Database storage managers have long been able to efficiently handle multiple concurrent requests. Until recently, however, a computer contained only a few single-core CPUs, and therefore only a few transactions could simultaneously access the storage manager's internal structures. This allowed storage managers to use non-scalable approaches without any penalty. With the arrival of multicore chips, however, this situation is rapidly changing. More and more threads can run in parallel, stressing the internal scalability of the storage manager. Systems optimized for high performance at a limited number of cores are not assured similarly high performance at a higher core count, because unanticipated scalability obstacles arise. We benchmark four popular open-source storage managers (Shore, BerkeleyDB, MySQL, and PostgreSQL) on a modern multicore machine, and find that they all suffer in terms of scalability. We briefly examine the bottlenecks in the various storage engines. We then present Shore-MT, a multithreaded and highly scalable version of Shore which we developed by identifying and successively removing internal bottlenecks. When compared to other DBMS, Shore-MT exhibits superior scalability and 2-4 times higher absolute throughput than its peers. We also show that designers should favor scalability to single-thread performance, and highlight important principles for writing scalable storage engines, illustrated with real examples from the development of Shore-MT.
lunes, 7 de septiembre de 2009
Shore-MT: a scalable storage manager for the multicore era
Database storage managers have long been able to efficiently handle multiple concurrent requests. Until recently, however, a computer contained only a few single-core CPUs, and therefore only a few transactions could simultaneously access the storage manager's internal structures. This allowed storage managers to use non-scalable approaches without any penalty. With the arrival of multicore chips, however, this situation is rapidly changing. More and more threads can run in parallel, stressing the internal scalability of the storage manager. Systems optimized for high performance at a limited number of cores are not assured similarly high performance at a higher core count, because unanticipated scalability obstacles arise. We benchmark four popular open-source storage managers (Shore, BerkeleyDB, MySQL, and PostgreSQL) on a modern multicore machine, and find that they all suffer in terms of scalability. We briefly examine the bottlenecks in the various storage engines. We then present Shore-MT, a multithreaded and highly scalable version of Shore which we developed by identifying and successively removing internal bottlenecks. When compared to other DBMS, Shore-MT exhibits superior scalability and 2-4 times higher absolute throughput than its peers. We also show that designers should favor scalability to single-thread performance, and highlight important principles for writing scalable storage engines, illustrated with real examples from the development of Shore-MT.
The Fiscal Politics of Big Governments: Do Coalitions Matter?
This paper uses a closed voting rule to explain how gradual changes in the socio-economic structure of developed societies shift political coalitions and lead to rapid expansions of transfer programs. Equilibria in a lifecycle economy with three homogeneous groups of voters (retirees, skilled young workers, unskilled young workers) have three properties. One, if income inequality is sufficiently high, unskilled workers and retirees will form a dominant coalition which raises both intragenerational and intergenerational transfers. Two, when capital is abundant relative to labor, government transfers will be strictly intergenerational. Three, all transfers increase when the voting franchise is extended to less affluent individuals.
Permanence of a Semi-Ratio-Dependent Predator-Prey System with Nonmonotonic Functional Response and Time Delay
Sufficient conditions for permanence of a semi-ratio-dependent predator-prey system with nonmonotonic functional response and time delay x˙1(t)=x1(t)[r1(t)−a11(t)x1(t−τ(t))−a12(t)x2(t)/(m2+x12(t))], x˙2(t)=x2(t)[r2(t)−a21(t)x2(t)/x1(t)], are obtained, where x1(t) and x2(t) stand for the density of the prey and the predator, respectively, and m≠0 is a constant. τ(t)≥0 stands for the time delays due to negative feedback of the prey population.
Existence and Uniqueness of Periodic Solutions of Mixed Monotone Functional Differential Equations
This paper deals with the existence and uniqueness of periodic solutions for the first-order functional differential equation y′(t)=−a(t)y(t)+f1(t,y(t−τ(t)))+f2(t,y(t−τ(t))) with periodic coefficients and delays. We choose the mixed monotone operator theory to approach
our problem because such methods, besides providing the usual existence results, may also sometimes provide uniqueness as well as additional numerical schemes for the computation of solutions.
Experimental EPR-Steering of Bell-local States. (arXiv:0909.0805v1 [quant-ph])
Entanglement is the defining feature of quantum mechanics, and understanding
the phenomenon is essential at the foundational level and for future progress
in quantum technology. The concept of steering was introduced in 1935 by
Schr"odinger as a generalization of the Einstein-Podolsky-Rosen (EPR) paradox.
Surprisingly, it has only recently been formalized as a quantum information
task with arbitrary bipartite states and measurements, for which the existence
of entanglement is necessary but not sufficient. Previous experiments in this
area have been restricted to the approach of Reid [PRA 40, 913], which followed
the original EPR argument in considering only two different measurement
settings per side. Here we implement more than two settings so as to be able to
demonstrate experimentally, for the first time, that EPR-steering occurs for
mixed entangled states that are Bell-local (that is, which cannot possibly
demonstrate Bell-nonlocality). Unlike the case of Bell inequalities, increasing
the number of measurement settings beyond two--we use up to six--dramatically
increases the robustness of the EPR-steering phenomenon to noise.
Quantum-circuit guide to optical and atomic interferometry. (arXiv:0909.0803v1 [quant-ph])
Atomic (qubit) and optical or microwave (modal) phase-estimation protocols
are placed on the same footing in terms of quantum-circuit diagrams. Circuit
equivalences are used to demonstrate the equivalence of protocols that achieve
the Heisenberg limit by employing entangled superpositions of Fock states, such
as N00N states. The key equivalences are those that disentangle a circuit so
that phase information is written exclusively on a mode or modes or on a qubit.
The Fock-state-superposition phase-estimation circuits are converted to use
entangled coherent-state superpositions; the resulting protocols are more
amenable to realization in the lab, particularly in a qubit/cavity setting at
microwave frequencies.
Chirped-pulse interferometry with finite frequency correlations. (arXiv:0909.0796v1 [quant-ph])
Chirped-pulse interferometry is a new interferometric technique encapsulating
the advantages of the quantum Hong-Ou-Mandel interferometer without the
drawbacks of using entangled photons. Both interferometers can exhibit
even-order dispersion cancellation which allows high resolution optical delay
measurements even in thick optical samples. In the present work, we show that
finite frequency correlations in chirped-pulse interferometry and
Hong-Ou-Mandel interferometry limit the degree of dispersion cancellation. Our
results are important considerations in designing practical devices based on
these technologies.
Classical analogues of two-photon quantum interference. (arXiv:0909.0792v1 [quant-ph])
Chirped-pulse interferometry (CPI) captures the metrological advantages of
quantum Hong-Ou-Mandel (HOM) interferometry in a completely classical system.
Modified HOM interferometers are the basis for a number of seminal
quantum-interference effects. Here, the corresponding modifications to CPI
allow for the first observation of classical analogues to the HOM peak and
quantum beating. They also allow a new classical technique for generating phase
super-resolution exhibiting a coherence length dramatically longer than that of
the laser light, analogous to increased two-photon coherence lengths in
entangled states.
"Quantum-optical coherence tomography" with classical light. (arXiv:0909.0791v1 [quant-ph])
Quantum-optical coherence tomography (Q-OCT) is an interferometric technique
for axial imaging offering several advantages over conventional methods.
Chirped-pulse interferometry (CPI) was recently demonstrated to exhibit all of
the benefits of the quantum interferometer upon which Q-OCT is based. Here we
use CPI to measure axial interferograms to profile a sample accruing the
important benefits of Q-OCT, including automatic dispersion cancellation, but
with 10 million times higher signal. Our technique solves the artifact problem
in Q-OCT and highlights the power of classical correlation in optical imaging.
Experimental violation of Svetlichny's inequality. (arXiv:0909.0789v1 [quant-ph])
It is well known that quantum mechanics is incompatible with local realistic
theories. Svetlichny showed, through the development of a Bell-like inequality,
that quantum mechanics is also incompatible with a restricted class of nonlocal
realistic theories for three particles where any two-body nonlocal correlations
are allowed [Phys. Rev. D 35, 3066 (1987)]. In the present work, we
experimentally generate three-photon GHZ states to test Svetlichny's
inequality. Our states are fully characterized by quantum state tomography
using an overcomplete set of measurements and have a fidelity of (84+/-1)% with
the target state. We measure a convincing, 3.6 std., violation of Svetlichny's
inequality and rule out this class of restricted nonlocal realistic models.
Topological Order Following a Quantum Quench. (arXiv:0909.0752v1 [quant-ph])
We determine the conditions under which topological order survives a rapid
quench. Specifically, we consider the case where a quantum spin system is
prepared in the ground state of the Toric Code Model and, after the quench, it
evolves with a Hamiltonian that does not support topological order. We provide
analytical results supported by numerical evidence for a variety of quench
Hamiltonians. The robustness of topological order under non-equilibrium
situations is tested by studying the topological entropy and a novel dynamical
measure, which makes use of the Ulhmann fidelity between partial density
matrices obtained from different topological sectors.
A new exoplanet to test tide theories
Almost 400 extrasolar planets have been found to date (see Physics Today, May 2009, page 46), but a new planet reported by Coel Hellier (Keele University) and colleagues stands out. Like many exoplanets, theirs, dubbed WASP-18b, is massive (10 times the mass of Jupiter) and has a small orbital radius (only 1/50th of Earth's). But its orbital period of only 0.94 day is the shortest for any "hot Jupiter" yet observed. Moreover, its large mass and small orbit are predicted to cause the strongest tidal interactions of any known star–planet system. According to current theory, the tidal bulge that the planet raises on its host star exerts a torque that will drain angular momentum from the planet and cause it to spiral inward. (For more on tidal interactions, see Physics Today, August 2009, page 11.) If the star's tidal dissipation rate is comparable to what's been measured for binary stars and for the gas giants in our own solar system, the infall will be quick: WASP-18b has less than a million years left in a lifetime, estimated from the age of its host star, of about a billion years. Over the next decade, WASP-18b's death spiral should produce a measurable shift in the planet's observed transit time. The absence of tidal decay—a notable possibility, given the rarity of finding a planet so close to the end of its life—would constitute direct evidence for a different class of tidal interactions in the host star and provide new constraints on models of stellar interiors. (C. Hellier et al., Nature 460, 1098, 2009.)—Richard J. Fitzgerald
High-Dimensional Non-Linear Variable Selection through Hierarchical Kernel Learning. (arXiv:0909.0844v1 [cs.LG])
We consider the problem of high-dimensional non-linear variable selection for
supervised learning. Our approach is based on performing linear selection among
exponentially many appropriately defined positive definite kernels that
characterize non-linear interactions between the original variables. To select
efficiently from these many kernels, we use the natural hierarchical structure
of the problem to extend the multiple kernel learning framework to kernels that
can be embedded in a directed acyclic graph; we show that it is then possible
to perform kernel selection through a graph-adapted sparsity-inducing norm, in
polynomial time in the number of selected kernels. Moreover, we study the
consistency of variable selection in high-dimensional settings, showing that
under certain assumptions, our regularization framework allows a number of
irrelevant variables which is exponential in the number of observations. Our
simulations on synthetic datasets and datasets from the UCI repository show
state-of-the-art predictive performance for non-linear regression problems.
Ternary Codes Associated with O^-(2n,q) and Power Moments of Kloosterman Sums with Square Arguments. (arXiv:0909.0811v1 [math.NT])
In this paper, we construct three ternary linear codes associated with the
orthogonal group O^-(2,q) and the special orthogonal groups SO^-(2,q) and
SO^-(4,q). Here q is a power of three. Then we obtain recursive formulas for
the power moments of Kloosterman sums with square arguments and for the even
power moments of those in terms of the frequencies of weights in the codes.
This is done via Pless power moment identity and by utilizing the explicit
expressions of "Gauss sums" for the orthogonal and special orthogonal groups
O^-(2n,q) and SO^-(2n,q).
An Infinite Family of Recursive Formulas Generating Power Moments of Kloosterman Sums with Trace One Arguments: O(2n+1,2^r) Case. (arXiv:0909.0809v1 [math.NT])
In this paper, we construct an infinite family of binary linear codes
associated with double cosets with respect to certain maximal parabolic
subgroup of the orthogonal group O(2n+1,q). Here q is a power of two. Then we
obtain an infinite family of recursive formulas generating the odd power
moments of Kloosterman sums with trace one arguments in terms of the
frequencies of weights in the codes associated with those double cosets in
O(2n+1,q) and in the codes associated with similar double cosets in the
symplectic group Sp(2n,q). This is done via Pless power moment identity and by
utilizing the explicit expressions of exponential sums over those double cosets
related to the evaluations of "Gauss sums" for the orthogonal group O(2n+1,q).
SKorean TV giants tout differing technologies
The world's top two makers of flat-panel televisions are stressing the energy-saving virtues of different display technologies in their race to dominate a huge global market.
West Coast fishermen embark on new wave of fishing
(AP) -- The West Coast groundfish fleet has struggled to stay afloat during major cutbacks to reverse long-standing problems with overfishing and to protect the seafloor from damage caused by bottom trawling gear.
Apple's iPhone launch in China no easy task: experts
Apple's iPhone will soon officially go on sale in China, more than two years after its US debut, but it may not make much of a splash, with smuggled units and similar devices available, analysts say.
Researchers restore missing protein in rare genetic brain disorder
UCSF researchers have successfully used protease inhibitors to restore to normal levels a key protein involved in early brain development. Reduced levels of that protein have been shown to cause the rare brain disorder lissencephaly, which is characterized by brain malformations, seizures, severe mental retardation and very early death in human infants.
Two more genetic risk factors for Alzheimer's disease found
An international team of scientists has identified two more genetic risk factors for Alzheimer's disease. The findings are reported in the online edition of the journal Nature Genetics.
Scientists discover new genetic variation that contributes to diabetes
Scientists have identified a genetic variation in people with type 2 diabetes that affects how the body's muscle cells respond to the hormone insulin, in a new study published today in Nature Genetics. The researchers, from Imperial College London and other international institutions, say the findings highlight a new target for scientists developing treatments for diabetes.
Ozone: Climate change boosts ultraviolet risk for high latitudes
Climate change will disrupt Earth's precious ozone layer, boosting ultraviolet (UV) radiation in the deep southern hemisphere and reducing UV in far northern latitudes, a study warned on Sunday.
Suscribirse a:
Entradas (Atom)