Computer Technology in 2020

In 2010, I made some predictions about the state of computers in January 2020.  Let's see how I did:

  Prediction Reality Score
Storage capacity* 120 to 200 Tb for $150 6 Tb Wrong
Standard RAM** 500 to 750 Gb 16 Gb Wrong
Monitors*** 500 to 1000 dpi 182 dpi Wrong
Portable r/w media capacity 256-512 Gb 256-512 Gb Correct
Aggregate CPU speed**** 120-200 GHz 28.8 GHz Wrong

*Internal non-volatile read/write, random access, general storage of any form.

**Average RAM on newly purchased systems costing under $1,000.

***Any form of display.

****Number of cores times clock speed for top-of-the-line newly purchased consumer systems.

Source for prices:

Well, despite my last predictions being (mostly) spot-on, I did rather poorly this time. Before we examine why, let me explain how I've been doing my predictions. Moore's law basically says that the number of transistors that can be fit onto a given amount of silicon real-estate doubles every 18 months. It is thus a simple calculation to project out ten years. Likewise, storage capacity of hard disks has been increasing by 60% a year since the 1980s. For display resolution and portable media capacity, my predictions were simply educated guesses based on historical trends.

Many have been predicting the demise of Moore's law for years. But it is not yet defunct. For a long time, this was very accurate in terms of making predictions. So, what went wrong with my predictions for 2020? I mentioned in my 2010 predictions that I had concerns about the slow economy. The economy had been slow since 2001 (and, in fact, only just improved a few years ago). But even so, the slow economy hadn't affected the predictions for 2010 made in 2000. Therefore, although it was probably a factor, it doesn't really explain how wildly wrong I was.

So, what does explain it? In terms of CPU and RAM, technology did not advance at all as predicted by Moore's law. I think that it was a societal change. This last decade saw the cell phone come into its own - to the point that I believe desktop sales slid because many people used desktops primarily for email, surfing the net, and playing games. Phones do all of that now and also can be used as phones. Mobile platforms are battery-bound and thus power efficiency is far more important than raw processing power. Market demand drives technology, and so desktop technologies did not advance as quickly as before. On the other hand, mobile-related technologies have been quickly advancing. Moore's law is still valid - it simply moved to a new market. The numbers support this. Desktop sales have remained steady, or even dropped a little over the last decade. Mobile PCs, tablets, and smartphones each increased every year throughout the decade. Desktop-related technology improvements seem to be primarily driven by high-end gaming and data center server demand. This will make my predictions for 2030 a little more difficult.

Market and economy also explains, to some degree, the fact that storage did not increase as much as predicted. But the market here was more than merely a move to the mobile platforms - there was also a disruptive technology that took away some of the demand for hard disk drives: solid-state drives. In 2010, SSDs were expensive and low capacity. That has changed by 2020. People began preferring the smaller-capacity and more-expensive, but much faster, SSDs to hard disks. Futher, they have lower power requirements, thus making them appealing to some mobile devices (primarily tablets and laptops). I should note that SSD capacity increased along at the rate that hard disks had been increasing previously, but since they started behind they have yet to catch up with hard disks. I wouldn't be surprised if SSD capacity exceed hard disk capacity by 2030, although I imagine they will remain more expensive on a byte to byte comparison.

As far as display pixel density is concerned, I based my estimate on past performance. Even with the market changes, I would have expected the mobile market to drive the desktop display technology forward at the same rate, but it didn't. In the case of mobile devices however, I was on target. The Sony Xperia XZ Premium (released in 2017) has a pixel density of 806.9 pixels per inch. However, since I was referring to desktop montiors, I have to qualify my prediction as a failure. Why didn't this translate into desktop monitors? I'm not sure, but it probably has to do with flat demand for desktops. Additionally, it seems that the technology advances for desktop screens went into making larger, and flatter, displays. Further, although touch-screens have been around for a long time (the first one I saw was in the 1970s), they have finally become mainstream in the last decade. Sometimes technology follows the line of more features rather than more capacity.

Now, I will make some predictions for January 2030. I'm not going to make predictions related to the mobile market, because that will be driven largely by battery technology, and I have no handle on how that segment of the technology industry is advancing. Thus, I will be making predictions only for the desktop market. For the aforementioned reasons, I will abandon Moore's law and base my estimates on how the last decade has progressed.

  Prediction for 2030
Internal storage 18-20 Tb for $150
Standard RAM 64 Gb
Monitors 200 to 800 dpi
Portable r/w media capacity 512 Gb-1 Tb
Aggregate CPU speed 50-60 GHz

The wide range for monitors reflects the slow pace of the last decade (thus 200 dpi on the low end), but I have to imagine that at least some of the mobile market display improvements will make it over to the desktop market (hence the high range of 800 dpi).

Meet me back here in ten years and see how close I came!