Many have the perception that the established behemoths (OK, VISA) have unassailable banks of technological resource that make it impossible for ‘normal’ sized companies to compete. The reality, with tech as advanced as it is, is not nearly as complicated/expensive/unscalable as you might – and they would like – you to think.
Oh really, Charles Big Potato, show your cards then? OK, Yee Of Little Faith, here I go.
Without getting too Terabyte on you (because it’s dull), you may, understandably think you need mega, not to mention, fast storage and that’s expensive. Yes and no. Yes you need storage but, no, it doesn’t need to expensive, in fact, you can get it from standard ‘high street’ retailers like PCWorld or relatively cheaply online.
I know you didn’t bother looking at the links so let me quickly summarize for you: For $300, we can have a low-cost HDD at the 10TB level. To be roughly the like for like of VISA, we need annually, 365/10 HDDs. Let’s round it up (for redundancy) to 42 drives at a cost of $12,500 (USD). Ooh, not so crazily intimidating is it?
Still with us? Nearly there:
- Add servers and other costs, we have around $20,000 in total: Hence, the annual costs for a VISA challenging node is a server that costs $20,000
- Let’s scale that up: So, we have 1,000 nodes, this is a cost of $20 million per annum. VISA costs around $500 million per annum
Want to skip to the juicy bit?
In essence, Bitcoin can not only challenge VISA for under $20 million PA; at only $20,000 each node to a miner, it even makes money.
For the real techy:
…and that’s today. It actually gets better. If we take Moore’s law, we scale to double every 18 months. This is a growth factor of just over 100 times the computational power and storage every 10 years.
Thus, the cost decreases exponentially and we see that competing with VISA (per node) will cost:
- Oct 2018 – $10,000
- Apr 2020 – $5,000
- Oct 2021 – $2,500
- Apr 2023 – $1,250
- Oct 2024 – $625
So the costs to compete with VISA are within any company’s reach. Any user can be a miner.