$PEARL is still the fastest growing POW network in history
Once FP8 goes live, I expect this pink line to go vertical
Friendly reminder:
$ZEC in Jan 2018:
• ~3.0M supply, minting 7,200 ZEC/day
• ~86% annualized inflation
• only ~14% of the 21M cap existed
It still ran to ~$880.
@crypto_bobby A man buys egg futures. The price rises, so he buys more. It keeps rising, and he keeps buying, until he’s sitting on an enormous position and a huge paper profit.
He calls his broker: “Right, let’s sell everything.”
The broker replies: “Sell to whom? You’re the egg man!”
> Even Pearl's own documentation highlights overhead costs at 4-15%.
Peryn is seeing below 1% depending on how you define it. Lots of different job types exist.
And the separate but real thing is no factory or production line is ever used 100%.
So miners have a choice to 1/ mine during downtime 2/ mine during batch jobs where net revenue would work out better and 3/ mine during full priced work IF the PoUW can sidecar inside the GPU without diminishing inference.
It's likely going to be a combination of all three. Number 3 is the most interesting and it depends if the work can stack inside the GPU without lowering throughput. It likely can in some instances.
And yes I'm being a little snarky too because "negative overhead" means more throughput as a result of implementing the Pearl SASS.
Best case I think implementing the Pearl SASS makes inference faster anyway, and then adding on mining has no impact on throughput after that because it uses different parts of the GPU.
RFQ is coming soon
The first integration will be with Silhouette (@silhouette_ex) to allow for RFQ access for traders on Hyperliquid (@HyperliquidX)
The algorithmic trading engine (VWAP/TWAP/etc) will utilize RFQ to execute blocks if the price is preferable relative to the liquidity on the limit order book
Traders will also have the ability to manually RFQ
This opens the door for future RFQ platforms to be integrated to @tread_fi
1. Honestly wild that Pearl already has an equivalent energy footprint as Zcash. Bitcoin next.
2. A common misconception with Pearl is that price needs to constantly increase for hashrate to grow. I.e. "hashrate follows price". This is true for traditional PoW networks like Bitcoin or Zcash, because higher prices are required to incentivize the continued buildout of their ASIC hardware base. But this is not true for Pearl, whose hardware base will continue to grow due to accelerating AI demand. These GPUs can mine Pearl for extra margin regardless of price.
Hashrate will continue to grow in all conditions
Pearl looks like it has saturated all the int7 hashrate it can find. FP8 has about 400x as much energy available.
Once FP8 launches we will see if Pearl can flip Bitcoin in gigawatts (roughly a 76x from here).
The rest, including any new data centers built, is pure energetic TAM.
At least a 25pct chance Pearl flips Bitcoin in energy in the next few years IMO. On top of that, a high chance Pearl clones key features from other blockchains.
Execution is a risk. Consensus changes are risks.
What if proof of work didn’t have to be wasteful?
Join us this Tuesday (11/17) as we host the @prlnet research team for a conversation on Proof of Useful Work.
Open to researchers, students, and community members! Food and drinks will be provided.
RSVP: https://t.co/bCHAh9MnGL