9 ms·

Sequoia: 20 Petaflops, 1.6 million cores, 1.6 Petabytes RAM

25 points by timf 18y ago

biohacker42 18y ago
It's interesting to think about the possible uses for a machine like that. My own personal interest is protein folding. Proteins are folding right now, as you read this, in every cell of your body, and they're doing it in linear time. But computationally we're stuck with exponential time. The day we find an algorithm for linear time protein folding, is the day the world changes at least as much as when penicillin was discovered.
- anamax 18y ago
  > My own personal interest is protein folding. > But computationally we're stuck with exponential time. It's not that bad and it "recently" got a lot better. Search for "shaw protein folding neutral territory".
  - bbgm 18y ago
    It's not even close. Shaw's work is good, but we are still not close to folding proteins from a linear sequence. The physics is just not there
- fizx 18y ago
  Protein folding is NP-complete (Berger et al 1998). This implies that we can get fast heuristics for certain subclasses of proteins, but there is no general linear time protein folding solution.
  - biohacker42 18y ago
    I don't know of anything in nature which solves the traveling sales person problem, but protein folding does progress in linear time in nature. That may mean that proteins quickly collapse the probability space or have some other "trick" by which they are not NP-complete, or whatever, it is not magic. There is a way and I'm hopeful we can find it.
    - kurtosis 18y ago
      Hey this is from the abstract of: Sequence optimization and designability of enzyme active sites - PNAS v. 102 (2005) Full text is here: http://www.pnas.org/content/102/34/12035.full http://www.pnas.org/content/102/34/12035.full """ Since it was first observed that all known natural proteins adopt one of only ≈1,000 distinct protein folds (1) [emphasis mine], researchers working in the field of computational protein design have endeavored to understand the basis of this biophysical phenomenon and apply the resultant knowledge to the construction of unnatural proteins. Motivated by the observation that certain folds are associated with more natural sequences than others, the principle of fold designability (the number of unique sequences compatible with the backbone structure of the fold) was introduced (1). To place the problem on a mathematical foundation, an analogy has been drawn between sequence space and phase space in statistical physics, where the backbone structure represents a macroscopic state of the system and distinct sequences are microscopic states whose occupation of the macroscopic state is associated with an energy, the folding free energy of the protein. In this language, a fold's designability is correlated with its sequence entropy (2). """
    - kurtosis 18y ago
      For what it's worth I attended a seminar by this guy http://www.princeton.edu/~rajchak/ http://www.princeton.edu/~rajchak/ the other day and he was talking about using optimal control theory to optimize protein sequences for specific enzyme activity. He had a very sophisticated model for how the amino acid sequence determines the shape of the protein's active area. It was way over my head but he treated it as a solved problem. You may find his papers interesting. The combinatorial search results on protein folding are theoretically interesting, but I don't think that they are show stoppers for finding, or optimising the folded configuration of a protein. As you mentioned, most of the degrees of freedom which lead to the exponential size of the search space are irrelevant. A detailed understanding of the physics of how amino acids interact in water with ions can allow one to make approximations that vastly reduce the size of the search space. Plus how much time does it take a protein to find it's native conformation? Let's say the folding time is 10 ms. I think ultrafast pump probe spectroscopy has shown that alpha helices and other secondary structural motifs form on the time scale of 100 ps. Then the protein explores about 10^8 configurations before it finds the native one.
- bbgm 18y ago
  I do think that really understanding molecular machinery (known proteins and their dynamic interactions in atomistic glory) and true molecular recognition problems are what we should be doing. That said, we need better physics to represent these interactions. The fields been relatively static when it comes to forcefields that really work.
lst 18y ago
...and is it finally coming close to .00001 % of a human brain? If someone is excited about that, what about the human brain itself? Just a simple information: The human brain is the most complex thing in the whole universe, or: all of physical universe is less complex than a single human brain.
- acro 18y ago
  May I ask what is the basis of your estimate on the computational capacity of the human brain?
  - timf 18y ago
    Especially pointed question since the statement implies someone has crossed the chasm of the physical universe in order to measure this. (since "all of physical universe is less complex than a single human brain" and the physical universe contains around 6.7 billion brains (that we know of))
    - lst 18y ago
      I said "physical" to exclude the misunderstanding.
      - ryanwaggoner 18y ago
        Pretty sure you failed.
      - timf 18y ago
        My point was that if you are restricting the comparison to the physical universe then logically you must be implying the brain extends beyond the physical universe somehow -- since there is more than one brain in the physical universe. Your subsequent statements in this thread confirmed you have this opinion. If true, then it has an undefined and immeasurable complexity. You can't make all these statements about it, you can't measure it, and you have no context to even begin to talk about it. You can't even make assertions about what the measurement of complexity itself would be ("infinitely more" is just as nonsensical as a finite number more because we do not know what the units of a complexity measurement would be -- if "units" even makes sense in the "beyond").

25 points by timf 18y ago

It's interesting to think about the possible uses for a machine like that. My own personal interest is protein folding. Proteins are folding right now, as you read this, in every cell of your body, and they're doing it in linear time. But computationally we're stuck with exponential time. The day we find an algorithm for linear time protein folding, is the day the world changes at least as much as when penicillin was discovered.

anamax 18y ago
> My own personal interest is protein folding. > But computationally we're stuck with exponential time. It's not that bad and it "recently" got a lot better. Search for "shaw protein folding neutral territory".
- bbgm 18y ago
  It's not even close. Shaw's work is good, but we are still not close to folding proteins from a linear sequence. The physics is just not there
fizx 18y ago
Protein folding is NP-complete (Berger et al 1998). This implies that we can get fast heuristics for certain subclasses of proteins, but there is no general linear time protein folding solution.
- biohacker42 18y ago
  I don't know of anything in nature which solves the traveling sales person problem, but protein folding does progress in linear time in nature. That may mean that proteins quickly collapse the probability space or have some other "trick" by which they are not NP-complete, or whatever, it is not magic. There is a way and I'm hopeful we can find it.
  - kurtosis 18y ago
    Hey this is from the abstract of: Sequence optimization and designability of enzyme active sites - PNAS v. 102 (2005) Full text is here: http://www.pnas.org/content/102/34/12035.full http://www.pnas.org/content/102/34/12035.full """ Since it was first observed that all known natural proteins adopt one of only ≈1,000 distinct protein folds (1) [emphasis mine], researchers working in the field of computational protein design have endeavored to understand the basis of this biophysical phenomenon and apply the resultant knowledge to the construction of unnatural proteins. Motivated by the observation that certain folds are associated with more natural sequences than others, the principle of fold designability (the number of unique sequences compatible with the backbone structure of the fold) was introduced (1). To place the problem on a mathematical foundation, an analogy has been drawn between sequence space and phase space in statistical physics, where the backbone structure represents a macroscopic state of the system and distinct sequences are microscopic states whose occupation of the macroscopic state is associated with an energy, the folding free energy of the protein. In this language, a fold's designability is correlated with its sequence entropy (2). """
  - kurtosis 18y ago
    For what it's worth I attended a seminar by this guy http://www.princeton.edu/~rajchak/ http://www.princeton.edu/~rajchak/ the other day and he was talking about using optimal control theory to optimize protein sequences for specific enzyme activity. He had a very sophisticated model for how the amino acid sequence determines the shape of the protein's active area. It was way over my head but he treated it as a solved problem. You may find his papers interesting. The combinatorial search results on protein folding are theoretically interesting, but I don't think that they are show stoppers for finding, or optimising the folded configuration of a protein. As you mentioned, most of the degrees of freedom which lead to the exponential size of the search space are irrelevant. A detailed understanding of the physics of how amino acids interact in water with ions can allow one to make approximations that vastly reduce the size of the search space. Plus how much time does it take a protein to find it's native conformation? Let's say the folding time is 10 ms. I think ultrafast pump probe spectroscopy has shown that alpha helices and other secondary structural motifs form on the time scale of 100 ps. Then the protein explores about 10^8 configurations before it finds the native one.
bbgm 18y ago
I do think that really understanding molecular machinery (known proteins and their dynamic interactions in atomistic glory) and true molecular recognition problems are what we should be doing. That said, we need better physics to represent these interactions. The fields been relatively static when it comes to forcefields that really work.

lst 18y ago

...and is it finally coming close to .00001 % of a human brain? If someone is excited about that, what about the human brain itself? Just a simple information: The human brain is the most complex thing in the whole universe, or: all of physical universe is less complex than a single human brain.

acro 18y ago
May I ask what is the basis of your estimate on the computational capacity of the human brain?
- timf 18y ago
  Especially pointed question since the statement implies someone has crossed the chasm of the physical universe in order to measure this. (since "all of physical universe is less complex than a single human brain" and the physical universe contains around 6.7 billion brains (that we know of))
  - lst 18y ago
    I said "physical" to exclude the misunderstanding.
    - ryanwaggoner 18y ago
      Pretty sure you failed.
    - timf 18y ago
      My point was that if you are restricting the comparison to the physical universe then logically you must be implying the brain extends beyond the physical universe somehow -- since there is more than one brain in the physical universe. Your subsequent statements in this thread confirmed you have this opinion. If true, then it has an undefined and immeasurable complexity. You can't make all these statements about it, you can't measure it, and you have no context to even begin to talk about it. You can't even make assertions about what the measurement of complexity itself would be ("infinitely more" is just as nonsensical as a finite number more because we do not know what the units of a complexity measurement would be -- if "units" even makes sense in the "beyond").