Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Businesses Data Storage Facebook The Almighty Buck Hardware

Should a Teenage Entrepreneur Sell Out To Facebook? 358

colinneagle writes "Andrew Mayhall is 19 years old and is running a server company, called Evtron, whose product has reportedly set the world record for data density (4.6 petabytes per server rack) and has begun attracting attention from investors. One of those interested parties is reportedly Facebook, with whom the young CEO claims to have had casual discussions about a potential acquisition/hire agreement (Facebook did not respond to a request for comment on the talks). He says the opportunity to speak with Facebook was simply one he couldn't pass up, and seems more impassioned by entrepreneurship. He speaks often of building his company into an EMC or NetApp, and could very well compete with them soon. But if an offer from Facebook ever comes, should he accept, or try to build something on his own?"
This discussion has been archived. No new comments can be posted.

Should a Teenage Entrepreneur Sell Out To Facebook?

Comments Filter:
  • Retire at 20 (Score:5, Insightful)

    by smprather ( 941570 ) on Tuesday November 06, 2012 @01:23PM (#41895453)
    Sell for $5mil and be done with earning a living. Relax and enjoy the rest of your life.
    • Re: (Score:2, Insightful)

      Five million bucks won't keep you for life unless your very prudent.

      • Re:Retire at 20 (Score:5, Insightful)

        by X0563511 ( 793323 ) on Tuesday November 06, 2012 @01:36PM (#41895697) Homepage Journal

        What planet you living on? Most people don't even make half that through their whole lives.

        • Re: (Score:3, Insightful)

          by js33 ( 1077193 )

          What planet you living on? Most people don't even make half that through their whole lives.

          You still have to be very prudent with it if it's going to last you your whole life. Most people who win the lottery and take a lump sum are not prudent with it, and they end up broke in a few years. Just like some of the high-paid sports stars when they enter middle age.

          • Re:Retire at 20 (Score:5, Insightful)

            by cayenne8 ( 626475 ) on Tuesday November 06, 2012 @01:54PM (#41895917) Homepage Journal
            Yeah...we call them stupid.

            Get $5Million clear....invest it in a non-agressive manner, which could still get you like %5 interest annually.

            As long as you don't buy leer jets...you could live easily on about $250K a year pretty readily.....

            I know I could.

            • Re: (Score:3, Funny)

              by Jawnn ( 445279 )

              As long as you don't buy leer jets....

              Are those the ones with the bigger windows? Are they more expensive than Citations or Gulfstreams or... Learjets?

            • by dtmos ( 447842 ) *

              Is this [modellversium.de] a leer jet? Is this [luxuryproperty.com] a Learjet?

            • Re:Retire at 20 (Score:5, Insightful)

              by WhiplashII ( 542766 ) on Tuesday November 06, 2012 @02:17PM (#41896247) Homepage Journal

              If he started a company in his teens, I doubt he wants to relax the rest of his life...

              He should sell. The reason small companies sell to large companies is to decrease concentration of risk for their owners. He, as an owner of a small company, needs to deconcentrate his risk. He will have another company he wants to work on - he probably already has some ideas. It is far easier to do that after selling your first company, and far harder to do that after missing the only opportunity to sell.

              The simple fact is a cash out event gives you great options for your future work. If you don't sell, there is a high risk that the company will fail before you can personally cash out - this is true of all start-ups.

          • It's really not that hard.

            Put $5M in the bank earning 2%. Live reasonably well off the $100k/year for the rest of your life.

            Or bonds earning 5%. Live very nicely off the $250k/year for the rest of your life.

            • You're eating into your principle doing that. Because there's a little thing called in-fla-tion. At 2% inflation, that 100k a year at the end of 50 years is worth only what about 37k a year is worth now. Not quite so much any more. And that's a very conservative estimate of inflation.

              • by h4rr4r ( 612664 )

                We all know no one could ever live on 37k a year.

                In 50 years the house is paid for, college loans long gone, multiple cars are owned outright. Not a huge challenge living off that kind of money in that case.

              • You have a valid point, but it's not eating into your principal. It's devaluing your future earnings.

                But really, he could go get a regular job, and add that onto the principal or add it to the interest income. Still set for life.

              • by obarel ( 670863 )

                It's really important not to eat into your principal, because you want to make sure that you still have $5 million in the bank when you're dead.

      • Re:Retire at 20 (Score:5, Informative)

        by WilliamGeorge ( 816305 ) on Tuesday November 06, 2012 @01:38PM (#41895725)

        Assuming you lost nearly half of that to taxes, 2.5 million invested at even a low 2% (in CDs, for example) return is $50k per year. I bet you could do much better than that if you invest wisely, and even if you didn't $50k is enough to live comfortably on if you don't have any debts. There would have been plenty in there to buy a home, nice car(s), etc - at that point simply living off the interest is certainly doable.

      • Re: (Score:3, Informative)

        77 years at $65,000/year. You can buy a house and a car in cash and avoid paying a huge chunk in interest. You may be able to avoid paying taxes on it since it's an asset changing hands, not sure, kind of doubt it (the business liquidating would work this way, but a CEO getting cash monies is going to pay income at around 30% just like everyone else). You could ask them to payroll you at and disburse a portion of the money into 401(k) at the maximum contribution per year until the account balance is enou
        • 65k will be nothing in 77 years, at the rate of inflation it'll be below the poverty line in 77 years
      • 4% interest on 5 mil would be $200,000 a year. Ya I could live on that for awhile
    • Re:Retire at 20 (Score:4, Insightful)

      by Baloroth ( 2370816 ) on Tuesday November 06, 2012 @01:35PM (#41895673)

      That's what I'd do. Calculate first to make sure that after taxes what I got was enough to live moderately comfortably for the rest of my life, with some margin for error, then sell for that if I could. Too much risk when building a business to assume it'll work, especially given the size of the players in the field. If you can't sell for enough, well then try to build the company up.

      The fact that he has a server business at 19 says he is pretty motivated, though, which means he probably won't sell. He'll try to build it to get more money or a constant stream, motivated types usually do. Might work out, might not.

      • The problem is they aren't necessarily going to pay cash. They like to give you a bunch of stock options, restricted so that you'll have to work for them for 5 years for the stock to vest.

        Meanwhile Facebook stock is going down, down, down.

    • by ron_ivi ( 607351 )
      But hire a lawyer experinced in small business sales to make sure heJ:

      doesn't get paid in $5-mil in non-voting stocks with bizzare restrictions on when/how he can sell it that might be worth nothing before he can cash out.

      doesn't sell something worth $50-mil for $5-mil. It'd be pretty sad if tricky legalese makes his share worth nothing, and he needs to try a Paul Ceglia like lawsuit to get his facebook stock cashed out.

  • by Andy Prough ( 2730467 ) on Tuesday November 06, 2012 @01:24PM (#41895459)
    I asked my teenage daughter (the demographic that drives all technology spending) if she or her friends use Facebook anymore. She said almost never - they use Instagram to share pictures, and some other services I can't remember right this second. Is Facebook in danger of falling off the MySpace cliff?
  • No (Score:5, Insightful)

    by Anonymous Coward on Tuesday November 06, 2012 @01:24PM (#41895461)

    Don't sell...license.

    • by xxxJonBoyxxx ( 565205 ) on Tuesday November 06, 2012 @01:50PM (#41895877)

      >> Don't sell...license.

      OK, some perspective from the "buyer" side. When I encounter a company with technology I want, I almost always attempt to LICENSE the technology, not buy the company. Why? Because I can usually get what I want out of the company cheaper and with less hassle than if I bought the company. (Think retaining/motivating employees, etc. too.) Plus, if I become a significant stream of revenue to that company, I can often dictate what the company does with the majority of its development resources if I'm needy enough; in effect, I can get that companies' other customers subsidize my desires.

    • A bird in the hand is worth two in the bush... SELL!
  • He should sell out (Score:4, Interesting)

    by Ukab the Great ( 87152 ) on Tuesday November 06, 2012 @01:24PM (#41895467)

    And have plenty of beer money for when he goes to college.

  • by multicoregeneral ( 2618207 ) on Tuesday November 06, 2012 @01:25PM (#41895469) Homepage
    If one hypothetical things happens, should you do some other hypothetical thing? Sure. Why not.
  • by xxxJonBoyxxx ( 565205 ) on Tuesday November 06, 2012 @01:27PM (#41895497)

    >> if an offer from Facebook ever comes, should he accept?

    Yes, but...

    >> One of those interested parties is reportedly Facebook, with whom the young CEO claims to have had casual discussions about a potential acquisition/hire agreement

    ...I wouldn't count on that now. Yeesh.

  • by DreadfulGrape ( 398188 ) on Tuesday November 06, 2012 @01:27PM (#41895507)

    Sell, absolutely. Then take the money and build something even bigger.

    • Re:Sell! (Score:5, Insightful)

      by bhlowe ( 1803290 ) on Tuesday November 06, 2012 @01:31PM (#41895601)
      Yeah, companies go up and down quickly in tech. If you have an opportunity to get your technology into the hands of pros who will pay you big money.. its a no brainer.. With cash in the bank, you'll be able to start a new business, relax about meeting payroll, and take time off to finish college or travel abroad.
    • Personally, I would put enough away to live on and use the free time that gives you to make something bigger.
      Not turn around and spend all your earnings on a risky venture.

    • there is an implicit assumption here that he can build something bigger.

      Lots of people believe that they can be serial successes, but lots of people fail to do anything special after their first good idea.

      Kinda like the many many one-hit-wonders in the music industry.

      still - if he gets a good offer here, then selling sounds smart. It's probably easy enough for Facebook to just reproduce his work, open source it and maim his business.

  • Yes. Money = food.

    No. OSS = nirvana.

  • Yes. (Score:5, Funny)

    by cpotoso ( 606303 ) on Tuesday November 06, 2012 @01:27PM (#41895527) Journal
    The answer is yes. There is nothing magical about putting off the shelf items densely packed. Once somebody sees it, it is not hard to reproduce. So, yes, sell.
    • And when it comes to storage, it is the management software, not the hardware that is tricky.

      30PB is a lot of data... Managing it is a big headache.

      Tracking down how much is allocated and actually used, who owns what and how it is zoned on the fabric takes sophisticated software.

      I work for a DC hosting company and when we take over operations from a non IT company and start accounting for storage, the amount of space that is mis-allocated or just plain orphaned runs about 20%?

      Or more... It is amazing how mu

  • by dmomo ( 256005 ) on Tuesday November 06, 2012 @01:29PM (#41895561)

    If he's ambitious and hungry for innovation, could he not use the money from Facebook to bootstrap something bigger? If it's just dumb luck, perhaps if his product is of any real value, and he doesn't sell, the big players might just jump in and roll their own faster, cheaper solution. Competing with them will be a different sort of job. Moving forward on his own may involve more work with Lawyers than with Technology and Innovation. So, the answer depends what kind of person he is. Some people are driven by the business end of things. Others want to innovate. Rarely is someone good at both.

  • by OzPeter ( 195038 ) on Tuesday November 06, 2012 @01:29PM (#41895567)

    as to what the best business strategy should be between two 3rd party companies?
     
    You might as well ask what I think your neighbor should give their kids for Christmas. Go ahead .. use that as a Slashdot poll.

    Obviously we have not yet reached the bottom of the Slashdot story barrel.

  • so maybe you take down $5 or $10 million and you miss the magic second in which it could have been $50 million.

    I can get through the weekend just fine on $5 million... and another 60 years.

    the smart investor knows it's worse odds than a lightning strike to go for the last penny. even if you never have another good idea in your life, let alone one you can commercialize, have a deals lawyer in the room with you and work out a series of code taps under the table, and shake hands even if there's a little taste

  • by jeffmeden ( 135043 ) on Tuesday November 06, 2012 @01:33PM (#41895633) Homepage Journal

    Let's say this is the best case scenario, that you actually figured out something in your garage, with little to no experience in other high density storage applications, that EMC, NetApp, and the other major players simply couldn't come up with despite having hundreds (thousands?) of very talented engineers. If you manage to get a patent on it (you don't have one yet... interesting) then just license the rights nonexclusively. But then again, hopefully you have at least one lawyer around to give you the same advice.

    Worst case scenario, is you just "invented" something that is already patented (this is highly likely) and your visibility just isn't high enough to have the hellhounds attack yet. In this case, again, a lawyer is your friend.

    • Exactly.

      I'm sure EMC, NetApp, and the other major players like the guys who build the disk drives themselves have stuff in their lab that's 10x the capacity of what they sell. In the lab it's still proving itself, and some of that lab stuff isn't going to work out. They'll only take the stuff out of their lab when competitive pressure forces them to.

      I read that Google is the world's 7th largest computer server manufacturer, but they don't' sell *any* servers, they use all they make themselves. Googl
  • I'd love to see a creative young genius like Mayhall become an independent force in the technology sector. But 21st-century capitalism just has no room for small independents. They can't get the money to grow. The capital markets like big names and known brands.

    Go to local mall. Thirty years ago it would have been all mom+pop operations. Now it's all big-branded chains and franchises. It sucks, but that's the way it is. Sell out while you can!

  • by mveloso ( 325617 ) on Tuesday November 06, 2012 @01:35PM (#41895675)

    Building cool hardware is great. Selling cool hardware is totally different.

    If someone wants to buy you at a point before you sell, do it. The summary says you'll compete with EMC or NetApp. You won't. You're able to do what you're doing because you have time to think about the product. Someone else in the field can look at what you're doing and figure it out quickly. Someone like the people at backblaze.

    Can you offer 24x7 support? How is your manageability and maintenance? Recovery? How are you going to make the thing? Those are basic questions. Are you going to sell direct or via channels? blah blah blah.

    OTOH, if you get eaten by facebook you get to help them design and build their systems, which is great if that's what you want to do. The thing is, your story is what's getting you the PR, not your product. Leverage off that PR as much as you can, since it's all you've got right now.

    • Can you offer 24x7 support? How is your manageability and maintenance? Recovery? How are you going to make the thing? Those are basic questions. Are you going to sell direct or via channels? blah blah blah.

      This is the first thing I thought of, too. The only thing killing storage density today is serviceability. Everyone wants a box that offers per-disk replacement. This is a different model entirely; most likely they would just write off the failed disks and at some point as the chassis ages just throw the whole thing away. Going to work for Facebook to make this work in their model is a great path. Trying to do it on your own, to sell to the masses and put up with the constant headaches of support, is g

  • If he wants to be an innovator then I'd say avoid selling out to a large corporation. If he doesn't sell out, but makes a deal to service Facebook, then his company could outlive them and he could gain income from servicing other companies. More businesses will benefit from his innovations and the world will be a better place. If he wants to be a venture capitalist entrepreneur, then sell out to Facebook, take the money and move onto different things. But keep in mind that when Facebook's end comes it w
  • by retroworks ( 652802 ) on Tuesday November 06, 2012 @01:36PM (#41895691) Homepage Journal
    If Facebook was actually serious, he would have gotten a Non-Disclosure Agreement by now, as FB has way too much exposure to insider trading shenanigan penalties. I assume if he's seeking advice on slashdot, he has either already blown it or is prematurely bragging to a girlfriend.
  • Do you want to be a bitch to facebook, or do you want to own facebook.
    • Do you want to be a bitch to facebook

      It depends entirely on how many zeroes they put on that cheque.

      For high anything over $50 mil ... I'd fuck Zuckerburg at the Super Bowl half time show. Anything over $100 mil, and he can bring a friend. ;-)

  • Has he developed a service or a feature?

    The former has growth potential, is more valuable in the long-term and easier to monetize. The latter is valuable from a licensing / R&D perspective. Given his probable lack of resources and experience, it would probably be easier to sell his work to the highest bidder than spend years and tons of money developing it into something else that might equal the buyout value after expenses.

  • Comment removed based on user account deletion
    • by hawguy ( 1600213 )

      ...unless you know your product won't/can't become and isn't worth more than they're paying. 5 million now is not as good as 500 million later; interest rates are low (lol). Possible exceptions are when you just want to get out, or need cash now in order to start a bigger and better venture.

      Conversely, unless you know your product will become worth more than they are paying, sell out.

      If Netapp or EMC thinks there's money to be made in very high density storage systems, I'm sure they have an engineer or two that can figure out how to squeeze 120 disks into a 4U rack (or they have the money to hire someone [wikipedia.org] who can) - and it'll work with the rest of their software stack.

  • by electrosoccertux ( 874415 ) on Tuesday November 06, 2012 @01:39PM (#41895737)

    You should know the answer before you ask the question if you're really entrepreneur quality.

  • by hawguy ( 1600213 ) on Tuesday November 06, 2012 @01:39PM (#41895745)

    If he doesn't sell to Facebook, then he'll give up control to a VC firm.

    If he really wants to compete in the enterprise storage market, he'll need lots of money, so he'll be giving large parts of his company to some VC (or multiple VCs).

    Hardware storage density is only part of the answer, no one buys Netapp or EMC for the raw price per GB or how many disks they can fit into a rack shelf, they buy for the hardware resiliency and software stack that surrounds the disks. So he's got a lot of work ahead of him to prove that it really is a viable solution.

    There's probably not a whole lot of overlap between people that need 5PB of storage in a rack, and people that are willing to use an underfunded 19 year old's product to store the data. He's going to need serious funding to make any inroads into the increasingly crowded dense storage space.

    If I were him, I'd take whatever money FB is offering and retire.

  • Is worth two in the bush.

  • So the storage is more densely packed, but you probably need twice the cooling to keep it from melting. So do you even save any space all things considered and do you use more power per byte because of this density?

  • by axehind ( 518047 ) on Tuesday November 06, 2012 @01:43PM (#41895803)
    A 19 year old doesnt know anything about running a business. Also it is likely someone will beat the density in the near future, so sell if you get a fair offer.
    • A 19 year old doesnt know anything about running a business. Also it is likely someone will beat the density in the near future, so sell if you get a fair offer.

      This is very short-sighted of you. Don't discount this 19 year old's business sense. Hell, I'll be he's doing a heck of a lot better financially at 19 than most people 30 years older then him. I agree sell, but not because he's not capable, but because he will have the capital to move on to a bigger and better business.

  • But if an offer from Facebook ever comes, should he accept, or try to build something on his own?"

    He's 19 years old for crying out loud ... pick a giant magic number like $200 mil or so, and go to work for them ... he can always become a billionaire later. Think of it as startup capital. :-P

  • by Mike Buddha ( 10734 ) on Tuesday November 06, 2012 @01:48PM (#41895861)

    ... he shouldn't take business advice from Slashdot. It's filled with ankle-biters and mavens who are sure they've got all the answers despite not having ever been close to being in a similar predicament.

  • That's multiple lifetimes worth of money. SELL. Don't live to work, work to live.

  • Have a chat with Kim Dotcom.

    Retire in New Zealand.

  • by broknstrngz ( 1616893 ) on Tuesday November 06, 2012 @01:58PM (#41895967)

    And this is one.

  • Would the payment be in cash or Facebook stock?

  • ...is ask a bunch of random slackers who have time to waste posting to Slashdot how to handle a multi-million dollar, life-changing decision.

  • If he's getting paid Facebook stock that he can't sell for two years, there's no telling how much it will be worth by the time he can sell it.

  • maximizing storage density isn't something that other competitors will have a hard time with, and they will have more experience with hardware design, and so will have lower production costs and longer MTBF.
  • by Zontar_Thing_From_Ve ( 949321 ) on Tuesday November 06, 2012 @02:02PM (#41896031)
    He should consider selling but for different reasons than others have suggested.

    1. I work for what was once a very successful start up that a Fortune 500 company acquired, so I am an employee of the Fortune 500 company. I was not hired by the start up until near the time that the acquisition happened, so I was not around in the early days. However, one of the things I saw from the early employees was this supreme arrogance that the company was successful only because they were all geniuses and that everything they did afterward could not possibly end in failure. I know that a few of the founders, all of whom left after the company was sold, have tried to start new businesses and none of them have yet taken off. One or two of them might, but the jury is out. My point is that it's actually hard to build a successful business, but everyone who does it fails to recognize that they beat the odds and they become convinced that they simply cannot ever fail. There are a few guys who really can turn every business they start into a success, but most can't repeat the success.
    2. It's really hard to compete with bigger, older companies. Mayhall may truly have the best product, but he may be limited in sales because some clients may prefer to go with bigger, more established companies just in case. The start up I briefly worked for was sold because the owners had basically grown the business as far as they could on their own and they needed a larger partner with more and (truth be told) better sales people if it was going to grow. After the buyout of our company, our sales went through the roof and we grew at a rate we could never have achieved on our own.
    3. The insight he had to offer more density may be patented (I don't know), but someone else will eventually come up with the same idea even if they never see his patent. They might make it just different enough to get their own patent on it. Or they may simply willingly infringe it, gambling that they can win a court battle or that Mayhall won't have the money to stick it out in a protracted fight.
    4. There are companies that didn't sell when they could have and they lost market share over things they couldn't see happening when they were at the top. There's always a risk.
    5. Mayhall may well have the arrogance that youth has (ie. Zuckerberg) that everybody older than him is an idiot and only young people have any idea what they are doing and he can beat his competition because they are old and stupid. That may actually end up being true, but it probably isn't going to. He could always prove incompetent as the head of a larger company and make a lot of bad decisions. Jim Balsillie was king of the world for a while and now he's just the guy who ran RIM into the ground. Jim wasn't as young as Mayhall, but he certainly had the same "I simply cannot fail" attitude.
  • An Instagram is defined as Facebook what paid for a Photo Touchup service. His stuff could actually make them more competitive. A bird in the hand is worth 5 of what some VC will claim you will make if you hold out for an IPO.
  • by Revotron ( 1115029 ) on Tuesday November 06, 2012 @02:15PM (#41896211)
    http://evtron.com/ [evtron.com]

    Based on what I see on their website and how much I've heard about the company (Answer: Nothing and nothing) an offer from Facebook would be a miracle.

    I have a feeling this high-density storage idea is just a creative way to stick as many drives in a rack as possible, which means someone will easily beat it in no time because, face it, advances by the big storage companies means drives are getting more and more dense every day. Also, it won't be long before someone else comes along with an even-more-efficient arrangement of hard disks in a custom-built chassis in a 48U rack.

    If Facebook wants to buy what's on your drawing board, go for it. You have two viable options at this point:

    1. Sell for as much as you possibly can. Get every penny out of it that they're willing to give. Then take that money, hire people smarter than you, and make something even better than what you sold Facebook.

    Or,

    2. Negotiate a payoff for your idea, and employment. Go work for Facebook and help them bring your dream to life. You'll still get money for your idea but you'll make it come alive through their bankroll. They'll hire on very bright people and bring in storage experts to tweak things. I know it might suck to watch them come in and change your pet project, but watch and learn. You'll learn from these people, see flaws you never took into account, get to know them, work with them, and (forgive the schmoozing business term) network with them. If you choose this option you'll have a few good things going for you:

    a. A job. You didn't finish high school - if you don't start another business, there's sadly not much else you can do.
    b. Solid work experience for a Silicon Valley company. This will pay off in the long run.
    c. Relationships with people in the industry. It's invaluable.
    c. Knowledge. You'll learn things you don't yet know about the storage industry and the tech behind it.

    These are second-degree hypotheticals though. If you're talking about it publicly that means you haven't signed an NDA, which means they haven't made you an offer yet. Don't count your chickens before they hatch, but always look a few steps ahead.
  • by lga ( 172042 ) on Tuesday November 06, 2012 @05:44PM (#41899811) Journal

    This product doesn't quite add up.
    I've watched the video of their presentation at DEMO [youtu.be] and the concept is basically a 4U box containing 120 hard drives and has three fans on the front. They claim it takes 45% less power, 66% less space and 38% less heat output.

    I can see that requiring less power for cooling would reduce the overall power consumption, but can't see how pumping the same heat out of three fans would achieve that. They also don't say how these drives are interfaced. Presumably there is some kind of controller in the box since there aren't 120 SAS connections coming out of it, but somewhere you have to have a server in charge of those disks and a connection with enough bandwidth to run them all. And then we need to know if there is any RAID intelligence in the boxes or if the server gets them as JBOD.

  • by pikine ( 771084 ) on Tuesday November 06, 2012 @09:13PM (#41902211) Journal

    Let's do a little math here.

    • Largest capacity 3.5" hard disk on market today is 4TB. It takes 4.6PB / 4TB = 4600TB / 4TB = 1150 hard disks to achieve this.
    • Suppose a rack is 48U, 19" wide and 42" deep [server-rack-online.com]. 1U is 1.75" height. This gives 67032 cubic inch. Each hard drive is allowed 67032 cubic inch / 1150 = 58 cubic inch.
    • A typical 3.5" hard drive is 4" x 1" x 5.75" = 23 cubic inch.

    This does leave a comfortable amount of space for ventilation, wiring, and RAID controller. You basically build a whole rack full of hard drives.

    But this is extremely impractical. You'll need a fast interconnect between other computing node racks and this storage rack because all storage access to raw data has to go through this pipe. On the other hand, if you basically keep it one or two hard drive per computing node, then raw data never needs to travel across the interconnect, and you get better locality as result.

I have hardly ever known a mathematician who was capable of reasoning. -- Plato

Working...