It is important for computer professionals to be able to forecast future technology performance and pricing. In order to fulfill future purchasing and support requirements, the IT professional should be able to analyze historical data and arrive at an approximated, though accurate, figure.
Though it is beyond the scope of this paper to find an elegant solution to the problem, I will examine if it is feasible using the process of averaging the average biennial difference (aBD) and linear slope value (sV) of historical data rather than a complex statistical model.
Method
To forecast current prices and performance of various computer hardware components, I will use historical data from O’Brien & Marakas (2007) and attempt to closely approximate current prices and technological performance growth by finding an average of the aBD and sV based on this data. I will then check current prices and compare my results with a published price list available on the internet.
Design and Procedure
This experiment will be isolated to the growth in performance and pricing as it relates to computer processing units (CPUs) (see Figure 1), random-access memory (RAM) chips (see Figure 2) and hard-disk drives (HDDs) (see Figure 3) over a 15 year period from 1991 to 2005. With biennial data, I will find the sV of each set, the aBD of the values of each set, then average the two. I will repeat this process for each component for performance growth and cost.
To find the current prices, I will use Newegg, Inc. (n.d.), known for reliable and consistent pricing, and compare the results of the historical data analysis.
Results
Computer Processors
The performance growth analysis of CPUs shows a sV of 263.01 MHz with an aBD of 269.64 MHz, a variation of 6.63 MHz with an overall average of 266.32 MHz. The cost analysis shows a sV of $23.28 with an aBD of $26.36, a variation of $3.08 with an overall average of $24.82. Continuing this trend shows that a CPU in 2009 should have a performance of 4.8 GHz with a cost of $626.73.
The typical performance of a CPU today is 2.79 GHz with an average cost of $244.29. The price range per processor is $59.99 – 1,039.99.
A CPU with a performance speed of 4.8 GB today is currently in development and unavailable for price comparison. Though the accumulated speed of multiple core processors can achieve this speed, it is outside the scope of this paper.
Random-access Memory
The performance growth analysis of RAM chips show a sV of 103.2 MB with an aBD of 142.79 MB, a variation of 39.59 MB with an overall average of 122.99 MB. The cost analysis shows a sV of $1.14 with an aBD of $6.71, a variation of $5.58 with an overall average of $3.93. Continuing this trend shows that a RAM chip in 2009 should have a performance of 2,214.84 MB with a cost of $125.66.
The typical performance of a single RAM chip today is 2 GB with an average cost of $33.20. The price range per chip is $22.00 – 56.00.
Hard-disk Drive Storage
The performance growth analysis of HDDs show a sV of 18.82 GB with an aBD of 22.85 GB, a variation of 4.03 GB with an overall average of 20.84 GB. The cost analysis shows a sV of $-24.62 with an aBD of $-26.07, a variation of $1.45 with an overall average of $-25.35. Continuing this trend shows that a HDD in 2009 should have a performance of 375.17 GB with a cost of $23.79.
The performance of a HDD today ranges from 18 GB to 2 TB, though typically 500 GB with an average cost of $91.29 (which is also typical of drives up to 1.5 TB). The price range per 500 GB drive is $49.99 – 379.99.
A HDD with a performance capacity of 375 GB today costs approximately $50.00.
Discussion
Though some of the results approximate the current technology and pricing, the data set used is too small to draw any meaningful conclusions at this time. Differences in device stability, manufacturing technologies and branding complicate this issue even further. More research needs to be done to find any meaningful correlation between historical pricing and technological performance growth with that of the future. The use of larger data sets analyzed with statistical methods may prove useful, but at this time, the use of an average of the sV of each set and the aBD of the set values does not provide any realistic outlook on technology in the future.
Newegg, Inc. (n.d.). Newegg.com online store. Retrieved June 5, 2009 from http://www.newegg.com/Store/Computer.aspx?name=Computer-Hardware.
O’Brien, J. A., & Marakas, G. M. (2007). Computer hardware. In S. Mattson, S. Isenberg & T. Hauger (Eds.), Introduction to information systems (14th ed.) (p. 109). New York, NY: McGraw-Hill/Irwin.
