First off, just wanted to give a shout out to whoever staked on my node around a week ago. (If the guy/gal is on this forum) Thanks so much for putting such a high stake on my node.
Before I get into the details, I’ll just briefly explain my background. I am a graduate of NYU Stern who worked in Portfolio Analytics at a Hedge Fund and Trading Group for a couple of years. I really wasn’t satisfied with the complexity of work and purpose of financial institutions, so after gradually setting up mining rigs, I left the industry to work on my operations full time. Many of my friends called me insane for leaving a high paying job, and perhaps rightly so. However, as somebody who grew up with an interest in technology very early on in my life, the value proposition that the blockchain space offered was undeniable. Luckily for me, my career decision seems to have been the right move. One of the most exciting things for me in regards to cryptocurrencies was when I found a security vulnerability on one of the services that Coinbase was involved with called Shift Card. Basically, I was able to withdraw unlimited amounts of cash from ATMs using the Shift debit card. I reported this to Coinbase via Hackerone where I received a bug bounty for the highest level. (Reward was 1 bitcoin which was equivalent to ~$10,250 at the time) Sadly, they are no longer in business.
Currently, I am running several Ethereum mining rigs (most AMD cards unfortunately, hopefully OpenCL can be somehow better implemented in Linux systems in the future), a Filecoin rig, and since 3 weeks ago, the LPT orchestrator node. As for NVIDIA cards, we have an RTX 3090 and 2 RTX 3070’s. Depending on future stakes we receive, I am strongly leaning towards adding another RTX 3090. Although expensive, the RTX 3090 is the best card to maximize profitability on services that heavily utilize the VRAM.
Right now, we are easily be able to do around 70 concurrent streams (probably can do 140+ without dual mining) but the problem is we aren’t getting enough work due to our stake value. We average around 16-20 concurrent streams daily which utilizes VRAM slightly lower than the maximum capability of the 2 RTX 3070’s. The RTX 3090 has rarely been utilized in transcoding.
We are right below the Livepool US server on the staking leaderboards, so my goal for now is around a 1+k stake to overtake them. Next hurdle is receiving 35k+ in LPT, which is the gap between 28th and 29th place on the staking leaderboards.
I’ve also noticed that a significant amount of orchestrators near the top of the leaderboard have very small fees collected presumably due to lack of GPU power. We should be able to quickly overtake them in fees as we rise to the top.
Rewards are being called every day now so the fees are currently 20/80. (for the Last eth winning ticket though, 35% was paid out to the staker. As we rise higher in the leaderboards, my plan is to gradually increase the reward cut until we reach a cap of around 38-50%, while paying out 75%-85% of ETH fees collected. (Around a 40/20 (LPT/ETH cut)) It’ll provide financial stability to me, while increasing the profits of stakers. It seems to me there is an overemphasis on the Reward Cut, but over the long run, I do not think that is the best way to provide significant value to stakers. As the Livepeer platform matures, the profitability of orchestrators will increasingly become more dependent on the ETH fees generated.
Thanks for taking the time to read my campaign. I’ll leave some screenshots of my grafana dashboards down below. And good news: We’ll also be making our first donation this Friday (6/11/2021) to The Met for our philanthropy initiative.