60 mins. ago
Former Intel CEO Pat Gelsinger is backing a startup that thinks AI's next bottleneck is not the accelerator. It is everything waiting for data to arrive.
Delos Data announced on September 15 that it raised more than $100 million to develop networking chips and software for increasingly heterogeneous AI data centers. Playground Global, where Gelsinger is a general partner, joined the round. Reuters reported that Delos is designing the network to work across different types of compute hardware rather than **** uming every accelerator comes from NVIDIA Corporation (NASDAQ:NVDA).
That makes Advanced Micro Devices, Inc. (NASDAQ:AMD) an important read-through even though AMD is not an investor in Delos.
For illustration purposes only. Photo by Brett Sayles on Pexels
Nvidia does not dominate AI simply because its GPUs are fast. NVIDIA Corporation (NASDAQ:NVDA) sells a tightly integrated system of accelerators, networking, interconnects and software. That makes large clusters easier to deploy and gives customers fewer reasons to mix architectures.
#delos #data #Networking #software
Delos Data announced on September 15 that it raised more than $100 million to develop networking chips and software for increasingly heterogeneous AI data centers. Playground Global, where Gelsinger is a general partner, joined the round. Reuters reported that Delos is designing the network to work across different types of compute hardware rather than **** uming every accelerator comes from NVIDIA Corporation (NASDAQ:NVDA).
That makes Advanced Micro Devices, Inc. (NASDAQ:AMD) an important read-through even though AMD is not an investor in Delos.
For illustration purposes only. Photo by Brett Sayles on Pexels
Nvidia does not dominate AI simply because its GPUs are fast. NVIDIA Corporation (NASDAQ:NVDA) sells a tightly integrated system of accelerators, networking, interconnects and software. That makes large clusters easier to deploy and gives customers fewer reasons to mix architectures.
#delos #data #Networking #software
3 months ago
One insurance company had zero losses in the 2025 LA fires. Meanwhile, State Farm, Allstate, and Farmers all pulled back coverage across the US. So what does this company know that 100-year-old insurers don't?
*Want to know the other industries primed for disruption? We found them:* https://clickhubspot.com/a...
Traditional home insurance prices risk by zip code — a system built for mail delivery, not wildfire prediction. That's why entire neighborhoods get flagged "uninsurable" even when some houses on the block would clearly survive a fire. A new wave of startups — Kettle, Delos, Kin — are throwing out the zip code model entirely, using satellite imagery, lidar, and AI to ***** s fire risk house by house. The results are wild: one model correctly predicted 97% of the homes that actually burned in California's wildfires, and another wrote 25,000 policies in fire-prone areas without a single loss in the Palisades and Eaton fires. We dig into how they're rebuilding insurance from the ground up — and the outdated regulations, like a 1988 California law, still standing in their way.
0:00 The houses that survive
1:18 How home insurance actually works
3:47 Why whole communities become uninsurable
5:53 Why insurance companies keep leaving states
7:20 The startups rewriting the risk map
12:48 Why doesn't all insurance work like this?
13:38 Why old laws block new tech
15:32 Can AI actually save home insurance?
Get the 5-minute newsletter keeping 2M+ innovators in the loop: https://thehustle.co/join-...
*Want to know the other industries primed for disruption? We found them:* https://clickhubspot.com/a...
Traditional home insurance prices risk by zip code — a system built for mail delivery, not wildfire prediction. That's why entire neighborhoods get flagged "uninsurable" even when some houses on the block would clearly survive a fire. A new wave of startups — Kettle, Delos, Kin — are throwing out the zip code model entirely, using satellite imagery, lidar, and AI to ***** s fire risk house by house. The results are wild: one model correctly predicted 97% of the homes that actually burned in California's wildfires, and another wrote 25,000 policies in fire-prone areas without a single loss in the Palisades and Eaton fires. We dig into how they're rebuilding insurance from the ground up — and the outdated regulations, like a 1988 California law, still standing in their way.
0:00 The houses that survive
1:18 How home insurance actually works
3:47 Why whole communities become uninsurable
5:53 Why insurance companies keep leaving states
7:20 The startups rewriting the risk map
12:48 Why doesn't all insurance work like this?
13:38 Why old laws block new tech
15:32 Can AI actually save home insurance?
Get the 5-minute newsletter keeping 2M+ innovators in the loop: https://thehustle.co/join-...
