Your GPUs could be earning you a lot more money.
Inference providers, you're already paying for your GPUs, whether they're working or not.
Idle hours, low-demand periods, and unused compute capacity during inference are all opportunities to earn more.
Peryn lets you mine $PEARL alongside your inference workloads, generating additional revenue from the same GPUs.
That extra income could cover a significant part of your GPU rental costs, or even all of them.
Higher margins. Lower inference prices. A competitive advantage.
Same GPUs. Two revenue streams.
Reach out to us to see what Peryn can do for your fleet.
Elon just described the exact problem pearl-2:native solves.
Neoclouds need revenue beyond AI contracts.
With Peryn, GPUs can mine pearl-2:native while idle, between inference jobs, and even during inference.
No customer, no contract, no waiting.
Stranded compute just needs to plug in.
Pretty much.
As of the past few weeks, debt markets have almost entirely shut off access to money for any company that doesn’t have substantial revenue independent of AI data centers.
SpaceX closed $40B of debt for GPUs, but we’re investment grade and have Starlink & launch revenue.
That means the 200+ neoclouds are going to see guppy-level death and consolidation. Some will make it through, but not many.
What's next: in-band mining. Fine-tuning and batch jobs that mine pearl-2:native while they run, up to 41 % of pure mining output on top of the job, quality checks passing.
Already passing on GLM-5.3-Flash. DeepSeek V4.1 Flash and MiniMax M3.1 Flash are next.
Founding Miner signups are open: 0 % PPLNS / 1 % FPPS pool fees until 6 months after Pearl’s FP8 upgrade, plus a permanent Founding Miner badge.
pearl-2:native
peryn.ai/mine
Our Pearl FP8 miner on Hopper is now within 2% of the hardware's theoretical limit under Pearl's mining rules.
29% faster than Pearl's open-source miner. Even faster than NVIDIA's cuBLASLt on mining workloads.
And it runs alongside live inference, delivering over 50% of its full mining hashrate during prefill-heavy serving with zero loss in inference throughput.
FP8 fork will transform Pearl mining.
We've built the world's most efficient duplex mining infrastructure for clouds.
Serve inference, secure Pearl, earn pearl-2:native.
All with the same compute.
AI unit economics will never be the same.
xitter.cf/aliomerhorzum/status/2…
Peryn retweeted
Pearl Safe 1.1 is live on Google Play.
bitcoin:native and zcash:native , next to pearl-2:native In one self-custody vault.
Safe by design.
Simple by nature.
Marketplaces & enterprises: wholesale AI capacity, mining-subsidized.
→ Volume pricing
→ Choose where requests run
→ Private deployments, never mined
[email protected]
$17B a year to mine math nobody needed.
$700B a year on GPUs for AI.
Pearl made it one computation. Peryn makes it pay twice.
peryn.ai
