Apple Podcasts
Spotify
Castro
RSS Feed
Breaker
PocketCasts
YouTube Channel
The player is loading ...
December 14, 2023
It started with RAID... Storage Path to Success with Jetstor | Episode #75
In this GTwGT episode, we focus on JetStor's long and interesting journey in data storage solutions, JetStor's founder, Gene Layzarovich, discussing JetStor's expansive range of data storage solutions. Layzarovich shares his journey from immigrating with minimal resources to pioneering storage technologies that challenge industry giants like EMC. JetStor, officially trans…
In this GTwGT episode, we focus on JetStor's long and interesting journey in data storage solutions, JetStor's founder, Gene Layzarovich, discussing JetStor's expansive range of data storage solutions. Layzarovich shares his journey from immigrating with minimal resources to pioneering storage technologies that challenge industry giants like EMC. JetStor, officially transitioning from AC&NC, boasts a history of innovation, customer loyalty, and adaptability, offering solutions from small deployments to multi-petabyte setups. The conversation also delves into the evolution from RAID to NVMe and flash storage, addressing the burgeoning data demands of AI, and the importance of reliable data architecture..
Preparing for future challenges with AI data demands.
In this GTwGT episode, we focus on JetStor's long and interesting journey in data storage solutions, JetStor's founder, Gene Layzarovich, discussing JetStor's expansive range of data storage solutions. Layzarovich shares his journey from immigrating with minimal resources to pioneering storage technologies that challenge industry giants like EMC. JetStor, officially transitioning from AC&NC, boasts a history of innovation, customer loyalty, and adaptability, offering solutions from small deployments to multi-petabyte setups. The conversation also delves into the evolution from RAID to NVMe and flash storage, addressing the burgeoning data demands of AI, and the importance of reliable data architecture..
TL;DR: Gene Layzarovich, JetStor's founder, details the company's growth from its early days to its current status as a provider of diverse and reliable storage solutions. JetStor addresses the dynamic needs of data storage with a customer-centric approach, ensuring longevity and trust in the tech industry.
Main Takeaways:
- JetStor has a rich legacy of overcoming market challenges with innovative storage solutions.
- The company's product diversity caters to a broad spectrum of storage needs, from DAS to all-flash arrays.
- JetStor emphasizes the importance of data protection and advises on best practices for data redundancy.
- The company is actively adapting to the latest trends in storage technology, preparing for future challenges like AI data demands.
- JetStor is officially rebranding from AC&NC to focus on its core storage brand identity.
☑️ Support the Channel by buying a coffee? - https://ko-fi.com/gtwgt
☑️ Technology and Topics Mentioned: JetStor, AI-Driven Storage, Scalable Solutions, Data Protection, RAID, NVMe, AI, Generative AI, Storage, SANs, SSD
☑️ Web: https://jetstor.com/ ☑️ Crunchbase: https://www.crunchbase.com/organization/advanced-computer-network-corporation
☑️ Interested in being on #GTwGT? Contact via Twitter @GTwGTPodcast or visit https://www.gtwgt.com
☑️ Subscribe to YouTube: https://www.youtube.com/@GTwGTPodcast?sub_confirmation=1 ☑️ Subscribe to Spotify: https://open.spotify.com/show/5Y1Fgl4DgGpFd5Z4dHulVX
• Web - https://gtwgt.com • Twitter - https://twitter.com/GTwGTPodcast • Apple Podcasts - https://podcasts.apple.com/us/podcast/id1519439787?mt=2&ls=1
☑️ Music: https://www.bensound.com
Transcript
RAW
so there's a lot of data that will be generated and it's already generated for AI it's going to require a lot of you know storage a lot of capacity hello and welcome to episode 75 of great things with great Tech the podcast highlighting companies doing great things with great technology I'm your host an sper and in this episode we're talking to a company offering a comprehensive range of data storage solutions designed for organizations looking to optimize their data infrastructure their products blend
robust performance reliability and scalability with a focus on advanced technology and userfriendly management their Storage Solutions cater to a wide array of computing platforms and operating systems that company is jet store joined today by Jane Lear Le zich I'm going to get that right sooner or later Jan the founder of what was Advanced Computer and Network Corporation the driving force behind jet store welcome to the show Jane thank you so much thanks Anthony for having me and um you know great to be with you and uh
you know we should be should have a good show yeah and you think with me working with a lot of Russians that I should be able to to have pronounced your surname correctly off the bat but tell us give me the correct pronunciation for your surname Jean lazarovich lazarovich lazarovich I wasn't too far off that's okay you're excellent yeah there good there you good I shouldn't have doubted myself awesome all right just J just before we get in good stuff just before we get into the history of of jet store
and especially yourself you got an amazing history yourself in terms of you know your your journey towards seone founding of this company but as a reminder if you love great things with great Tech would like to feature in future episodes please click on the link on the show notes or head to jtwj t.com you can set up to date with all the episodes all 75 of them now on the major podcasting platforms Google spole Apple and Spotify um all hosted and distributed by Spotify podcast and also go to YouTube hit jtw JT podcast hit the
like subscribe button and the alert button and do all that good stuff so Jane with that out of the way let's let's talk about jet store let's talk about yourself first though because um your your story is quite remarkable in terms of just um a refugee from Russia um coming to the to the United States and making it big um so give a little bit of background about yourself and you know those early years uh before you founded this company yeah absolutely um yeah it was um a long journey um you know it was
1991 October 1991 um you know I woke up in Moscow and um you know my mom says you know there are tanks on the street and uh you know it was the first coup you know we actually were supposed to get out um you know few days later to go to immigrate to United States you know we had a refugee status and everything and then uh the airports were closed um you know we we couldn't get out so um but you know luckily the coup only lasted three days and um you know we were able it was 91 it was the first coup that they had
um you know and then we were able to get out and um actually made our way to um the states to Pittsburgh um and then that's how we end up here you know immigrated in 1991 and then um basically I was actually working in a medical field at first um actually working at the hospital uh but then you know got really interested in uh technology and then in 94 um you know we actually um found um you know for incorporate Advanced computer network um and then in 99 uh we um 1999 96 we started we actually created a a brand jet store and
we started manufacturing and um you know basically creating appliances and design and appliances and uh name jet store so so you yeah so that's interesting so how old were you when you um had to um come over to the US I came I when I was I was 18 I was 18 years okay so so not speak my English yeah there's no English you know I we did not have um you know we came literally with $100 now pockets um so we didn't have you know any financial means you know we just literally just came off the boat and um you know we
just started working from you know for ground up it sounds that way and did you have did you have an education in Russia that was based in in some sort of engineering or Computing um so I I went to a school I went to like engineering school in uh in Russia for two years and then I uh went for another two years here in in University of Pittsburgh um so but I you know I got really interested in technology um you know in computers uh my first computer was Gateway 2000 the one there was the cowpots remember those those things yeah
yeah yeah I I do it's I was I was young but I do remember yeah so um you know and then we you know we started you know obviously we started getting interested in um technology and then we heard about this famous white paper by three professors you know we raid white paper Okay um and then you know we actually um acquired raid.
com um you know and then you know we started getting into data storage and you know doing all this raid stuff which is now of absolete as you know um not as cool anymore so but that's significant I think it's significant because you know we're talking about someone who bought the raid.
com domain you know in the early days of the incident and we won't we won't give away the story about that just yet in terms of where that is today because I think that's a really cool story but we'll get into it later but I really wanted to touch on you know the fact that you found at AC and NC in in Pittsburgh you know in '93 it's only two years after you came to America it's it's pretty crazy to think you found that a company but what what were you guys doing in '93 because he mentioned in 99 that you know the jet store
product which is what people would would know you guys for now in terms of storage and whatnot but what were you doing that in that first six years right um so we were mostly uh doing because you know the company was called Advanced computer network so we were doing just general it so we actually were you know we actually had a partner actually back in Russia and then we were exporting some of the um you know just it related equipment you know software um you know we were just doing mostly just the kind of export to our
Russian partner um you know because you know there was a lot of kind of interest and there was a lot of lack of you know it infrastructure in Russia so we were supporting actually like Russian Banks and stuff like that because there that you know that industry was just coming up at that time uh um so and then we were able to help them to you know get their infrastructure you know kind of in a better shape so we we we were dealing in were you dealing in STO just like storage systems at yes we were actually dealing in storage a lot of it was
storage like the beginning you know there was just a simple you know we actually were working with a company called storage computer uh out of Nash New Hampshire um and they had um the product called rate seven um and you know it was early days and you know we started you know were doing the N9 gigabyte drives 4 gigabyte drives you know back then um so which is now just USB sticks so um yeah the early days were very very interesting you know we had a huge box you know like maybe like this small like a college size
refrigerator that was and it would hold probably about 50 gigabytes yeah so you so you were shipping these things off um to to Russia obviously you know just maybe explain a little bit about about that cuz I think I I've I've done a little bit of research and interest in you know Russian technology and how it's generally a couple of years to maybe a few years behind where the rest of the world is due to fabrication and whatnot so the idea of importing you know systems from the US meant that companies
were getting more advanced Computing and more advanced storage than what they would have got locally right correct correct yeah we were definitely on um you know at that time on the bleeding edge you know they didn't have anything that that was you know anything close to uh what we were you know shipping them and uh you know they were getting ahead you know obviously they were trying to get stay you know current with technology and you know we actually had at that time you know the bleeding technology that was that was there so
very very successful um you know until the market until the there there was a collapse in financial markets in Russia uh and then we brought over our engineer who's still with us he's been with us for 24 years oh wow okay yeah he's been working for us for 24 years and um um you know we brought him over like 98 and then we created you know brand jet store and then we started marketing jet store in the states and then pretty much our business in Russia kind of you know was shut down at that point so we closed it
out and then we started doing a lot of marketing here and then you know the rest is history yeah yeah like you started so so you started to focus locally and is was that around about that 99 time period when jet store as as a service or as a brand started correct yeah 99 um we started we created a brand you know brand jet store which you know a lot of people um you know it's not related to airplanes it's jet you know means fast and then store means storage so it's fast storage um and for 99 I think that was a pretty cool name you
know and still I think it's still Rel still pretty cool name yeah for sure yeah yeah so um um you know and then um you know we started doing we started doing a lot of marketing here you know we went to local bank you know try to get a loan and you know we didn't get any loans at all so you know we were bootstrapping it from our apartments from then we you know got small office and then we you know we had a bigger office where we start doing production and and so forth and actually our first customer uh was actually here in
Pittsburgh and they actually were ISP I don't ISP right like inter provider so now they actually are very big cloud provider here in in like a regional provider they have about 15 data centers well who's that it's called expedient oh I know them very well there you go interesting you know expedian of course that's the CL provider world my yeah yeah it used to be called Stargate Industries and uh our first sale was to starget Industries then later became expedient so yeah that's that's you know
I think yeah so probably clarify a little bit about you know we talked about the Russian area of of the raid controls and the storage but as jet store started and as you sold into this ISP what we us was it actually I guess is it was it custombuilt storage arrays or what exactly was the product what is what was jet store shipping to customers back then correct so yeah we were working with um a few oems uh one was called storage computer and one was called CMD um so in storage computers they were making um an appliances that
was already pre-built and then we had to put in the drives you know we had to burn them we had to test them it was all back skazzy the 68 pin um you know parallel scuzi and then you know we were doing like I said 4 gigabyte and 9 gigaby drives um so and then the CMD was a little bit more complex because actually had a rate controller and then we had to build get chassis from a company called cane engineering and then we put it all together with um drives and then we would ship it to a customer and those systems were bigger because
they would have maybe like 20 to 40 to 60 drives but the drives are still small so these things were you know like a mini refrigerators and um you know they were they were quite large but the capacity was really small but the speed was good at that time you know and customers were happy so yeah so what what was the differentiating factor back then when we think about um storage arrays back then probably didn't didn't quite hit the era of the big Sands at that point right I mean they for my end they they kind of started in the early
to mid 2000s where you started to see the biggest Sands dominating you know data centers so who was your competition back then and why were you winning customers back in the sort of turn of the 21st century yeah exactly so those were the units I know they call they were they still were called Death they direct touch storage so they would connect directly to the the host you know the computer the server um you know via 68 pin cables and um you know there would just be direct attached storage like you would attach your USB discs now
um so and you know we were actually competing with uh companies as like Dak digital um you know digital storage Corporation and then also EMC uh we were do you know we were competing a lot with EMC at that time it was not you know there was not Del MC but then you know we actually competing with EMC um you know and that time you know the it was very obviously we had a we were much smaller um and you know we were able to provide affordable solution to um to you know to our customers which we still do to this day and we still compete with
DMC now yeah of course see everyone kind of does that does yes everyone absolutely um so in terms of your your value ad that you had in terms of you know starting dread store back then and it's Evolution we're going to talk a little bit about you know how you've evolved into quite a quite a big range of of storage array with different componentry different types different protocols um was there a component in there that was your secret source that was giv you mentioned you've used OEM and whatnot but what was the
what was the R&D component that was that was jet store yeah so we were um we did a lot of work with GUI um you know we had like a lot of input you know how the GUI was designed we designed you know some of the guy components and some of the ideas for data protection um you know we we designed as well um so that's kind of where we came in you know was our value ad where you know we had a lot of say you know how the the layout will be and um the actual um features of the product you know we were actually contributing to the engineering
uh resources as well okay and in terms of Technology shifted we mentioned we can talk about raid.com now I think it's a pretty cool story um but you know from the point of view of how we used to think about About Storage I mean you talked about in the '90s we just talking about discs and maybe we're talking about direct attached storage and we would see a small array of discs that were on a shelf that were connected in parallel and that kind of gave you a bit of bit of uh protection a bit of parody
uh a bit of bit of storage as well you know in terms of the size of it and then raid started to come into play and everyone was talking about raid and raid was a thing for a long time and then we started to sort of not talk about raid and raid cut got a bad a bad rap and then we started talk about eraser coding and now we're kind of in this world where we almost don't talk about those things specifically it's become a bit of a black box I mean object storage has probably had a lot to play with that in
recent times um but how did how did you see that because you've obviously you've ridden that whole wave right of of of change and Trends in the storage world so um firstly tell us about raid.com that's cool and then tell us about you know your point of view on raid to eraser to object storage to where we are today with cloud and hybrid yeah absolutely um so yeah first about raid.
com so we know obviously raid was getting hot and you know everybody was looking for raid and uh we found a company that actually owned raid.com which they actually were kind of in distress so we approached them they said you know they wanted I think it was like $6,000 and you know we like well we'll just settle for 3,000 so I think mental note everyone just a mental note on the $3,000 outlay go continue the story yeah so we um you know we acquired it for I think for like $5,000 at that time and we've had it for uh I think you know we sold it uh
probably about uh 10 10 years ago so uh maybe even eight years ago for like 200,000 you know we had a lot of people I and my investment advisor told me that it was probably my best invest I you know told my investment advis you know if I ever invest you know if you ever invest like this you know that's what I want to do you know for the rest of my life so it was probably yeah we we look for great investments in those sort of unicorn Investments and I I know a lot of people did make money on domains right I think
about the guy that owned x.com when Elon bought Twitter right and he was trying to sell it for millions and millions of dollars because he owned x.com and it's just a funny thing and it just kind of kind of be lucky right in that sense but to turn $3 or $5,000 into 200,000 yeah you're not going to get a better investment almost ever like no no no crypto no no actual proper company nothing nothing nothing's going to give you a sustained return on that so it's a really cool story but to actually yeah
to hold raid.com for so long best part about it that it was not bought by competitor so it actually was bought by a company that does this the bug spray that's even better yeah I wasn't yeah I wasn't worried about it at all so you know I said hey fine just have it you know have fun with it so you sometimes yeah go ahead their their site was kbug dead.
com ah a really long site so they're like you know we wanted and on the bottle they had raid.com I mean raid raid you know that was the product yeah you know but they couldn't you know obviously it was not a trademark because it's abbreviation you know redundant R of independent disc so they couldn't prove you know they couldn't say anything about that it was their name yes um so they had to buy it and you know they offered me like 50k you know like you know I'm like no no no that's not this more is much more so then they
finally came back you know many years later and you know gave us what we were looking for and uh you know then now they have ray.com on all their bottles there you go do you know I told them send me a box up the spray and they're like no no you can buy it that's that's a shame you should have you should have had that in perpetuity actually that's what you should have said you give me a you give me a box every year do you sometimes wonder how much how much I could do you wonder how much I could
lifetime supply yeah exactly of raid of raid and raid but do suers go oh8 years ago if it was like now maybe I could have maybe sold it for a million or something you know but anyway you can't you did pretty well I wouldn't complain I think I mean yeah it's you know just like we talked you know basically the raid evolved you know is like you said object and other plat you know Technologies and everything so you know we kind of into into that you know kind of um different technology you now so it's it's it's pretty cool and so
talk about that talk about you know that the early days of raid what it meant and how it became such a thing and then it transitioned into the Eraser and and where we are today where the black box kind of exists but certain people still need to understand what's happening underneath the surface right of these storage controllers and arrays yes a lot of people you know so obviously you know now using eraser coding um you know object storage you know the you know all kinds of ZFS file system um and everything so they're
different level of protection and um you know obviously now with the the cloud adoption you know in the past five 10 years you know we've been doing a lot of hybrid Cloud Integrations with storage and it's you know it's obviously the capacity have you know have has AOL from you know 4 terabyte to to now 30 terabyte drive so we can do multi paby deployments you know in a very small footprint so the obviously the capacity the density um you know have evolved you know and now you can get you know quite
massive you know amounts of storage in a very small footprint and that's what we're doing you know and then of course the now the the sand and the the the Nas and the object those are three main um you know the components of storage platforms and that we pretty much offer all of it um you know because we have a wide like you said we have a wide variety of uh products you know that go from kind of small to medium to large and you can um You Know cover very wide um you know variety of customers because
we cover different verticals um you know we have customers all over um you know the place you know kind of a lot of educational customers um high performance Computing media entertainment video surveillance and and now we in the past five years we've been doing a lot of work with Cloud providers and backup providers right and okay yeah so in terms of the evolution so you know you started with s and sand but then you've got all flash arrays now you got nazas um and then you've also got Pure nvme arrays as well so You' you've got a
fairly wide array of of products I I think one of the interesting things if I look off on the C on the on the um vendors that I've had on the show this year specifically a few of them have been very narrow in their focus in terms of you know we're only going to release two or three different versions of our product it's going to be software we're going to do this and that um but I think what you guys have done you seem to not so much be a jack of all Trad but you sort seem be very Broad in your product
set so is that something that happened uh as a purposeful strategy or is it something that just happened just via normal evolution of the people that you were talking to in your customer set yeah I think it's kind of all of the above I mean you know we had demand for you know something sort of medium mid-range and something large and then all you know we have all some some customers that have like remote sites remote offices that have something small um so like this year we actually deploying 80 locations um for a government customer
and you know they're doing like a private Cloud at 80 different locations and need something uh fairly small but um you know it has to be very liable so we actually doing a private Cloud for them uh but the actual requirement is about 50 to 100 terabytes per site but it has to be very liable and you know they need U more you know those kind of storages as well so we have to provide this you know the smaller um deployments but then we have some customers that wanted you know half a paby two paby five paby six paby 10 pyte you know and
we have to um you know accomplish that as well so um you know the customers's demand like you said you know is is pretty much kind of you know requesting where you know we need to be um and that's where we see that you know you have to offer different solutions um you know kind of for different customers yeah and how do you balance um performance reliability and scalability with with that in mind because we we all know in this industry that there there's a challenge in offering two out of three or three out of three you know you might
lose some because that's just the law that storage platforms work under believe me I know I've had I tell this to all the storage guys I have on is that the scars that I have from Storage platforms gone wrong um you know I could write a book on um when you're looking after when you're looking after hosting platforms that are supposed to be up and the storage goes down for for all different sorts of reasons and and whatnot um how how did you know you get the loyal customer base because I feel
from what the research and looking at the customer base and the product set you had you were able to basically deliver on the promise and that's pretty rare in this industry to deliver on the reliability the performance and the scalability correct yeah we've we've been fortunate you know we've have customers that been with us for over 15 years which is very remarkable you know this industry it's huge um and you know customers been with us you know because right now we on like our eighth generation of product and um
you know they've been buying from us you know we have some customers that had like three or four generations of products so they whenever the three five year cycle runs you know then they buy new jet store and you know obviously now it's all flash all nvme like you said so you know we are doing um The Flash and nvme um and then for the you know production data and then for the backup you know we're doing hard drives which is they're still by the way alive very much alive and you know like people have been
saying dead you tape is going to be dead you know 10 years ago you know whatever tape is still alive and now people are saying that you know the hard drives are dead and you know the hard drives are still alive but you know obviously we're doing a lot of flash a lot of flash you know for production it's all Flash and this this deployment that we're doing right now you know it's all flash um and yeah go ahead and then like to you know touch up touch on your question about the you know how do we you know achieve
the the actual reliability is that you know I always tell people my biggest fear when a customer calls and something happens with array you know we ask them where's your backup and they say oh the backup is on that array and I said where's your production data oh it's on that the same array yeah so at that point you know that's my biggest fear because you know we tried to um you know the 3 to1 Rule still there it's just never going to be you know you have to have primary you have to have Dr and you have to have
archive there's just no way around you know you have to have Dr site you have to have secondary storage you have to have an ARA yeah and a lot of customers think that when they buy raid or they buy storage array or whatever you call it you know it's all going to last forever and it's all going to be one piece you never you you you got to have doneen see and it's all about architecture so we try to tell you know we consult people we provide them Solutions we tell them you know you got to have multiple um you know copies of
your data you cannot rely on one single piece of equipment and think that it's you know it's going to be there forever and um you know most of our customers recognize that and you know you know they have multiple rates and that's the only way is to to really be you know up 20 you know 247 99.
999 um you just can't design a single product you know as as reliable as two product you know sort of two two different pieces and I like that Honesty the honesty there is fantastic I think it's probably what's gotten you as far as you have right because yeah I just tell people the way it is you know you just got to have you just can't have one you know you can't you know you can't ride one horse I mean you just have two horses because the horse dies I mean what are you gonna do yeah I I bet there's some sort of Russian saying or
translation for that one the two horses dying and whatnot there's got to be something I do like I do like those Russian stories and those translations that come through there but but yeah but I think it's good though because what I do see and what we have seen especially in the last 10 years if I think about the the startups that have come through from the storage in the storage World they all promise the world they all promise ultimate you you know resiliency and up time and we never go down yeah and you know there's all these
guarantees and you know not not to not to badmouth a lot because there has been a few to their credit that have done very well I think of you know the peers of this world and and whatnot they have actually done quite well in in getting up in front of it right but there's not too many purs of this world um but what I'm saying you know in talking to different people you know I had infin that on last last episode which seemed to be company just done their thing they've gone about it very quietly they've got a great yeah they got
they've got a great they've got a great loyal base it's a great team I mean yeah they around it's a it's all about you know everybody in storage knows everybody so it's a whole when you start talking to people they all come from somewhere and you're the same right and I think jet storees the same you got this low customer base you mentioned your customers are there for 15 years that's unheard of in this world because if anything the people come different people come through so when you think
about that you've survived generational change in the IT department as well because typically no one staying there for 15 years especially people at heads of Department that are making these technical decisions and sometimes the technical decision actually becomes a financial decision as well when it comes to storage so you know for for that for you to have that longevity across the board I think is very cool um talk a little bit about are you as excited today about NV and Flash as you were back in the days
of you know the raids and you know holy crap we've got raid 1 plus zero now you we got raid five like what what excites you more like if you could sort of pick a period of time um I mean um we still you know we're still obviously very excited about the future and um you know we we seeing what's happening I mean obviously in the past five years I mean the market has just kind of just leaped you know there's just really big jumps you know on the market you know where you know at first it was kind of like
going fairly sort of not slow but you know kind of in a normal Pace but you know the the how it's all evolving now the past five years it's just really fast you know I've been fascinated by it you know because like the Flash and then it goes to nvme and then it goes to you know something else from this point you know so the from that point on you know from that standpoint it's just really really been going fast um you know how we see the technology evolving because you know as soon as the flash came in
now the nvme and then now the you know like maybe Fabric and you know this NV me over Fabric and and this and that um so there's a lot of um there's a lot of you know technology advancements but you know we we try to stay kind of on the bleeding edge but not on like we don't want to bleed out so we kind of you know we want to be established you know we don't want to be like all so new that things may not work out because obviously there's been a lot of you know those kind of technologies that came in
and kind of left um so but yeah there's a lot of things like like containers you know like the obviously the cloud the you know all those things that are just coming up you know with the hybrid Cloud adoption and uh yeah you know so much data being generated um you know in the past five years that you know which haven't generated in the past you know thousands of years so the the data explosion is just you know is just amazing and of course you know we haven't touched on but I mean that could be another episode is AI and ml you know
AI you know AI is gonna generate tons of data it's already generated there's people already is yeah you know huge farms AI farms and you know they it's very expensive to to go on a public CL with AI so there's a lot of initiative that you know lot of GPU clusters that being generated but not only doesn't you know it needs compute but it also needs a lot of storage and that's actually something yeah that that's something that probably gets forgotten actually there's a lot of people that talk about
the compute and the GPU side of that um yeah but I don't think about the storage element of it you're very you're very correct there so I mean how how how's um how do you think about that from a a leader a Visionary how do you feel about you know where jet store will be to ride that wave of AI is it is it to be able to find a new sort of happy medium between performance and scale and and and capacity to be able to store more data but needs still to be accessed at a fairly quick clip like is that kind
of where you see it going and then obviously you've got to try and mold the Technologies to get that balance as well but where do you see it uh yes OB you know we definitely seeing um you know projects coming up you know with AI and ML and autonomous vehicles and robots you know and all all those new technologies as well as even like some Medical Imaging projects as well um so there's a lot of data that will be um you know generated and it's already generated for AI and um you know the the the thing is that it needs a lot of
performance so we seen you know scalability how you know we're going to scale it to you know to en normous scale and um but the it's all going to be flash it's all going to be in vme all Flash Point yeah yeah it's all going to be flash um some of it may be sort of secondary but a lot of it is going to be on vme um and um it's going to require a lot of um um you know storage a lot of capacity um which um now it's I think it's going to create some shortages uh supply chain issues that's interesting
yeah well well yeah I mean that's what seen that yeah we've seen that already um so we actually um you know we work with Samsung and Kia on the flash side and um you know right now Samsung has supply chain issues um so we can't get any drives we're waiting until January yeah right because and that that is the problem between say spinning spinning disc and and you nvmn SSD right corre it's chips it's chips versus platters effectively right correct and they cannot make they cannot make the chips
fast enough so you know there's basically the factory you know is just just full completely booked you know with with and they can't they don't have enough factories to build it so um and it's not a resource thing is it it's it's not a resource situation it's it's it's a it's a production issue it's a production issue yeah it's a production issue they just can't the factory you know cost you know like few billion dollars and they you know even if they have the money to do it it just takes
three or five years to build it yeah so we build it or do we not so and then they're just kind of going back and forth with it uh but in the meantime there's a demand because you know obviously when Co hit the demand went down you know hyperscalers went down and then now hyperscalers are back up and then you know of course they China PR prioritize for hyperscalers and for the smaller guys you know there's always um you know supply chain issues so um the hard drives are okay but the flash is now in in in shortage do you think
there's a opportunity there to create a better type of hot tier cing scenario then where you know you and it's it's it's a traditional way of thinking right we I mean the the vnx um from EMC was was a classic where it had the the hot tearing where it would promote the hot blocks 10% whatever you set it up to over over an algorithm that ran every night into the Flash and you know in theory that if that was that was a quicker ZFS with a zil Cas sh works the same way so is that do you think a solution for the AI as well where yeah
you you need to but then I think about the way that AI just ch chunk pieces of random information from everywhere it almost be impossible but there in in that lies some some opportunity for somebody to create something that can work with that to be able to do that sort of cing on the Fly and not just use all nvme or Flash and leverage a bit of spinning disc to be able to promote but yeah there's a there's a problem for you to solve if you can solve it right I to be okay yeah there's definitely thinking about um you know how do we
tiar it sort of speak but the tiering is like you know they they don't like that word in storage anymore because there's companies now saying you know no tear you know it's all flash you know um you know somewh you know partners and competitors you know saying you don't need hard drives anymore so you just kind of stay with all flash but the problem is you just can't have flash is still sort of expensive so a lot of people still want to use utilize this but like you said the hot data is going
to be the key and if you're going to have so much hot data data that everybody's always accessing you still need you know you're going to need the fast access to that data so you're going to overrun the The Flash so you have to be very careful how you design it um you know because in the in the early days I don't remember company called Nimble storage they had tearing they were like the first ones that were doing tearing and a lot of times the problems that they have is that they have very small
cash so the customers started having at first it was fast but then when everything went into Cash overflow the cash and it will get in disc and the system got slower so all the customers are complaining you know it's too slow it's too slow and we actually replacing some of this Nim old Nimble arrays now this Cloud customer they saying it's too slow now because you know we we have a lot of demand you know High you know databases that you know need to go into flash um so it's it's a very tricky thing because you have to
figure out how much hot data you're going to put um you going how much Splash you're going to put in this array yeah that's a great conversation I think I think like you said we could have almost spent the whole time talking about yeah we could spend another episode on this it was F it was fascinating though because you know I think it made me think and hopefully the listeners think about you know yeah this this we're not only got a problem with GPU and computer there's a problem with storage here that needs solving so
hopefully jet store can do that and hey just to finish off because we only got about a minute left talk about the fact of you know officially you you know are moving officially to the jet store name um versus the name that you founded so just quickly talk about that and then we'll wrap up yeah so um we uh this year you know this this week actually we um you know decided finally decided you know that to finally rename the company to Jet store um so the acnc is pretty much going to be the you know doing business as jet store so we are jet
store officially we're jet store we're going to have everything is now going to be jet store and and um you know we are officially jet store from this point on so this is the big announcement on your show done thank you feel feel feel grateful for announcing it but hey it was a really good conversation I love chatting to you and you know hopefully we can we we can stay in touch and hopefully everyone can you know be more involved and take a look at what jet store is doing in storage because you know you've got great lineage great
history and I'm sure you're going to be successful moving forward so thank you for being on just as a final reminder if you love great things a great Tech please head to Anthony material at jtjt podcast or jtw j.com if you're not subscribed Please Subscribe and again a big thank you to Jean and Jet store this has been episode 75 of great things with great [Music] te
Join the Email List
Listen On
Support On
Top Episodes
Here are some great episodes to start with. Or, check out episodes by topic
