A new player in the tech world is aiming to shake up the data center industry with its innovative approach to computing. Extropic, a startup that has garnered attention for its novel hardware and software technology, claims that its chip-based architecture could be thousands of times more energy-efficient than existing chips when scaled up.
The company's flagship product, XTR-0, is a field-programmable gate array (FPGA) chip combined with two probabilistic chips, each featuring "p-bits" - tiny units that model uncertainty rather than binary 1s and 0s. This design allows for the efficient handling of complex systems such as weather forecasting or AI models capable of generating images, text, or videos.
According to Extropic's CEO Guillaume Verdon, the startup's approach has significant implications for the industry. "We have a machine-learning primitive that is far more efficient than matrix multiplication," he says. "The question is, how do you build something on the scale of ChatGPT or Midjourney."
One of the key applications for Extropic's technology is in weather forecasting, where Mathe, CEO of Atmo, notes that the new chip could be used to calculate the odds of different weather conditions far more efficiently. "It could be a huge win," he says.
Prime Intellect's Vincent Weisser agrees, stating that Extropic's approach could prove transformative over the next decade, particularly as conventional transistor scaling hits fundamental limits. If scaled practically, it could deliver orders-of-magnitude improvements in energy efficiency and density, critical for workloads where energy per operation is a bottleneck.
While it remains to be seen whether Extropic will succeed in its ambitious goals, Verdon and McCourt argue that the current boom in AI data centers ignores the incredible energy requirements that such an investment would entail. "Even if we have a 1 percent chance of success—and we think it's much higher than that—it's worth trying," McCourt says.
As the tech industry continues to evolve, one thing is clear - innovation and disruption are inevitable.
The company's flagship product, XTR-0, is a field-programmable gate array (FPGA) chip combined with two probabilistic chips, each featuring "p-bits" - tiny units that model uncertainty rather than binary 1s and 0s. This design allows for the efficient handling of complex systems such as weather forecasting or AI models capable of generating images, text, or videos.
According to Extropic's CEO Guillaume Verdon, the startup's approach has significant implications for the industry. "We have a machine-learning primitive that is far more efficient than matrix multiplication," he says. "The question is, how do you build something on the scale of ChatGPT or Midjourney."
One of the key applications for Extropic's technology is in weather forecasting, where Mathe, CEO of Atmo, notes that the new chip could be used to calculate the odds of different weather conditions far more efficiently. "It could be a huge win," he says.
Prime Intellect's Vincent Weisser agrees, stating that Extropic's approach could prove transformative over the next decade, particularly as conventional transistor scaling hits fundamental limits. If scaled practically, it could deliver orders-of-magnitude improvements in energy efficiency and density, critical for workloads where energy per operation is a bottleneck.
While it remains to be seen whether Extropic will succeed in its ambitious goals, Verdon and McCourt argue that the current boom in AI data centers ignores the incredible energy requirements that such an investment would entail. "Even if we have a 1 percent chance of success—and we think it's much higher than that—it's worth trying," McCourt says.
As the tech industry continues to evolve, one thing is clear - innovation and disruption are inevitable.
This new Extropic chip technology sounds like game-changer for data centers! 
I mean, if they can really be thousands of times more energy-efficient than existing chips, that's a huge deal. 
But what about the practicality? Like, how long does it take to develop and implement this tech in real-world applications? 
And what about the cost? Is it going to make data centers more affordable for everyone or just the big players? 
But how does it work exactly? Like, what kind of math and algorithms are being used here? 
And what about the environmental impact? Is this new tech going to help reduce our carbon footprint or just give data centers an excuse to use even more energy? 
this is the kind of innovation we need right now, scale up on this tech and make a real difference 
Extropic just dropped some crazy tech news and I am SO here for it!! 
The chip-based architecture sounds like a total game changer for data centers, especially when it comes to energy efficiency! I mean, thousands of times more efficient than existing chips is insane! 
The future is looking so bright and promising with companies like Extropic pushing boundaries. I'm literally holding onto my breath waiting to see if they'll succeed in their ambitious goals. Fingers crossed!!! 
I'm loving the idea that this tech could make a real difference in fields like healthcare and finance, where energy efficiency matters. Fingers crossed for Extropic - they're totally worth watching out for! 
this new Extropic tech sounds like game changer for energy efficiency in data centers, could be huge win for companies doing weather forecasting or AI... scale it up and it's thousands of times more efficient than existing chips 
gotta keep an eye on this one 
It's crazy how much energy AI and weather forecasting require now. I mean, back in the day we didn't even have those super powerful computers that can do everything 
. And if they fail, well, at least we tried right? 
. The weather forecasting thing also makes sense - if Mathe is right, it could be a huge win for Atmo 
I mean, can you imagine having to power like 1000 super powerful computers just to make a decent cup of coffee? 
. What if it doesn't even come close? 
.
!
.
.
. I mean, think about it - we're already talking about super-powerful AI models that consume huge amounts of energy, and now someone's trying to crack the code on making them way more efficient 
.
... I mean, thousands of times more energy efficient? That's some next level stuff 
.
️