@elonmusk@elonmusk you should bring Starlink Mobile to Canada asap.
Canadians are getting absolutely fleeced by the telco duopoly here that can’t even figure out how to protect customers from SIM swapping.
Predatory pricing. Third-world security. First-world monopoly profits.
Neurophos spent years shrinking optical transistors to fit a million in a 5 x 5 mm area instead of a square meter.
A close look at the hardware we developed to tackle AI’s power consumption and capital cost.
We built a new kind of robot for the underdogs.
Today, robotics is stuck between ~$100k systems and low-cost robots not meant to survive deployment.
We think builders deserve better. Our mission is to develop the best hardware platform for others to deploy on.
Meet Feather, an affordable general-purpose robot designed for real work:
• 1m reach
• Human strength
• 10-hour battery
• Shipping for the past year + available today
Founded by @whoishoa and Parsa Bakhtiari, Feather is coming out of stealth with $7.6m in pre-seed funding led by @GradientVC, with participation from @BuilderVC, @geometryvc, @SEEDInnov, and Virgo VC.
We’re hiring across engineering, ops, and go-to-market. If you’re an absolute maverick, join our cause today – https://t.co/91NL7AKvCc
My in-depth, 35 minute conversation about @neurophos's optical prefill accelerator with @PatrickTBowen (CEO).
0:00 - What's a metamaterial?
1:51 - Why should we care about optical compute?
5:47 - Why it's not about the speed of light
7:17 - Optical compute is a graveyard, why is Neurophos different?
12:08 - Free space optics & silicon photonics - what do these things mean?
16:21 - What's the catch?
19:56 - Compute vs memory constraints in inference
24:25 - The Handoff Problem: How difficult is it for customers to integrate specialized/disaggregated accelerators?
26:11 - What will be improved in the metasurface over time?
27:44 - A surprising feature of the semi industry
30:19 - An unexpected customer reaction
31:26 - If Neurophos were to fail, what would the reason be?
33:51 - & what if it succeeds?
@geometryvc is an investor.
@octra can someone explain how mapping ciphertexts to hypergraph vertices actually lowers FHE cost? is HFHE doing something fundamentally different or just more efficient?