a perspective on the future of massively parallel computing presented by: cerise wuthrich june 23,...
DESCRIPTION
Outline Intro & Background of Current Models –Limits of Sequential Models –Tightly Coupled MP –Loosely Coupled DS Fine-Grain Parallel Models –ANN –Cellular Automata Fine-Grain vs Coarse-Grain –Architecture –Functions –Potential Advantages Summary and ConclusionsTRANSCRIPT
A Perspective on the Future of Massively Parallel ComputingPresented by: Cerise WuthrichJune 23, 2005
A Perspective on the Future of Massively Parallel Computing: Fine-Grain vs. Coarse-Grain Parallel Models
Predrag T. TosicProceedings of the 1st Conference on Computing FrontiersApril [email protected]
Outline Intro & Background of Current Models
– Limits of Sequential Models– Tightly Coupled MP– Loosely Coupled DS
Fine-Grain Parallel Models– ANN– Cellular Automata
Fine-Grain vs Coarse-Grain– Architecture– Functions– Potential Advantages
Summary and Conclusions
Introduction and Background of Current Models Hardware limitations There are physical limits to how fast we can
compute– Limits to increasing the densities and
decreasing the size of basic microcomponents– No signal can propagate faster than the speed of
light
Introduction and Background of Current Models Limitations of Von Neuman Model
– There is a clear distinction (physical and logical) where data and programs are stored (memory) and where the computation is executed (processor)
– Sequential
Parallel Processing Realization that parallel processing was a
necessity Classical Models
– Multiprocessing Supercomputers– Networked Distributed Systems– Both models are actually coarse-grain
Proposal– “Truly fine-grain connectionist massively
parallel model”
Characteristics of Multiprocessing Supercomputers Communication Medium
– Shared, Distributed, Hybrid Nature of Memory Access
– Uniform vs. NUMA Granularity Instruction Streams (single or multiple) Data Streams (single or multiple)
Characteristics of Distributed Systems Loosely Coupled Heterogeneous collection Networked by middleware Scalable Flexible Energy dissipation not an issue Harder to program, control, detect errors
and failures
The model we should really consider Current supercomputers use thousands of
processors Current DS (like WWW) can use hundreds
of millions of computers We shouldn’t base parallel computing on
CS or engineering principles Instead look at the most sophisticated IP
device engineered – the human brain
Human Brain Tens of billions (1010) of processors
(neurons) Highly interconnected with 1015
interconnections Each neuron is a very simple basic
information processing unit
Artificial Neural Networks Best known and most studied class of a
connectionist model 1942 – Linear Perceptron Multi-Layer Perceptron Radial Basis Function NN Hopfield NN
Neural Network Diagrams
http://www.nd.com/neurosolutions/products/ns/nnandnsvideo.html
Artificial Neural Networks Each neuron (processor) computes a single
pre-determined function of its inputs Neuron similar to a logical gate Neurons connected with synapses Each synapse stores about 10 bits of info Each synapse fired about 10 times/sec Receptors are input devices Effectors are output devices
Artificial Neural Networks Just as the brain grows, changes, and
adapts, ANNs allow for – creation of new synapses– Dynamic modification of already existing
synapses ANNs – Memory
– No separate place for memory– All info stored in nodes and edges – Dynamic changes in edge weights
Cellular Automata The state of a cell at a
given time depends only on its own state one time step previously, and the states of its nearby neighbors at the previous time step. All cells on the lattice are updated synchronously.
Another Connectionist Model
Cellular Automata Model inspired by physics Grid where each node is a Finite State
Machine– Edge-labeled directed graphs– Each vertex represents one of n states– Each edge a transition from one state to the
other on receipt of the alphabet symbol that labels the edge
Cellular Automata Only 2 possible states
– 0 is quiescent– 1 is active
Only input is current states of its neighbors All nodes execute in unison A one-dimensional infinite CA is a
“countably infinite set of nodes capable of universal computation”
Connectionist Models Appear to be a legitimate model of a
universal massively parallel computer ANNs are suitable for learning, but not
Cellular Automata CA find most of their applications in
studying paradigms of dynamics of complex systems
Coarse-Grain vs. Fine-Grain Architectures
Coarse Fine
# of Proc Thousands Billions
Type of proc Powerful, expensive,
dissipate energy
Simple, cheap, energy efficient
Capabilities Complex Single, predefined function
Memory Separated from processor
Virtually no distinction between memory and
processor
Coarse-Grain vs. Fine-Grain Functions At the very core level, connectionist models
are different in how they:– Receive information– Process information– Store information
Limitations of ANNs ANNs aren’t necessary in all domains
– ANN can’t computer more or faster than the human brain
– The power of a human brain is an asymptotic upper bound on a connectionist ANN model
– Not needed for:• Computation tasks• Searching large databases
Suitable domains for ANNs Pattern Recognition
– “No computer can get anywhere close to the speed and accuracy with which humans recognize and distinguish between, for example, different human faces or other similar context-sensitive, highly structured visual images.”
Problem domains where computing agent has on-going, dynamic interaction with environment or where computations may have fuzzy components
Potential Advantages of Connectionist Fine-Grain Models Scalability Avoid slow storage bottleneck since there is
no physically separated memory Flexibility (not necessary to re-wire or re-
program with additional components) Graceful Degradation – neurons keep dying
in our brains and yet we continue to function reasonably well
Potential Advantages of Connectionist Fine-Grain Models Robustness – If one component of a tightly
coupled supercomputer fails, the whole system can fall apart
Energy consumption – dissipate much less heat
Summary and Conclusions Connectionist models such as ANNs or CA
are capable of massively parallel information processing
They are legitimate candidates for an alternative approach to the design of highly parallel computers of the future
These models are conceptually, architecturally and functionally very different from traditional models
Summary and Conclusions Connectionist models are:
– Very fine-grained– Basic operations are much simpler– Several orders of magnitude more processors– Memory concept is totally different
Summary and Conclusions Connectionist models are:
– Yet to be built– Idea is in its infancy– Currently still too far-fetched an endeavor– Promising future as the underlying abstract
model of the general-purpose massively parallel computers of tomorrow
Questions