A scientific large-scale application that performs weather pattern prediction is deployed on a 100-core machine i.e.,...
Fantastic news! We've Found the answer you've been seeking!
Question:
Transcribed Image Text:
A scientific large-scale application that performs weather pattern prediction is deployed on a 100-core machine i.e., a machine with 100 CPUs. To save power, some CPUs can be turned off by the system. In which case, the application is restricted to run only on the CPUs that are powered on (online). In a single run, the application performs the following sequence of operations. (1) it initializes its state which takes 5 seconds and cannot be parallelized; (2) it launches one prediction heuristic per each square-foot of covered area. These are kept independent from each other and thus can be executed in parallel, with each taking 1 second to complete. The total area covered by the weather prediction is 100 square feet in size. (3) It serializes the result obtained from each local prediction into a global prediction. This step takes 15 seconds and cannot be parallelized. Answer the following and motivate your answers. a) What is the speed-up of the entire application when it operates on a single CPU, compared to the case where it operates on all the available CPUs? b) How many CPUs we would need to keep online to ensure that we are able to achieve a weather prediction throughput of 2 predictions per minute? c) What is the weather prediction capacity of this systems? d) Does Amdahl's Law provide a good approximation to compute the speed-up of the system when 30 CPUs are turned on, compared to the case when only 1 CPU is online? e) Assume now that you have the option to turbo-boost only one CPU. Turbo-boosting one CPU makes that CPU twice as fast. When a CPU is turbo-boosted only a total of 50 CPUs can be kept online, while the other 50 must be powered off. Is turbo-boosting beneficial from a standpoint of throughput maximization? A scientific large-scale application that performs weather pattern prediction is deployed on a 100-core machine i.e., a machine with 100 CPUs. To save power, some CPUs can be turned off by the system. In which case, the application is restricted to run only on the CPUs that are powered on (online). In a single run, the application performs the following sequence of operations. (1) it initializes its state which takes 5 seconds and cannot be parallelized; (2) it launches one prediction heuristic per each square-foot of covered area. These are kept independent from each other and thus can be executed in parallel, with each taking 1 second to complete. The total area covered by the weather prediction is 100 square feet in size. (3) It serializes the result obtained from each local prediction into a global prediction. This step takes 15 seconds and cannot be parallelized. Answer the following and motivate your answers. a) What is the speed-up of the entire application when it operates on a single CPU, compared to the case where it operates on all the available CPUs? b) How many CPUs we would need to keep online to ensure that we are able to achieve a weather prediction throughput of 2 predictions per minute? c) What is the weather prediction capacity of this systems? d) Does Amdahl's Law provide a good approximation to compute the speed-up of the system when 30 CPUs are turned on, compared to the case when only 1 CPU is online? e) Assume now that you have the option to turbo-boost only one CPU. Turbo-boosting one CPU makes that CPU twice as fast. When a CPU is turbo-boosted only a total of 50 CPUs can be kept online, while the other 50 must be powered off. Is turbo-boosting beneficial from a standpoint of throughput maximization? A scientific large-scale application that performs weather pattern prediction is deployed on a 100-core machine i.e., a machine with 100 CPUs. To save power, some CPUs can be turned off by the system. In which case, the application is restricted to run only on the CPUs that are powered on (online). In a single run, the application performs the following sequence of operations. (1) it initializes its state which takes 5 seconds and cannot be parallelized; (2) it launches one prediction heuristic per each square-foot of covered area. These are kept independent from each other and thus can be executed in parallel, with each taking 1 second to complete. The total area covered by the weather prediction is 100 square feet in size. (3) It serializes the result obtained from each local prediction into a global prediction. This step takes 15 seconds and cannot be parallelized. Answer the following and motivate your answers. a) What is the speed-up of the entire application when it operates on a single CPU, compared to the case where it operates on all the available CPUs? b) How many CPUs we would need to keep online to ensure that we are able to achieve a weather prediction throughput of 2 predictions per minute? c) What is the weather prediction capacity of this systems? d) Does Amdahl's Law provide a good approximation to compute the speed-up of the system when 30 CPUs are turned on, compared to the case when only 1 CPU is online? e) Assume now that you have the option to turbo-boost only one CPU. Turbo-boosting one CPU makes that CPU twice as fast. When a CPU is turbo-boosted only a total of 50 CPUs can be kept online, while the other 50 must be powered off. Is turbo-boosting beneficial from a standpoint of throughput maximization? A scientific large-scale application that performs weather pattern prediction is deployed on a 100-core machine i.e., a machine with 100 CPUs. To save power, some CPUs can be turned off by the system. In which case, the application is restricted to run only on the CPUs that are powered on (online). In a single run, the application performs the following sequence of operations. (1) it initializes its state which takes 5 seconds and cannot be parallelized; (2) it launches one prediction heuristic per each square-foot of covered area. These are kept independent from each other and thus can be executed in parallel, with each taking 1 second to complete. The total area covered by the weather prediction is 100 square feet in size. (3) It serializes the result obtained from each local prediction into a global prediction. This step takes 15 seconds and cannot be parallelized. Answer the following and motivate your answers. a) What is the speed-up of the entire application when it operates on a single CPU, compared to the case where it operates on all the available CPUs? b) How many CPUs we would need to keep online to ensure that we are able to achieve a weather prediction throughput of 2 predictions per minute? c) What is the weather prediction capacity of this systems? d) Does Amdahl's Law provide a good approximation to compute the speed-up of the system when 30 CPUs are turned on, compared to the case when only 1 CPU is online? e) Assume now that you have the option to turbo-boost only one CPU. Turbo-boosting one CPU makes that CPU twice as fast. When a CPU is turbo-boosted only a total of 50 CPUs can be kept online, while the other 50 must be powered off. Is turbo-boosting beneficial from a standpoint of throughput maximization?
Expert Answer:
Related Book For
Foundations of Financial Management
ISBN: 978-1259024979
10th Canadian edition
Authors: Stanley Block, Geoffrey Hirt, Bartley Danielsen, Doug Short, Michael Perretta
Posted Date:
Students also viewed these programming questions
-
Managing Scope Changes Case Study Scope changes on a project can occur regardless of how well the project is planned or executed. Scope changes can be the result of something that was omitted during...
-
The Starbucks coffee shop on Sixth Avenue and Pine Street in downtown Seattle sits serene and orderly, as unremarkable as any other in the chain bought years ago by entrepreneur Howard Schultz. A few...
-
A bank reconciliation takes time and must balance. An employee was struggling in balancing the bank reconciliation. Her supervisor told her to plug (make an unsupported entry for) the difference,...
-
Why has the management of the environmental impacts of company operations become a major concern for businesses?
-
Which of the following is a co-requisite of a health savings account (HSA)?
-
Consider the following decomposition process: \[ A(\mathrm{~s}) ightleftharpoons B(\mathrm{~s})+C(\mathrm{~g}) \] The pressure at equilibrium is \(\ln P=12-\frac{3790}{T}\). Deduce the expression for...
-
Wiemers Products Company operates three divisions, each with its own manufacturing plant and marketing/sales force. The corporate headquarters and central accounting office are in Wiemers, and the...
-
Based on the following information, what is C/G/S for Moroni Industries. Inc. using FIFO under the periodic inventory method? Beginning Inventory 10 @ $120; Purchase 60 @ $112, Sale 40; Purchase 30 @...
-
Waylander Coatings Company purchased waterproofing equipment on January 6 for $1,031,600. The equipment was expected to have a useful life of four years, or 10,400 operating hours, and a residual...
-
E-max sold Dishwashers on credit, worth OMR 55,000 to customers, out of these Dishwashers customers return damaged Dishwashers for OMR 6,000. Record the entry for the Dishwashers returned? a. Debit...
-
Question 4: A system consisting of 1kg H0 undergoes the below processes: Process 1-2: Constant volume heating from P = 500 kPa, T =160 C to 1MPa Process 2-3: Constant pressure process to saturated...
-
4. You are required to design a composite in which the fiber efficiency will reach 90%. The fibers are carbon fibers with a tensile strength of 2,000 MPa and a diameter of 7m. In the first trial, you...
-
The planar mechanism know as Pantograph is shown in Figure 1. Pantograph has five links: the fixed link N (with zero length for the depicted system) and symmetric links A, D and B, C (lA = ID and lB...
-
Draw the Nyquist plot for the system with open loop transfer function, K s(s+2)(s+10) G(s)H(s)= a) Draw the s-plane map of the system. (2m) b) Draw the Nyquist plot of the system. Show the working...
-
4. Consider the Euler-Bernoulli beam model. Derive the distribution of the verti- cal normal stress oy(x, y) in an arbitrary longitudinal position x and the cross section is uniform rectangular one....
-
illustrate how E&P affects the character of the distribution to shareholders. Walk us through the example and explain the character of the distribution to each shareholder. What tax effect does the...
-
The Heese Restaurant Group manufactures the bags of frozen French fries used at its franchised restaurants. Last week, Heeses purchased and used 101,000 pounds of potatoes at a price of $ 0.70 per...
-
Assume you can buy a warrant for $5 that gives you the option to buy one share of common stock at $15 per share. The stock is currently selling at $18 per share. a. What is the intrinsic (minimum)...
-
Philips Rock and Mud is trying to determine the maximum amount of cash dividends it can pay this year. Assume its balance sheet is as follows: Assets Cash $ 312,500 Accounts receivable. 800,000...
-
Compare the balance based on IFRS with that based on ASPE, and describe limitations of the balance sheet information for financial analysis.
-
Several independent situations are described below. 1. The owner of the business included his personal dental expenses in the entitys income statement. 2. The company spent $25 000 on computer...
-
Gas Giant Australia Ltd signed a 20year deal to sell gas to a neighbouring country for \($1.8\) billion. The contract was signed on the last day of Gas Giant Australia Ltds financial year, 30 June....
-
Goode Medical Laboratory Ltd, GMLL, a medical research entity, has discovered a cure for a previously incurable disease. GMLL is protecting the drugs formula by keeping it secure in the company...
Study smarter with the SolutionInn App