Tensilica and Energy Efficient Computing


The proposed sustained 10 petaflop and 200 peak petaflop tensilica computer is based on 2008 technology. Over the next few years, Tensilica will have 45 nanometer and 32 nanometer chips with improved energy usage and performance, which would enable affordable supercomputers that are more powerful than 10 petaflops.


Other proposed, in development or completed supercomputers and their how specialized or general purpose they are.


Tensilica and their configurable processors could be used as the basis for exaflop and multi-petaflop computers. The main challenge for faster computers is increasing the amount of computer per watt of power. This power challenge is also critical because computing is using 6% of overall power.

What Tensilica is Doing for More Computing and Less Power

Overall Computing Energy Efficiency Trends

The Green list of supercomputers has a current peak of 536 Mflops per watt

Cell processors (in particular cell broadband engine, Cell/BE) are dominant in energy efficient supercomputers IBM will be producing the Cell at 45 nanometers early in 2009 and said it would require 40 percent less power at the same clockspeed than its 65nm predecessor and that the die area would shrink by 34 percent.

The trend is to mixing different types of processors (FPGAs, Cell processors, GPGPUs, floatingpoint accelerators and more) to get more energy efficient computing.

The configurable Tensilica processors provide a wider range of processor types and flexibility.

0 thoughts on “Tensilica and Energy Efficient Computing”

  1. Interesting topic. Decoupling happens, but it is sometimes the observed delay in an economic system’s “resetting” itself for dealing with the new technology.

    The moire pattern where economics overlaps the real world is difficult enough to describe, so that it is rare to find two economists who agree on many points.

    If only more people listened to Richard . . . .
    😉

    Reply
  2. Your input will not be missed Richard Kulisz grand master troll.

    RichardKuliszFanClub discussing the trolling gems of Richard.

    Discussing Richard as troll

    RichardKulisz

    133 entries for Richard Kulisz as troll Entries clearly are labelling Richard as troll.

    A reddit discussion of Richard getting banned

    Kulisz acts like his comments are so groundbreaking that they are
    worthy of censorship. He is, and always has been a blowhard looking
    for attention.

    Richard Kulisz has been nominated as a grandmaster of trolling.

    Reply
  3. I was talking about China’s high growth as an example of decades long higher growth related to the topic of the next step hypergrowth being achievable related to a technological next stage.

    China’s issues were not the main topic. Also, in spite of the environmental damage caused by China’s rise and the rise of the US and the UK back in the 1920-1960s when they also had a lot more pollution the growth happened.

    If the growth happened and if hypergrowth will happen (because of the profit to be made from it) then plans should be made to reduce any environmental damage.

    Do you have a prediction bet to make in regards to your belief in Michael Goldhaber writings? You believe strongly in it and seem quit passionate about it. What is the bet where your belief is right that GDP growth will stop. Give some dates and measures. You like to talk big, come on put some money where your mouth is.

    Reply
  4. > I decided not to take the time to correct commonly accepted GDP and economic measurements in my blog post. I also decided not to mention the negative net value of cigarettes and coal. I mention that a lot but not every post. Otherwise I would have to have some long standard boiler plate caveats to go onto every posting.

    It’d be one thing if you were talking about economic growth. But you’re talking about economic growth *in china* which is hugely reliant on coal. Come on, you don’t think that’s relevant? *Really*? You *really* don’t think the multi-trillion environmental debt China has amassed means nothing?

    > You are correct in that I meant to say increased decoupling.

    Michael Goldhaber makes an excellent case for near-total decoupling. In fact, he makes an excellent case that decoupling IS the next wave of economic growth. Not counting catch-up countries like China and to a lesser extent India.

    So I find it rather ludicrous for you to be talking about growth and hyper-growth and increasing growth rates, in a totally analytical (hence worthless) CRAP format when you have people like Goldhaber to give valuable insights into economics. And Goldhaber says the EXACT OPPOSITE of what YOU say. Analytics vs fundamentals: guess who’ll win?

    I also have nothing but contempt for analytics. A contempt that stretches long before the current housing bubble and its unwinding. So I don’t appreciate your usage of analytics in any way, shape or form.

    You’ve given no reason to think GDP growth will continue, or that it’s even meaningful. You can’t since there are ironclad reasons why it *won’t* continue and why it *isn’t* meaningful. Oh yeah, that’s another reason for my displeasure: your blatant violation of facts and logic as we both know them.

    Reply
  5. Richard

    thanks for commenting so politely.

    I decided not to take the time to correct commonly accepted GDP and economic measurements in my blog post. I also decided not to mention the negative net value of cigarettes and coal. I mention that a lot but not every post. Otherwise I would have to have some long standard boiler plate caveats to go onto every posting.

    You are correct in that I meant to say increased decoupling.

    Btw: I will keep commenting about economics and if you do not like that I will be really broken up over it.

    Reply
  6. > There might be some decoupling of progress from currency transactions

    Sometimes you say things that are so incredibly dumb. There might be some decoupling? There ALREADY IS some decoupling. What the hell do you think is happening to the publishing industry?

    Furthermore, your measure of economic growth is completely fucking idiotic. You don’t measure value, you measure mere activity.

    By your own figures, the cigarette industry provides negative value to the economy. And yet you count it as positive activity. You do the same thing for China’s coal industry.

    What’s aggravating is that you persist in trying to modify this measure of activity to account for increases of value. Meaning, you’ve got a bastard metric of the economy that measures NOTHING.

    Economics is one of your major blind spots. It is a subject on which you apparently know nothing. And you should really STFU about it instead of spreading lies. It would increase your personal signal to noise ratio.

    Reply

Leave a Comment