Businessweek Archives

Not Your Dad's Mainframe: Little Iron


? Rocketboom: What a Mess |

Main

| eBay's Google Checkout Ban: Hold the Rotten Tomatoes ?

July 06, 2006

Not Your Dad's Mainframe: Little Iron

Steve Hamm

Mainframe computers have long been the province of IT druids in huge corporate data centers and goverment agencies like NASA and the CIA. Now, though, the mainframe is heading towards a much larger potential market.In IBM's latest effort to keep Big Iron relevant in a fast-changing computing world, it's retooling the technology for small- and medium-size businesses. The new z9 Business Class mainframe, released this spring, can be bought for as little as $100,000. Think of it as Little Iron. And small outfits who can't afford to buy a mainframe can pay by the drink by using IBM's on-demand services. All the things IBM mainframes can now do will surprise you. Example: As the server for Taikodom, a massively multiplayer online game being developed by Brazilian upstart Hoplon Infotainment. It's a company with just 50 employees. I don't want this to sound like an ad for mainframes (not my role or inclination) but this cool application caught my fancy.

Tarquinio Teles, CEO of Hoplon, managed to shake off Brazil's World Cup loss to France and come to the phone to tell me his story. He and three university buddies started Hoplon at an inauspicious time: mid-2000. The idea was to create multiplayer online games like Everquest. There's never very much venture money in Brazil for tech startups, but the dot-com bust made the situation even worse. The four pals worked on their project part-time initially, then in 2003 raised a bit of money from an angel investor--enough to create a real company with employees and all that.

Work on Taikodom (I forgot to ask what it means) started in 2004. The idea was to create a virtual universe where players can have a great deal of flexibility in what they do, how they interact with each other, and what happens in the place. Rather than having a few people team up and go on prescribed missions, players are encouraged to explore this slice of virtual reality as pilots or crew members of space ships, conquer worlds, or set up mining or other commercial operations. Teles calls the concept a Massive Social Game. Also, potentially, they wanted to be able to have anybody in the world interact with anybody else. In most massively multiplayer games, you can only fraternize with people who land on the same server as you do.

This leads us to the mainframe. Back in 2003, Teles and some of his friends attended a grid computing conference put on by IBM in Sao Paulo. They met up with some IBM technical wizards and talked over their dreams and were stunned to hear that the mainframe might be the best computing choice for a small company with a massive project like they had in mind. (They started with the idea of using Linux running on commodity Intel servers) IBM offered to host their application at its data center outside Sao Paulo and deliver the service on demand and metered. The company tried it in a public beta test involving 15,000 people last year and liked it. Because they didn't have to buy computing hardware and software, they were able to invest in programmers, increasing the staff from 35 to 50 recently. "We are only paying for what we use," says Teles. "For a startup like ours that's vitally important." Teles says in North America it would probably take $20-40 million and 200 engineers to build a game like this. But, thanks to the on-demand setup and Brazil's low cost of programming, they have only spent $3.5 million in 2 1/2 years.

Now Taikodom is back in the workshop for some improvements suggested by beta users. They wanted some structured missions, for instance. Teles hopes to go commercial in Brazil and Portugal by the middle of next year, then, if the game catches on, expand to North America and Europe. Asia is a possibility later on.

What's next? The corner drug store running applications on a mainframe? Probably not. But it's a real possibility for drug discovery startups and the like--companies with relatively small budgets and really big computing projects.

11:35 PM

mainframes

TrackBack URL for this entry:

http://blogs.businessweek.com/mt/mt-tb.cgi/

Steve,

Nice article, but what is (are) the compelling reason(s) for picking a mainframe vs. the alternatives (which for startups is as you say a comodity Intel server with Linux).

Conceptually I like the idea of mainframes. The only reason I hated them in the 1970's and early 1980's is the desktop software to interact with them was terrible. And there is this huge wall of people bureaucracy place around them. I presume by now both of these fixed...

Posted by: Rob Schneider at July 6, 2006 01:41 AM

What? The Vapour is dense in this one, Luke!

Posted by: What? at July 6, 2006 11:11 PM

Dear Steve,

I would like to ask you: did you try this game?

this is a complete hoax, this piece of vapourware do not support more than ONE player at a time without the worst lag ever seen in any MMOG in the world

by the way, the money put in this project were completed wasted, 3.5 million dollars is a lot of money for such an awful game (do not be fooled by the so so screenshots or fancy website) the game per se completely sucks

I?would feel ashamed of publishing such a thing without verifying the facts first

Posted by: eu at July 6, 2006 11:13 PM

Have you checked if this game really works? Looks like extreme VAPOURWARE for me, or an ad for IBM mainframes...

I know it doesn't work... I'm from Brazil, this game is a disappointment, it never worked.

Just a waste of money...

Posted by: Check your sources at July 6, 2006 11:16 PM

two things:

1) I didn't try the game, but it sounded cool. Yes, its is vaporware until it exists--like Microsoft Vista.

2) The advantages of the mainframe architecture are manageablity and security. In the case of the massively multi-player game, it apparently works better than a grid--but the reasons are technical and I am not qualified to try to explain them.

Posted by: Steve Hamm at July 7, 2006 12:07 PM

I am very sure that this article came right from the mouth of Hoplon's CEO.

Check the progress of the game, the money and personal invested you start to think its a joke or theres something else goining on there.

Try contacting small people at Hoplon, the most comum answers about that that are: "fill your pockets and get out", "Its a mess", "Taikodom will never be true".

Steve, do some research about, you will be very surprised about how wrong your article is.

I post it becouse am very sad that the first big game project from Brasil is a Titanic, I hate to have it as a reference.

Posted by: Annon Silva at July 8, 2006 06:03 PM

Hey you guys,

How did you try to access this game???

I've been playing Taikodom for almost a year now, and its great!! There are hundreds of players online at any one time.

Whoo hoo!!!

Posted by: Manuel Hogaros at July 9, 2006 05:57 PM

The claim that this kind of game takes usually more than US$20-40 million to develop is misleading.

1) only a couple of MMOGs reach these kinds of budgets, games like "World of Warctaft", but in return they're huge and profitable products

2) most MMOGs out there (Ragnarok, Auto Assault) are developed with much smaller budgets, but try the Taikodom demo: by comparison it has less than 1/10 of the features, stability and content of a typical MMOG at the same development stage and costs, so, doing a simple math: Taikodom will probably take tens of millions to reach the same competitive levels.

The bottom line: inexperienced and 'cheaper' programming teams not always mean a smaller development cost.

Posted by: anuser at July 12, 2006 12:31 PM

Sorry if I gave you the wrong impression with the numbers, anuser. I agree with you we spend a large part of our dollars up to now climbing the learning curve, and not actually delivering a great game, yet.

And yes, it will take a couple million dollars more to get to Gold.

But I think you are missing my point here. We are not saying we delivered a fantastic game for a incredibly low budget. We are saying we have a incredibly tight budget, lots more to do, and still we are using a mainframe to do it.

What I meant is, we are not swimming in cash, and decided to have a go on the mainframe. We didn't go for a mainfame because we lacked challenges. We did it because exactly because we had to spend wisely, and we were willing to take risks.

Just to be clear, I never said we'd proven Hoplon can make massive multiplayer games. What we sure proved is mainframes can run them. Even when they are a long way off from completion.

In the end, there is no such a thing as free sandwiches, as you use to say in the US. So please stay tuned: Taikodom may not be ready, but it sure ain't dead, either.

Posted by: Tarquinio Teles at July 17, 2006 01:39 AM

I am not sure about Brazil, but in my country, spending a large part of the investor?s dollars into ?learning?, as you said, instead of the development of the product itself (Taikodom), could even lead to an investigation of some kind, with very negative consequences for your company.

Posted by: Mike J. Bernett at August 1, 2006 06:35 PM

well to be fair, brazil is out in the sticks with regards to the game industry. learning curve is inevitable.

Posted by: steve dobbs at August 3, 2006 09:25 PM

Hi, im one of the "Beta testers" for more than a year and what i can tell you about Taikodom?

Taikodom has a great ideia behind it, in this game we will have an freedom that we will never found in anyelse game...The economy will be huge, detailed, fascinating, noone?s will be auto-suficient... every player will need resources than other have, to build something...

the most important, THERE IS NO LEVEL on this game... the best player will be that who can manage better is spaceship, in the usuals MMO games that we see on the market, the best player is that who has the best itens...this will not happen in Taikodom...

The players organize themself in Corporations, like clans in other games, but here it?ll be diferent, it will see more like an small company, with employers doing things that is related in a contract, it will be a very interessant way to trys to reach in a game, what the companys does in the real world, as an leader of an Corp, im always thinking in ways to reach the best profit, the best ofensive force (and defensive too), how i?ll organize my army in corporations war?

This game sound fascinating for me, of course that the things that i said here, isnt "Oficial", these informations i reached talking with some of the Taikodom Game Masters that can be found in the IRC channel of this game...

If Hoplon reach all that they have in mind, be sure, Taikodom will be the Best Online Game Ever made...and as a Beta Tester, i can tell you that they are in the way to reach that!!!

(sorry, but my english isnt very good =/)

Bruno

Belo Horizonte, MG - BRAZIL

the best thing in brazil, is the brazilians =]

Posted by: Bruno (coiote) at August 12, 2006 03:42 PM

Bruno, indeed the concept behind Taikodom is great, but thanks to Mr. David Braben who created this concept of open space games in 1984 for his game "Elite". So, it is nothing new - games like Eve-Online and Taikodom just attempt to integrate this in an online environment. Only Eve-Online so far had success, and it didn?t costed so much more than Taikodom by what I read here...

Speaking in Eve-Online, they seem to have exactly the same kind of distributed IBM server thing like Taikodom, impressive, who copied who? Check this:

http://pc.gamezone.com/news/09_08_06_11_30AM.htm

I have to say Eve-Online has the upper hand IMO, they actually have a working game right now without any terrible lag effects....

Posted by: Guilherme C. Andrade at September 11, 2006 02:46 PM


Race, Class, and the Future of Ferguson
LIMITED-TIME OFFER SUBSCRIBE NOW

(enter your email)
(enter up to 5 email addresses, separated by commas)

Max 250 characters

 
blog comments powered by Disqus