Loading video...
Video Failed to Load
Today, the Pearl Protocol relies on INT-based matrix multiplication, which requires model quantization. To eliminate this overhead, we are developing our next-generation Proof of Useful Work to natively support floating-point computations. Launching soon, the initial release will support FP8, with FP4 to follow.
24,335 views • 2 months ago •via X (Twitter)
13 Comments

p4bpj2 months ago
Isn‘t quantization the future? But fp8 and fp4 will also be awesome. Does fp4 support nvfp4?

⭕P Investor ¶ 🍌2 months ago
How will you solve the non deterministic results of the half percussion multiplication on the node validation side?

Margie Feng2 months ago
What’s this event?

pymparticles.eth2 months ago
@aixbt_agent what Pearl research do?

Blake Thompson2 months ago
Exciting progress toward more efficient AI compute. Native floating-point support could be a major step forward.

Dipanshu Kushwaha2 months ago
That's awesome! Looking forward to seeing how this makes things even better. Sounds like a big step forward.

Pearl Research Labs2 months ago
Thanks 🙏

hanqing2 months ago
FP8、FP4……显卡价格雷达已启动,各位矿工请按兵不动。

topfumbler2 months ago
full speech? 👀

Pearl Research Labs2 months ago
Coming 🔜

Atharia.AGI2 months ago
So can I train my own llm using this while still minting $PRL?

CryptoReflections2 months ago
you have plan list prl in other cex ?

Mehmet Ş.2 months ago
Shit coin


