping SSDs and DRAM emulation products which consisted of flashmemory and controller IP integrated in a DRAM compatible DDR3 or DDR4 bus slot.Although Diablo's products failed to achieve any significant markettraction and the company was involved in a series of patent disputes and wentbankrupt - there were important market lessons to be learned from the outcome ofwhat I called in August 2015 the first salvo of SCM (StorageClass Memory) DIMM Wars.Before Diablo's Memory1 (announcement,shipping and customer benchmarks) there had been a genuine belief in the SSDindustry that new types of non volatile memory memory products shipped in DRAMcompatible bus slots would plausibly launch the next wave of performanceimprovements in the server market - in a similar way that the PCIe SSDs haddone earlier - and that somehow - the latency differences of these 2 types ofbusses and the ability to place more emulated memory in server motherboardswould make an order of magnitude difference and help to create substantial newmarkets.The immediate effect of the Memory1 announcement was to setoff a spate of competitor announcements about NVDIMM based products aiming atthe same idea but based on a variety of design approaches and memorytechnologies (not just flash).After the market failure of Memory1 myanalysis of the SCM DIMM wars phenomenon is that the architectural promise ofsuch products was fatally flawed and delivered marginal incremental benefitssometimes rather than sustainable order of magnitude improvements of the typespromised by the initial hype.The main reasons for the past failure andfuture limitations of such products are (in my view) the misconception that asingle component type of solution is in itself enough to take computingperformance to a sufficiently high next level - compared to the comparison pointof an already sophisticated SSD infrastructure.Instead - what I callthe memoryfication of enterprise computing - at the next level ofperformance will require changes in processor, memory and software architecturesworking together in new conceptual schemes. And I think suchsolutions will be agnostic with respect to form factors and may not indeed lookanything like a DIMM. Indeed they may work just as well delivered in PCIe slots(like Google's TPU) or be implemented at rack scale in fabrics.I'vediscussed these ideas in a bunch of articles including:-optimizingCPUs for use with SSDsintroducingMemory Defined SoftwareRAMin the post modernist SSD and Memoryfication Erashould we expectmore from memory? - after AFA's what's nextare we ready forinfinitely faster RAM? (what would it be worth)before and after the memory shortages in 2017Although the siliconbased semiconductor memory industry had experiencedmanyperiods of under supply and over supply capacity since its formation - theshortages of nand flash and DRAM which began in the 2nd half of 2016 and lasted through to the first half of 2018 were like no other which had happened before- in strategic impact and lasting consequences.Before the memoryshortages of 2017 - there had been a bunch of non volatile memories which (insome cases for more thana decade) had failed to ever emerge into sticky design wins and marketadoption from their seemingly never ending emerging status - dueto their lagging too many years behind successive improvements in mainstreammemory pricing per bit. During the memory shortages - the effect ofrising prices in flash and DRAM and the realization that those legacy memory roadmaps could no longer be relied on to follow the historicexpectations set by past experiences were beneficial to the competitivecomparisons with emerging competing nvms. It was as if the emerging nvms hadgot into a time machine and emerged looking 2-4 years better.After the memory shortages of 2017 - it became clear that (as I phrased it ina 2017 article) new notes had been added to the music of memory tiering. And while westill can't be sure which of these no longer emerging but emerged memories will have lasting power in the memory and SSD markets - it is clearthat a significant change had occurred and (for a forseeable bunch of years) there is no going back to the just 2 main types of memory model for future SSDdesigners and memoryfication architects. Memory designs will change. Processordesigns will change. Architecture will change to incorporate the capabilities ofnew types of memory.A much longer list of market impacts can be seenin my article -miscellaneousconsequences of the 2017 memory shortages.The causes andanalysis of market events leading up to the shortages along with contemporaneous commentaries can be seen in thenews archives ofStorageSearch.com and other historic web sites which discussed the memoryand SSD markets at that time.split 5, 6, 7, 8, 9, 10...insert your own hereI'm sure that many of you having reached thispoint may be wondering - why didn't I mention a particular thing which youbelieve is just as important as those above.I'd like to think thathaving written so many articles about SSDs - maybe I already did write aboutsome of those other things before. But if I didn't - then I'm sorry. I'm glad Ifinished this one. It's the last main article about SSDs I'll be writing aboutSSDs this side of retirement. (Or maybe ever.)Thanks for reading it.And if it's given you some ideas - then that's what I was hoping.Here'sa very old article I feel proud to have written.why buy SSDs? -pioneering use value propositionsBye for now and take care.........What's the connection betweenfast erase SSDs,storage drivesanitization and datarecovery?It was the recent news about the hurricane season which got me thinking again (as it does every year) about all those tragic stories of damage and loss.The data recovery industry is unlike otherparts of the computer market because no one sets out with the idea that theywant to include DR technology in their annual storage budgets. I thought back to some of the stories of amazing data survival and recovery whichI've reported on in my writings here. Drives being soaked in mud(that's one thing) - drives being drenched in chlorine (including the backups) that was another.we know whocan fix itAnd similar to the concept in the days when spysatellites intentionally crash landed their images on film cannisters in theocean, there have been science projects in the digital camera and telemetry agewhich were designed to bring back space data from crashed balloons in SSDs.Sometimes bad things happened to the SSD afterwards. I had forgotten about but still love this SSD data survival story about Memtech - from 2005.Re data loss - human error plays a big part in many stories of data loss (even in replicated servers) while humanerror is also one of the biggest vulnerabilities in data security.In a reflective article looking back at my conversations with the datarecovery market I note some of the things which surprised and impressed me atthe time. And having tracked the SSD data recovery market since beforeit began - I found a way to connect all the dots. ...read thearticle..20 years ago - in stealth mode byZsolt Kerekes,editor - StorageSearch.com - September 12, 2018I 've been perusing my emails of 20years ago (September 1998) which was the month before I was due tolaunch my new site StorageSearch.com and I was curiousto remind myself what kind of conversations I was having with people in therun up to doing that and - in effect - exiting stealth mode with an entirelyweb based publication which had no past print rooted brand strength.I had already been publishing a server market guide for 6 years at that time(called theSPARC Product Directory)and my company hadbeen operating a dotcom based business model funded by web advertisingsince 1996. But I wanted to make the new storage site look different and haveits own identity. Some of the new changes I planned to introduce inmy new server-agnostic storage guide were more pictures (for example thebranding images of thefirst few mice were already in hand) and the other innovation was to be the use of banner ads on the site - a design feature to which I hadoriginally said no a few years before.when dealing with such a smallcakepracticality dictates that one candle for each decade is quiteenough In those 20 year old emails I was looking atthis week a common thread was that none of my storage advertising customershad ever run banner ads to promote their enterprise products before and sotherefore many of the conversations were about finding people to designthem in time to launch the new site.I have preserved some of theseolder ads for future historians (pre modern SSD erastorage banner ads /premillenium SPARC banner ads) because they still tell us a lot about theproducts and the excitement about the key messages about them from aperspective which you don't see in dry historical narratives. (I'll add somemore of those early examples into an update of this blog so you can get an ideaof how much things have changed since then.)You can see below anexample of one of the earliest storage banner ads designed specifically to runhere.(The vendor was Dynamic Computer Products which soon after changed theirbusiness name toDataStorage Depot. This rev 1.0 ad design above - leaders inRAID, disk, tape and memory upgrades - was created in September 1998 andlike most ads underwent many design changes during the period that the adprograms ran. (Approximately 8 consecutive years with this customer - which wasnot unusual in my customer base during the 20 peak years of my web ad basedbusiness.)If it wasn't for the brave companies who were designing adsfor a new site which didn't exist yet and who weren't scared of mice then Iwouldn't have had such a rewarding job for the next 20 years and my readerswould've had to wait another 2 years to see the next wave of portals (that's what wecalled them back then) which covered the storage market as a disaggregated whole. StorageSearch.com's initial storage-wide focus(from raw storage chips, and RAID systems,backup softwareto tape libraries) itselfwould later change when in about 2007 - on seeing the scale of myearlier predictionsfor the SSD market coming true - I quietly resolved to reduce myeditorial coverage of things to do with the rotating storage market and insteadrefocus most of my energies on the rapidly changing SSD market. Which wasjust about as much as I thought I would be able to comfortably wrestle withgiven where the technology had started and where I thought it was going to go.Seealso:- SPARChistory, SSDmarket history, D2d (diskto disk backup) history..3d nand and new dimensions in SSD controller architectureresearch exploits layer based differences byZsolt Kerekes,editor - StorageSearch.com - August 28, 2018In the early years ofnand flash memoryadoption in the enterprise (for simplicity let's call this period the MLC(pre-TLC) / pre-adaptive R/W / pre-DWPD era) there wasn't thesame kind of established delineation of applicationroles for new SSDs as there is now -becauseSSDs were still carving out new reasons to be used in design wins (almost onestartup at a time) and it also happened quite often that when a new productwas announced there would be significant gaps in the datasheets compared to whatwas needed to be known to determine how the product might behave (without havingto invest large amounts of resources into benchmarking and evaluations). Tohelp my readers in this formative period I suggested several shortcutswhich could help potential integrators group such new SSDs into sets determinedby key design and architectural decisions in the new SSDs. Theseenabled anyone who thought a lot about SSD controllers to decide for themselves- yes this new one is in this set and so some its characteristics arepreordained - it's better at this, worse at at that - irrespective of whetherthere were any datasheets or benchmarks orwhether we believedthat such benchmarking had been correctly set up (which for a long time itwasn't). I know from the conversations I had with many systems designers thatthey found some of my filtering terms to be useful shortcuts - andmost of the companies which were creating these new products found it useful toanswer my questions about the internals of their designs and thinking.Butall such rules of thumb have a limited shelf life. And as I used to remindreaders in my year end articles - it's just as important to discard old ideaswhich at one time were useful as it is to adopt new ones. One of thesimplest SSD design filters which I wrote about was something I called thedifference betweenbig and smallSSD controller architecture (2011). At the heart of this was thequestion - how many memory chips has the controller been optimized for?Because if it can work with a single digit set of chips then the controllercan't employ as many clever strategies (to help reliability, performance andquality of performance) as another design which has been designed with a floorlevel of tens or hundreds of chips. It was a simple idea and it was a useful wayto look at controller designs over a 10 year period.But a paper I sawthis month made me reconsider whether that division still works. And even to askthe question - are there any small architectue SSDs left at all?Thepaper in question was -Improving3D NAND Flash Memory Lifetime by Tolerating Early Retention Loss and ProcessVariation (pdf) by Yixin Luo and Saugata Ghose (Carnegie MellonUniversity), Yu Cai (SK Hynix), Erich F. Haratsch (Seagate Technology) and Onur Mutlu (ETH Z rich) - which was presented at the SIGMETRICSconference in June 2018.This paper - among other things - suggests several new (not previouslypublicly written about) design approaches which can be adopted with tall (30layers upwards) 3D nand flash - which can leverage characterization assessmentswhich are made on a small sample of cells in a memory chip and leverage thosewith architectural support in an SSD controller to increase SSD reliability orperformance so as to make enterprise use of such memories more attractive.Oneof the ideas discussed in the above paper is the idea that the quality of cellsvaries in each layer. This in itself is not new. What is new however is thatthe authors show how the spread of reliability can be measured, modeled andharvested.The authors say - We are the first to provide detailedexperimental characterization results of layer-to-layer process variation inreal flash devices in open literature. Our results show that the raw biterror rate in the middle layer can be 6x the raw bit error rate inthe top layer. Among the many chip dependent design approachesin the paper here are 2 which I've singled out.LaVAR - Layer Variation Aware ReadingLI-RAID - Layer-Interleaved RAIDLayer Variation Aware Reading (LaVAR) - reduces processvariation by fine-tuning the read reference voltage independently for eachlayer. This idea - which properly occurs in the realm ofadaptive R/W technology (rather than big controller architecture) suggests a simplemodel which can predict a best guess threshold voltage for P/E based on a top/bottom samples extracted after endurance conditioning a small number ofblocks in the memory. On its own - this concept would be enough tomake the paper a must-read for controller designers.My gut feel isthis points the way to a middle course of run time controller design between 2well known philosophies:-the adaptive DSP ECC approach - which combines chip learnedcharacterization with heavy weight run time processing power in the target controller andthe machine learning / lifetime based characterization models proposed byNVMdurance in 2013 -which enables lightweight run time processing - based on a model whichextrapolates the best figures for a population of all memory chips - but islearned from a factory based characterization (rather than learned from thelocal chips attached to the controller). Layer-Interleaved RedundantArray of Independent Disks (LI-RAID) - improvesreliability by changing how pages are grouped under the RAID error recoverytechnique. LI-RAID uses information about layer-to-layer process variation toreduce the likelihood that the RAID recovery of a group could fail significantlyearlier during the flash lifetime than the recovery of other groups. This- to me - starts to look like another big controller architectureidea - but the authors say it can be used in an SSD with just a couple of chips.They also extend the concept to pairing the best predicted blocks in one memorychip with the worst predicted blocks in another memory chip in the same SSD.Youcan read about earlier uses of RAID thinking in SSD controller designs(including variable size planes) in myRAID systems page.Butit's clear that the interpretation of different layers in a 30 to 100 layer orso 3D memory chip starts to look a lot like big controller architecture.Previouslyit was the number of different identifiable conceptual toys in the box whichset the limits to system level design tricks. Now it's layers in the same chip..some earlier home page blogsre RATIOs in SSD architecture40 years of thinkingabout non volatile memory endurancemiscellaneousconsequences of the 2017 memory shortagesare we ready forinfinitely faster RAM? (and what would it be worth)introducingMemory Defined Software - yes seriously - these words are in the right order..Hmm... it looks like you're seriouslyinterested in SSDs. So please bookmark this page and come back again soon. ..If you could go back intime and take with you a load of memory chips and SSDs from today (along with compatible adapters so they could plug and play) how would thatchange the world? 1970s there have been 3 revolutionary disruptive influences in theelectronics and computing markets.Diskful Writes Per Daybegan as a shortcut to describe the endurance of SSDs in enterprise SSDs butwithin a couple of years it became adopted by other markets too - for industrial, military and even (surprisingly) consumer drives. It has been auseful metric but has its limitations too.what'sthe state of DWPD? ...the SSD market hasbeen the main incubator for disruptive memoryfication trends but now - as weapproach the series finale of theTop SSD Companies -I think a new list tracker is needed. the Top SSD Companies inQ2 2018genuine problem for the SCM (storage class memory) industry. How to describe performance.is it realistic to talkabout memory IOPS?The semiconductor memorybusiness has toggled between under supply and over supply since the 1970s. It was the best of times. It was the worst of times. 2017 was a year like no other in 40 yearsof SSD history. which way next for SSD?..Some of the winners andlosers from the memory shortages in 2017 were easy to spot. But there havebeen new opportunities created too.consequences of the 2017 memory shortages..Despite many revolutionarychanges in memory systems design and SSD adoption in the past decade we arestill not at the stage where it's possible to predict and plot the next decadeas merely an incremental set of refinements of what we've got now. Are wethere yet? - 40 years of thinking about SSDs..Enterprise DRAM has thesame latency now (or worse) than in 2000. The CPU-DRAM-HDD oligopolyoptimized DRAM for a different set of assumptions than we have today in thepost modern SSD era.latency lovingreasons for fading out DRAM....This may be a stupidquestion but... have you thought of supporting a RAMdisk emulation in your new flash tiered as RAM solution?whatcharacteristics could we learn?....How do we know anything?And how confident can we be when using that knowledge as the basis tomake important decisions? I'm not talking here cogito ergo sum but the rather moredown to earth matter of - how well can anyone today understand the SSD market? -and give you a reliable answer to a simple question like - what's the best wayof getting to SSD street from wherever your starting point happens to be rightCan you tell me the bestway to get to SSD Street?..The enterprise SSD story... why's the plot so complicated? and was there ever a missedopportunity in the past to simplify it?the elusive golden age of enterprise SSDs ..Why do SSD revenueforecasts by enterprise vendors so often fail to anticipate crashes in demandfrom their existing customers?meet Ken and theenterprise SSD software event horizon....I said to Rob Peglar - Theratio of processor cores to memory channels and local memory capacity is a solidpivot from which to leverage your forthcoming architecture blogs. I love ratiosas they have always provided a simple way to communicate with readers the designchoices in products which tell a lot to other experts in that field. re RATIOs in SSD architecture....Now we're seeing new trendsin pricing flash arrays which don't even pretend that you can analyze andpredict the benefits using technical models.Reliability is an importantfactor in many applications which use SSDs. But can you trust an SSD brandjust because it claims to be reliable in its ads? A couple of years ago - ifyou were a big company wanting to get into the SSD market by an acquisition orstrategic investment then a budget somewhere between $500 million and $1 billionwould have seemed like plenty.VCs in SSDs and storage.. Adaptive dynamic refresh to improve ECC and power consumption, tiered memory latencies and some otherideas.Are you ready torethink RAM?..With hundreds of patentsalready pending in this topic there's a high probability that the SSD vendorwon't give you the details. It's enough to get the general idea. SSD Market - Easy EntryRoute #1 - Buy a Company which Already Makes SSDs. (And here's a list of whobought whom.)3 Easy Ways to Enterthe SSD Market..Why can't SSD's truebelievers agree upon a single coherent vision for the future of solid statestorage? (They never did.) the SSD Heresies...The predictability andcalm, careful approach to new technology adoption in industrial SSDs was for along time regarded as a virtue compared to other brash markets.sayfarewell to reassuringly boring industrial SSDs..If you spend a lot of yourtime analyzing the performance characteristics and limitations of flash SSDs -this article will help you to easily predict the characteristics of any new SSDsyou encounter - by leveraging the knowledge you already have.flash SSD performancecharacteristics and limitations..The memory chip countceiling around which the SSD controller IP is optimized - predetermines theefficiency of achieving system-wide goals like cost, performance andreliability.size matters inSSD controller architecture..A popular fad in sellingflash SSDs is life assurance and health care claims as in - my flash SSDcontroller care scheme is 100x better (than all the rest).razzle dazzling flash SSDcell care ..These are the EditorProven cheerleaders and editorial meetings fixers of the storage andSSD industry.who's who in SSD andstorage PR?

TAGS:the way leading 

<<< Thank you for your visit >>>

Hot Websites