The Evolution of Data Storage: Punch Cards to the Cloud

-



This past weekend, my mom found herself in a modern-day crisis: her emails were bouncing!

No new messages were coming in, and the culprit was a little notification reading, “Storage Full.”

Her immediate, reflexive solution? She wanted to go to the store to buy another flash drive.

I had to gently talk her off the ledge. I explained that she couldn't just plug a USB drive in and solve all her problems.

Hoarding physical drives in desk drawers is a relic of how we used to think about data.

But as we sat there clearing out gigabytes of old forwarded videos and blurry photos, it got me thinking...

My mom’s instinct to hold a physical object to store her digital life makes perfect sense when you look at how we got here. 

The human race has spent over a century trying to figure out where to put our memories.

Here is the story of how our data evolved from heavy, physical objects you could hold in your hand, to invisible streams in the cloud. 


To put the timeline in perspective, think about a standard 3-Megabyte photo on your phone today. Here is how many of those photos each major breakthrough could hold:


1890: Punch Cards | 80 Bytes (~0.00003 of a photo)



In the beginning, data was literal, physical space. You stored information by punching holes in stiff pieces of paper. The data type? Simple numbers and text for census counts and early computation. It would take tens of thousands of these cards just to store the data of a single modern photo. And the biggest rule of early data storage? You can't unpunch a hole. Once the data was there, it was permanent.


1932: Magnetic Drum Memory | 62.5 KB (~0.02 of a photo) 





We realized paper was too limiting, so we turned to magnetism. Magnetic drums were massive metal cylinders coated in magnetic material. They stored early executable code and numerical data. They were the size of a water heater, incredibly loud, and still couldn't hold even a fraction of a single picture.

1951: Magnetic Tape | 2 MB (~0.67 of a photo) 




Next came magnetic tape, which was a massive leap but had a fatal flaw: it was sequential and slow. If your data was at the end of the spool, you had to physically fast-forward through all the data before it to get what you needed. It was great for backing up text and audio, but terrible for quick retrieval.

1956: The Hard Disk Drive (HDD) | 5 MB (~1.6 photos) 



IBM introduced the RAMAC, and random access was born. You didn't have to rewind anymore; a mechanical arm could jump to exactly where the data lived. The very first hard drive could hold just over one modern smartphone photo. It was the size of two refrigerators.


1970s: Cassette Tapes & Floppy Disks | 1.44 MB (~0.5 of a photo)


Storage finally came to the living room. Cassette tapes stored early PC programs, but the Floppy Disk changed the world. Everyday people could carry word processing documents and simple games in their pockets. We were storing mere kilobytes, which wasn't enough for a high-res photo, but it felt like freedom.

1982: The Compact Disc (CD) | 650 MB (~216 photos) 



This was a monumental shift. We moved from magnetic storage (which could degrade, stretch, or be erased by a strong magnet) to optical storage. Lasers read microscopic pits on a reflective surface. Suddenly, we were storing high-fidelity audio, complex software, and entire albums of photos on a single disc.

1989: The Dawn of the Solid State Drive (SSD) | 20 MB (~6.6 photos) 



While early SSDs existed before the 90s, this era saw the commercial beginnings of flash memory—no moving parts, just electrons trapped in grids. Early consumer units held a few dozen megabytes. They were incredibly fast but initially reserved only for aerospace and high-end tech before scaling down to consumers.

1994: The Zip Drive | 100 MB (~33 photos) 



A nostalgic staple of the 90s. When floppy disks were too small for graphic design files and early multimedia, the Zip Drive offered magnetic storage with some serious breathing room. It was the heavy-duty workhorse for creatives before optical drives became standard in every PC.

1995: Digital Versatile Disc (DVD) | 4.7 GB (~1,566 photos) 


Optical storage leveled up. The lasers got tighter, and the storage leaped into gigabytes. We went from storing text and audio to storing full-length, standard-definition movies and massive encyclopedias of data on a single plastic disc.

1999: Secure Digital (SD) Card | 64 MB (~21 photos) 



As gadgets shrank, so did storage. Flash memory became consumer-friendly. The earliest SD cards allowed us to store a handful of digital photos and MP3s on a microchip the size of a postage stamp.

2000: Universal Serial Bus (USB) Drive | 8 MB (~2.6 photos)



Enter my mom’s favorite gadget: the flash drive. It killed the floppy disk overnight. While the very first ones only held 8 Megabytes, they quickly scaled to gigabytes. You could eventually carry tens of thousands of documents, photos, and songs on your keychain.

2006: Blu-ray Disc | 25 GB (~8,333 photos)


The pinnacle of consumer optical media. By using a blue laser (which has a shorter wavelength than a CD/DVD's red laser), we could cram massive amounts of data onto a single disc. It was built for high-definition 1080p video and huge video game assets.

2006 to Present: Cloud Computing | Standard Capacity: 2 TB (~666,666 photos) 


With the launch of AWS in 2006, followed by the rise of Dropbox, Google Drive, and OneDrive, the era of holding your data ended. We stopped buying plastic discs and metal drives. Instead, our data lives on massive SSD and HDD server farms hundreds of miles away, delivered to us over the internet in milliseconds. Even a standard, basic consumer cloud plan today can hold over half a million photos.


The Economics of Data Storage

If you chart this timeline out, you'll notice a fascinating economic pattern. Every time we completely changed the state of our storage (moving from magnetic tape to optical CDs, or from spinning hard drives to flash-based SSDs) the cost per bit actually spiked at first.

Inventing a new medium requires massive research and development, and building new factories to manufacture lasers or silicon chips isn't cheap. The early adopters absorb those costs. But once the technology becomes standard, economies of scale kick in. Manufacturers figure out how to cram exponentially more data into the same physical footprint.

The price goes into a freefall, ultimately dropping far below whatever the previous generation's cheapest price was. This cycle of "expensive innovation followed by mass affordability" is exactly why cloud computing relies on massive, optimized server farms to give us the cheapest cost-per-bit solution the world has ever seen.

Which brings me back to my mom, sitting at the kitchen table, asking if a new flash drive would fix her email.

I explained to her that she doesn't need to hold her data anymore. Flash drives get lost in the laundry. Hard drives fail. CDs get scratched. 

I talked her out of buying more plastic and metal because the way we store data has evolved beyond what we can manually carry.

When I look at my own life, I realize I’m not just storing files; I’m storing my history and it's all weightless now. 

That is why, I cheerfully pay Google a few dollars every single month for storage for myself and now my family.

 

Resources




Comments

Post a Comment

Popular posts from this blog

File Types: Excel vs CSV

-
Image
They open in the same app. That’s where the similarities end. As someone who works with large amounts of data (and has made a lot of mistakes), I'm writing this for anyone still fuzzy on the fundamental differences between Excel (.xlsx) and Comma-Separated Values (.csv) files. If you’re expected to automate data flows , please don’t ignore this one. This post is for you! On the surface, these two look like identical twins: Tomayto.xlsx Tomahto.csv But if you look closely, you’ll notice the first difference. If you noticed that Tomayto.xlsx is 7× bigger in Size than Tomahto.csv —you’re right! And here’s the key detail: both files are EMPTY . So, what's going on? They both look the same to us, but the computer sees it differently. 
Read more

The Cloud: Why Does It Matter?

-
Image
  A floating cotton candy that brings rain, hail and snow. Since modern day technology is a relatively new invention, we tend to warp the meanings of words a lot.  For instance, a tablet used to be medicine, but now it's a portable computer with touch input.   Naming conventions also draw inspiration from nature, like "ecosystem," "virus," "bug," and "mouse." Today I would like to talk about the cloud as it is now a word that is being used outside of conversations about the weather.  I first learnt about the intricacies of clouds back in Geography class, from cirrus clouds to cumulonimbus clouds, as well as what they meant.   Spot a cumulonimbus cloud? Brace for a thunderstorm! (You're welcome!) Now, "cloud" metaphorically represents remote data storage and processing—ubiquitous, vast, and intangible, like an actual cloud. "So, Toni, what's so special about the cloud?" To answer, let's imagine a cloud-less 2023...
Read more

CIBC's Women in AI 2026: 3 Takeaways

-
Image
“The quality of your life is determined by the quality of the questions you ask yourself and others.”  — Tony Robbins Last Thursday, I attended CIBC’s Women in AI Event at UWI St. Augustine, and one of the most memorable parts of the experience was the Q&A session. The room was filled with thoughtful questions on topics such as data collection, AI history, performance metrics, and accountability. It was the kind of discussion that could have easily been its own event. While many of these topics were already familiar to me, it was refreshing to be in a space with women asking critical questions about the future of Artificial Intelligence. One question stood out to me most:  What are you doing to support the community when it comes to AI? That question stayed with me long after the event ended. Today, I’m using this blog post to reflect on that question and summarize three others raised during the session, blending key ideas shared in the room with my own perspective. F...
Read more

Communication : The Hardest Soft Skill

-
Image
  Two blobs talking to each other Hey there, tech enthusiasts!  Fun fact: I aced every subject in the Caribbean Advanced Proficiency Exam (CAPE) except one:  Communication Studies    ( even though I did it twice!) Now, before you jump to conclusions, let me explain how this happened: On my second attempt Question 1 could be interpreted in two different ways due to its confusing wording.  It even left the invigilator scratching their head in confusion. After that nerve-wracking experience, I swore I'd never have to worry about "Communication Studies" ever again.  However, little did I know that communication can be the ultimate secret weapon to make your tech job easier. Whether it's crafting emails or delivering project updates, an effective communicator will have the upper hand. You might think that having a degree in Computer Science is all it takes to succeed in your career. But let me tell you, I've found myself in situations where explaining ...
Read more

No Laptop November

-
Image
"Life is hard, but that's what makes it interesting" - Dad It's time to break the silence once more! Things did not go as I planned, but it makes for a good story.  So you're probably wondering... " Toni, what happened? I thought you were coming back? " and to that I would say, I thought the same. I thought that I would get back in my groove and post blog 109 weeks ago! My simple hardware failure quickly escalated into a complex and costly international shipping disaster. It all started on November 7th at about 5 AM while I was clacking away on my keyboard, the display for my newly purchased Zenbook S16 decided that it was time to say goodbye. I made every attempt to reilluminate it by doing the famous "turn-it-off-and turn-it-back-on" trick and confirmed that it was a hardware issue with Asus Helpdesk. I still held a warranty , which was a relief. However, I had to send the laptop to the States so the manufacturer could fix it and send it ba...
Read more

Data Wrangling : Best Practices For Working With Big Datasets

-
Image
  A matrix of dots that you did not bother to count Let's face it, working with exponentially expanding datasets can be both exciting and overwhelming. Imagine dealing with a dataset of 22 million rows - that's a lot of information to process!  The question is, can your ETL process handle it?  This is a problem that I faced this week, and I had to find a way to improve the performance of the process that updates a dashboard as it was taking a decade to update. The initial question that crossed my mind was, "What was the actual size of this dataset?" At first, I mistakenly assumed that the dataset contained between 1 and 6 million rows. A quick COUNT(*) query made it clear that my estimate was way off and also provided me with some clarity to the problem. That dataset was a behemoth, it probably had its own gravitational pull! The sad truth was that the process was not scalable, and it was clear that immediate improvements were necessary. Here are a three tips that I h...
Read more

FinTech : Avoiding Fines with Partial Automation

-
Image
A bitcoin spider (it does not exist) Fintech is an area of the financial industry that is constantly evolving and full of excitement and confusion. While you may enjoy the convenience of tapping your card on a point-of-sale terminal, your granny may struggle to adapt to the bank's new app. We can all agree that technology is infiltrating every sector and has led to the emergence of cryptocurrency and autonomous vehicles.  However, with innovation comes the need for regulation, and companies in the United States are now subject to more than 50,000 regulations according to page 6 of  SVB's Breakdown on The State of Fintech . Failure to comply with these regulations can lead to severe penalties, damage to the company's reputation, and even legal action. Coinbase is a prime example of this, as they were fined $50 million for failing to build and maintain a compliance program that could scale with their rapid growth.  Today, risk management is still a highly manual task that ...
Read more

Web Scraping: An Introduction

-
Image
  Wolverine was here Web scraping has been used to compile extensive datasets on everything from UFO sightings to product reviews on the internet! If you have ever gone to a website, copied data from it, and pasted it into an Excel file, then congratulations, you have done web scraping manually. However, what if you had a bigger problem to solve and required more data from a website? For example, the wristband for your cheap Casio watch broke, and you decided it's time to upgrade to a more premium model.  However, before spending your money, you wanted to compare the prices, reviews and materials of thousands of watches to ensure that you are making the best decision when it comes to durability and your budget.  I hope we can both agree that this task is humanly impossible to get done within a reasonable amount of time, and that's where libraries like Scrapy and Beautiful Soup come into play. With great power comes great responsibility! Here are some pros and cons of web ...
Read more

Robots : What They Can and Cannot Do (Yet)

-
Image
Exterminate... Humans! Today's topic is inspired by Apple's recent pivot to home robots after their electric vehicle project failure.  Industrial scenes come to mind when I think robots: a series of them churning out everything from cars to perfectly peeled pineapples.  (Who really likes peeling those?) But robots are far more diverse than those seen in factory lines, they can be classified in various ways based on design, function, autonomy, and application.  Take agricultural drones, for example, these bots hover around trees, using vision technology to select the best fruits for harvest based on color. Additionally, humanoid robots are no longer science fiction! As a result, these advancements in technology have led to workers performing these repetitive tasks being relieved of their jobs. While new jobs are created, the question still stands: "Will there be a balance between job displacement and job creation?" Those who are for robots often say that we will work a...
Read more

October Reflections

-
Image
The name October comes from the Latin “octo,” meaning eight . This morning, my dad asked about my blog. I winced — I’ve been dodging it since September! But that question was the nudge I needed. Because when I stopped to think, I realized:  even without posting, I’d been living. I hiked (and fell multiple times). Tried lobster for the first time. Met my coworkers in Barbados. And impulsively bought this... (I only had 1/4) All that living gave me plenty to reflect on — and two moments in particular stood out. 1. Women in Data Science (WiDS) TT Spotlight Talk Dr. Letetia Addison invited me to speak at the Women in Data Science TT Spotlight Talk — for the second time!  She actually found me through this very blog. I wasn’t publishing much here, but I was active on LinkedIn.  When it came time to prepare my WiDS talk, I had weeks, but I started only days before. Procrastination kept winning. Then came a call with my friend Carson White, we brainstormed ways in whic...
Read more