5 ms·

Why does almost every startup have to be a social web site or some other web app? The web is surely not the only thing in the hacker universe, is it? The bigges

by Mapou 18y ago

Why does almost every startup have to be a social web site or some other web app? The web is surely not the only thing in the hacker universe, is it? The biggest opportunities in computing today are in parallel programming tools and multicore architecture and design. How come nobody seems to care about those things? I note that your Phd dissertation was titled, "Prefetch mechanisms by application memory access pattern". Surely this qualifies you for all sorts of computing projects.

paul 18y ago
Parallel programming tools may make for interesting research, but that doesn't mean that it will be a good business. In general, tools startups haven't been terribly successful.
- Mapou 18y ago
  I see what you're saying and I think you're probably right. But if you can design and build a kick-ass multicore processor that blows everyone else out of the water in terms of performance, energy efficiency and ease of programming, it makes sense to give all your dev tools away for free. Heck, you could even throw in an OS and a nice browser for good measure. I realize that I'm biased since I happen to think that multicore design is the most exciting thing out there. In my opinion, whoever solves the multicore/parallel programming problem will dominate computing for the rest of the century. The rest of the industry, including Intel, IBM, AMD, APPLE, etc... would flock like migratory birds to come and worship at your feet.
  - immad 18y ago
    If I gave you $100k could you build me a kick-ass multicore processor that beats Intel+AMD? Seems like its not a good space to be a startup in. But I don't know that much about it...
    - Mapou 18y ago
      Not $100k, of course. But I think $5 or $6 million dollars could do it. For that, you'd get a scalable, self-balancing 16 or 32-core processor compatible with existing motherboards and chip sets, visual dev tools using multi-touch screens, and a fully built computer with an extensible parallel OS including a good set of device drivers. The whole engineering/design phase would take about two years. After that, you enter the marketing and production phase which will require several busloads of dough but that's not my department.
      - rw 18y ago
        Intel/AMD aren't exactly resting on their laurels, friend.
        Mapou 18y ago
        In my opinion, that's exactly what they are doing. They have adopted a paralllel computing model (multithreading) that is known to be a pain in the ass to program. So far, they have no plans to change the model. They either don't know how or they are stuck in a rut because they need to maintain compatibility with legacy systems. They are in a state of panic right now, spending money left and right trying to find a solution that does not exist for their chosen model. This is a great opportunity for some other company (or a startup) to come on the scene and make a killing.
  - paul 18y ago
    If you could design and build a flying car that runs on tap water, that would be even better. Unfortunately, that's easier said than done. Competing with Intel, AMD, etc in the processor market is also not easy -- many have tried, and generally failed.
    - Mapou 18y ago
      Yeah, but this is a different ball game. The big players have an Achilles' heel. They are clueless as to what the future holds for multicore processors and they don't seem close to a solution. Parallel programming is a pain in the ass. There is even panic in the air because the big vendors have no real idea how to proceed. The solution is out there though. Some unknown startup may sneak behind them and steal the pot of gold while they're busy fighting amongst themselves.
      - paul 18y ago
        Anything is possible, but so far the flying car and parallel programming startups have all failed :). (Moller and Peakstream come to mind)
        Mapou 18y ago
        Which parallel programming startups are you referring to? Sure, the established companies are adapting their exisiting tools to address the new multicore processors but I am not aware of any new parallel programming tools startups that failed. Most of the multicore processor startups are still around. Tilera, Ambric, Pico, etc... are still hanging in there. Pico is doing great. Israel's Plurality is just now beginning to get recognition for its self-balancing 64 to 256-core hypercore chip. My point is that there are very few new multicore startups and I think that the reason is not that too many of them failed in the past but that only a few have had anything really interesting or disruptive to offer. It remains that there is a big problem that needs to be solved right now and whoever solves it will grab a lion share of the CPU market in this century.
        aswanson 18y ago
  - axod 18y ago
    How is it exciting? What problem does it solve? How to write programs even lazier and introduce more complexity? Multicores seem like a bad hack to me - a stop gap until chip speeds increase. A single core has to be the most efficient. When everything is moving onto the web, consumers don't have much of a need to multicore processors. Maybe in high end db servers or something :/
babul 18y ago
Its not because nobody cares but more to do with business. Web apps, social sites, news, ecommerce and so on, is what consumers want and where the money is. The computer science and computing issues are the enablers, not the products.
- Mapou 18y ago
  I disagree that a kick-ass multicore processor is not a product, especially if you can get $500 for every one that comes out of the fab.
babul 18y ago
Also, forgot to mention the fact doing a web startup in these areas is cheap, quick, easy, and more likely to have a bigger payoff compared to building the kind of tools you mention.
maxklein 18y ago
The real killer app would be a low level development tool that does not require you to use 4-5 different tools and wait near to 5 minutes to test your design on hardware.
akkartik 18y ago
After 8 years bleeding my motivation waiting for week- or month-long simulations to run, I strongly believe nobody should work on problems where the feedback loop isn't insanely short. I care far more about working someplace small than about working on something web-based. There's just a lot less communication and bureaucracy, better odds of avoiding dysfunction. Web-based startups just happen to satisfy both these constraints - the web as delivery mechanism yields instant feedback[1], and the web as eco-system encourages small and agile organisms. [1] And high quality feedback. As an academic I was separated from my readers by an advisor and a dissertation committe, or by a set of anonymous reviewers and a program committee. Large companies separate you from the customer by layers of hierarchy. In a startup you're right next to your customers.

by Mapou 18y ago

paul 18y ago

Parallel programming tools may make for interesting research, but that doesn't mean that it will be a good business. In general, tools startups haven't been terribly successful.

Mapou 18y ago
I see what you're saying and I think you're probably right. But if you can design and build a kick-ass multicore processor that blows everyone else out of the water in terms of performance, energy efficiency and ease of programming, it makes sense to give all your dev tools away for free. Heck, you could even throw in an OS and a nice browser for good measure. I realize that I'm biased since I happen to think that multicore design is the most exciting thing out there. In my opinion, whoever solves the multicore/parallel programming problem will dominate computing for the rest of the century. The rest of the industry, including Intel, IBM, AMD, APPLE, etc... would flock like migratory birds to come and worship at your feet.
- immad 18y ago
  If I gave you $100k could you build me a kick-ass multicore processor that beats Intel+AMD? Seems like its not a good space to be a startup in. But I don't know that much about it...
  - Mapou 18y ago
    Not $100k, of course. But I think $5 or $6 million dollars could do it. For that, you'd get a scalable, self-balancing 16 or 32-core processor compatible with existing motherboards and chip sets, visual dev tools using multi-touch screens, and a fully built computer with an extensible parallel OS including a good set of device drivers. The whole engineering/design phase would take about two years. After that, you enter the marketing and production phase which will require several busloads of dough but that's not my department.
    - rw 18y ago
      Intel/AMD aren't exactly resting on their laurels, friend.
      - Mapou 18y ago
        In my opinion, that's exactly what they are doing. They have adopted a paralllel computing model (multithreading) that is known to be a pain in the ass to program. So far, they have no plans to change the model. They either don't know how or they are stuck in a rut because they need to maintain compatibility with legacy systems. They are in a state of panic right now, spending money left and right trying to find a solution that does not exist for their chosen model. This is a great opportunity for some other company (or a startup) to come on the scene and make a killing.
- paul 18y ago
  If you could design and build a flying car that runs on tap water, that would be even better. Unfortunately, that's easier said than done. Competing with Intel, AMD, etc in the processor market is also not easy -- many have tried, and generally failed.
  - Mapou 18y ago
    Yeah, but this is a different ball game. The big players have an Achilles' heel. They are clueless as to what the future holds for multicore processors and they don't seem close to a solution. Parallel programming is a pain in the ass. There is even panic in the air because the big vendors have no real idea how to proceed. The solution is out there though. Some unknown startup may sneak behind them and steal the pot of gold while they're busy fighting amongst themselves.
    - paul 18y ago
      Anything is possible, but so far the flying car and parallel programming startups have all failed :). (Moller and Peakstream come to mind)
      - Mapou 18y ago
        Which parallel programming startups are you referring to? Sure, the established companies are adapting their exisiting tools to address the new multicore processors but I am not aware of any new parallel programming tools startups that failed. Most of the multicore processor startups are still around. Tilera, Ambric, Pico, etc... are still hanging in there. Pico is doing great. Israel's Plurality is just now beginning to get recognition for its self-balancing 64 to 256-core hypercore chip. My point is that there are very few new multicore startups and I think that the reason is not that too many of them failed in the past but that only a few have had anything really interesting or disruptive to offer. It remains that there is a big problem that needs to be solved right now and whoever solves it will grab a lion share of the CPU market in this century.
        aswanson 18y ago
- axod 18y ago
  How is it exciting? What problem does it solve? How to write programs even lazier and introduce more complexity? Multicores seem like a bad hack to me - a stop gap until chip speeds increase. A single core has to be the most efficient. When everything is moving onto the web, consumers don't have much of a need to multicore processors. Maybe in high end db servers or something :/

babul 18y ago

Its not because nobody cares but more to do with business. Web apps, social sites, news, ecommerce and so on, is what consumers want and where the money is. The computer science and computing issues are the enablers, not the products.

Mapou 18y ago
I disagree that a kick-ass multicore processor is not a product, especially if you can get $500 for every one that comes out of the fab.

babul 18y ago

Also, forgot to mention the fact doing a web startup in these areas is cheap, quick, easy, and more likely to have a bigger payoff compared to building the kind of tools you mention.

maxklein 18y ago

The real killer app would be a low level development tool that does not require you to use 4-5 different tools and wait near to 5 minutes to test your design on hardware.

akkartik 18y ago

After 8 years bleeding my motivation waiting for week- or month-long simulations to run, I strongly believe nobody should work on problems where the feedback loop isn't insanely short. I care far more about working someplace small than about working on something web-based. There's just a lot less communication and bureaucracy, better odds of avoiding dysfunction. Web-based startups just happen to satisfy both these constraints - the web as delivery mechanism yields instant feedback[1], and the web as eco-system encourages small and agile organisms. [1] And high quality feedback. As an academic I was separated from my readers by an advisor and a dissertation committe, or by a set of anonymous reviewers and a program committee. Large companies separate you from the customer by layers of hierarchy. In a startup you're right next to your customers.