Skip to main content
Information Security

Return to Answer

added 182 characters in body
Source Link
Dewi Morgan
Dewi Morgan
  • 1.4k
  • 8
  • 15

The important thing about information security is that the idea is to think laterally before your foe does.

So, use that CPU time to mine bitcoins, or whatever makes money using a CPU. Do whatever it is, enough that you can get another CPU in 18 months. Buy that other CPU. Rinse, repeat.

Let's go with the 2^235 CPU-seconds from another post. That's 2^209 18-month Moore's law cycles (not a safe assumption that the law holds that long, but people have been claiming the law was dead since the 90s and it's still going, so let's go with it...), then in 23518 months, we'd crack that in a CPU-second. But we don't need to wait that long. There's a concept called the "Moore breakeven point", where the time-saving advantage of waiting another 18 months is less than 18 months. I think that's only 21018 months away.

So, overall, that's just 315 years.

But that thing we're doing with the bitcoin mining. Don't buy another CPU with it. Don't even buy the first computer. Save that $1,000 or so. Instead, just invest it in something that gives you 5% over the current interest rate.

At the end of those 315 years, after investing $1000 at 5% annual, compounded monthly, you'll have the equivalent of... $4.7 billion. Woah! That's enough to buy 4,700,000 PCs! If you also invest the money you're saving in electricity by not running a desktop, about $20/month, you end up with $27Bn. So this is another, smaller log factor to take into account, as well as Moore's law: at what point, investing your initial money, plus $20/month, is it worth cashing out and buying the best network of computers you can, to crack the hash fastest?

The important thing about information security is that the idea is to think laterally before your foe does.

So, use that CPU time to mine bitcoins, or whatever makes money using a CPU. Do whatever it is, enough that you can get another CPU in 18 months. Buy that other CPU. Rinse, repeat.

Let's go with the 2^235 CPU-seconds from another post. That's 2^209 18-month Moore's law cycles (not a safe assumption that the law holds that long, but people have been claiming the law was dead since the 90s and it's still going, so let's go with it...), then in 23518 months, we'd crack that in a CPU-second. But we don't need to wait that long. There's a concept called the "Moore breakeven point", where the time-saving advantage of waiting another 18 months is less than 18 months. I think that's only 21018 months away.

So, overall, that's just 315 years.

But that thing we're doing with the bitcoin mining. Don't buy another CPU with it. Don't even buy the first computer. Save that $1,000 or so. Instead, just invest it in something that gives you 5% over the current interest rate.

At the end of those 315 years, after investing $1000 at 5%, you'll have the equivalent of... $4.7 billion. Woah! That's enough to buy 4,700,000 PCs! So this is another log factor to take into account, as well as Moore's law: at what point, investing your initial money, is it worth cashing out and buying the best network of computers you can, to crack the hash?

The important thing about information security is to think laterally before your foe does.

So, use that CPU time to mine bitcoins, or whatever makes money using a CPU. Do whatever it is, enough that you can get another CPU in 18 months. Buy that other CPU. Rinse, repeat.

Let's go with the 2^235 CPU-seconds from another post. That's 2^209 18-month Moore's law cycles (not a safe assumption that the law holds that long, but people have been claiming the law was dead since the 90s and it's still going, so let's go with it...), then in 23518 months, we'd crack that in a CPU-second. But we don't need to wait that long. There's a concept called the "Moore breakeven point", where the time-saving advantage of waiting another 18 months is less than 18 months. I think that's only 21018 months away.

So, overall, that's just 315 years.

But that thing we're doing with the bitcoin mining. Don't buy another CPU with it. Don't even buy the first computer. Save that $1,000 or so. Instead, just invest it in something that gives you 5% over the current interest rate.

At the end of those 315 years, after investing $1000 at 5% annual, compounded monthly, you'll have the equivalent of... $4.7 billion. Woah! That's enough to buy 4,700,000 PCs! If you also invest the money you're saving in electricity by not running a desktop, about $20/month, you end up with $27Bn. So this is another, smaller log factor to take into account, as well as Moore's law: at what point, investing your initial money, plus $20/month, is it worth cashing out and buying the best network of computers you can, to crack the hash fastest?

Source Link
Dewi Morgan
Dewi Morgan
  • 1.4k
  • 8
  • 15

The important thing about information security is that the idea is to think laterally before your foe does.

So, use that CPU time to mine bitcoins, or whatever makes money using a CPU. Do whatever it is, enough that you can get another CPU in 18 months. Buy that other CPU. Rinse, repeat.

Let's go with the 2^235 CPU-seconds from another post. That's 2^209 18-month Moore's law cycles (not a safe assumption that the law holds that long, but people have been claiming the law was dead since the 90s and it's still going, so let's go with it...), then in 23518 months, we'd crack that in a CPU-second. But we don't need to wait that long. There's a concept called the "Moore breakeven point", where the time-saving advantage of waiting another 18 months is less than 18 months. I think that's only 21018 months away.

So, overall, that's just 315 years.

But that thing we're doing with the bitcoin mining. Don't buy another CPU with it. Don't even buy the first computer. Save that $1,000 or so. Instead, just invest it in something that gives you 5% over the current interest rate.

At the end of those 315 years, after investing $1000 at 5%, you'll have the equivalent of... $4.7 billion. Woah! That's enough to buy 4,700,000 PCs! So this is another log factor to take into account, as well as Moore's law: at what point, investing your initial money, is it worth cashing out and buying the best network of computers you can, to crack the hash?