Terabyte Blocks on Bitcoin Cash


“50 transactions per day, per human, for 10 billion humans. This is, I believe, the end game for Bitcoin Cash.” So says an energetic Joannes Vermorel, an engineer and founder of a supply chains company, Lokad.

“Supply chains are powered by commerce. It’s only because we have free commerce that supply chains actually work. So, my interest, is that if I can make Bitcoin Cash a reality, there would be even more commerce and even more complicated supply chains for my business to optimize.”

Vermorel thus concludes in a pronounced French accent after he introduced the maths of how Terabyte blocks can be a reality.

He said he put together the numbers of how to build a mining rig that can handle full terabyte blocks for 20 years. That includes bandwidth, storage costs, electricity, the hardware itself, and all that is needed.

The price tag, somewhat surprisingly, comes at just $25.6 million per node to handle seven million transactions per second.

“It may sound a lot, but… just to give you an idea, a commercial aircraft is $100 million, and people don’t think oh aircrafts are so expensive nobody can afford to fly. I mean, who came with an aircraft today” to the conference that took place in Japan, he rhetorically asked.

“The key idea is that an aircraft can serve the needs of many. Like a bridge, it’s expensive to build a bridge, but many people can use a bridge for a very, very long time. Being expensive is not a problem as long as everybody does not require his own aircraft,” he added.

The interesting part here is that he goes through optimizations to his initial calculations that reduce the cost by a factor of more than 20.

I realized looking back at the estimates that it is very hasty, it is not optimized at all, and actually we can do a lot better, he said.

So taking first bandwidth, which he estimated at 50 gigabyte per second costing $6 million for 20 years.

“If Bitcoin Cash becomes the world currency, this is not the retail price we are talking about, this is the big guys’ price, the price for bandwidth that Amazon and Google are actually paying…

Just to give you a price point number, Facebook has invested $400 million to build a trans-atlantic cable that will bring them 120 terabytes of bandwidth. So if we just linearly scale the price, Facebook, for the same bandwidth, would pay only $20,000.”

Bitcoin Cash would have the bargaining power if it was processing seven million transactions per second, Vermorel said, so taking off around $6 million from his estimate, bringing it to $20 million for 20 years.

Trans-atlantic cable is the worst for bandwidth, he said, anywhere else it would be cheaper. So I took the worst case, he argues.

Looking at the future, “I believe bandwidth is the most solved problem for bitcoin, literally. Two years ago, a lab in Netherlands managed to have a 1 kilometer 255 terabytes per second of a single strand of optical glass fibre. That’s two orders of magnitudes better than what Facebook is getting right now.

By the time Bitcoin Cash becomes a world currency, this bandwidth problem is going to become like nothing, I mean literally nothing,” the engineer told the dissenters conference.

Looking at CPU, he said, I calculated each transaction takes 250 microseconds, but now the software backing bitcoin has improved, so it’s down to 100 microseconds. So we don’t need as many CPUs anymore, he said. Too bad for Intel. Thus dropping the node’s price tag to $17 million.

Looking at the future, he says, if you want to be serious about computing you need Application-Specific Integrated Circuits (ASICs). In the two bitcoins specifically, what matters is the cryptographic ECDSA’s hashing of signatures, so you develop an ASICs to specialize in validating those signatures.

“I am not the first person to think of it. There is a lab that has a working prototype to do just that, and the performance is just astounding.

I mean we are talking of here, a quarter of a square millimeter, with a 40 nanometer process, vanishingly small consumption, which is just 46 milliwatt, and you can validate transactions in 1.5 microsecond.”

One card, a GPU like card, would be sufficient to validate 7 million transactions per second, he said. And this is not the future, it’s just not mass produced, that’s it.

Addressing data-storage, he said one terabyte per block, for 20 years, means one exabyte. That’s a lot of data to store. But bitcoin doesn’t need the whole blockchain, it only needs the UTXO set. So we need much less storage and we can save over $12 million of hardware equipment. Bringing the number down to just $4 million.

Vermorel, however, is not yet done. UTXO management optimizations bring his initial servers count of around 250 to keep it all running, down to just around 30. Costs too down to just $1 million.

That translates to just $50,000 a year right now, he says. Less than the cost of one employee. And that would give you the ability to process seven million transactions per second. Power the world’s money, commerce.

Terabyte. It’s difficult for the mind to see it, and quick is our urge towards caution. Decentralization is vital, not merely desirable. And just how vital it is, we have ourselves seen on August the 1st.

There is no liberty unless one can fork, and if all media, all corporations, the entire world seems to be against you as it was for many BCH supporters prior to August 2017, a coder must still be able to just chain-split fork because that is the only guarantor of freedom in the blockchain space.

Many can say plenty, and dogs can keep on barking, but no one cares and the caravan moves on because now, with the chain-split fork, you have the freedom of choice that escapes any real coercion.

That must be preserved if we are to leave our children what we enjoy, but what Vermorel is doing here, at what can be called nothing else than the dissenters’ conference, is what Nakamoto did in 2008.

He is outright refuting this suggestion that public blockchains can not scale by giving the maths and the facts to show how today they can handle one terabyte full blocks every ten minutes, let alone 10Mb.

Or far more ludicrously, just the 4Mb that was needed for bitcoin to provide an excellent experience to the many new users that only tried it for the first time last year.

Public blockchains can scale to 7 million transactions per second today, he said. Bitcoin can scale to Visa levels today, Nakamoto said in 2008.

Impossible, the mind insists. And with that one word adding fuel to the desire of engineers to prove just how it is true. Just how they’ll make it true. For it is.

When one glimpses a bit of reality. When one sees what can be seen. No is the answer of only fools.



Notify of