We’re pouring more and morepower into copper just to move bits a few meters – and paying for it in cooling. Silicon photonics slashes interconnect power per bit, easing the power/cooling crunch and freeing budget for compute, not heat. #Datacenters# #EnergyEfficiency#

🔥 ANTRACK lowers heating costs through efficient recovered heat utilization. 📍 In Martinville, Québec, Project BTU uses BITMAIN ANTRACK to recover heat generated by hydro-cooled ANTMINERs and deliver it to the local community centre's heating system. 📊 According to Radio-Canada, the project is expected to reduce the centre's heating-related energy bill by nearly 50%. Footage filmed on site shows the system operating at around 54°C, while the ANTRACK heat-pump version can deliver hot water at up to 80°C for a wider range of heating applications. ♻️ BITMAIN ANTRACK converts recovered heat into usable thermal energy—reducing heating costs, improving energy efficiency, and creating lasting value for mining operations and local communities. BITMAIN remains committed to consistently delivering better products and services to its clients. Learn more 👇 antrack@bitmain.com Selected news footage and project outcome information are credited to Radio-Canada. Power consumption comparisons are based on manufacturer-rated specifications. Actual operational cost savings depend on your local electricity rates, facility conditions, and operational practices and may vary. #BITMAIN# #ANTMINER# #ANTRACK# #ProjectBTU# #HeatRecovery# #HydroCooling# #EnergyEfficiency# #SustainableInfrastructure# 🌍💡♻️

Purdue University researchers have developed the world's whitest paint, reflecting 98.1% of sunlight and cooling surfaces by up to 8°F during the day. This innovation could reduce the need for air conditioning, offering a sustainable solution to combat global warming. #WhitestPaint# #PurdueInnovation# #PassiveCooling# #EnergyEfficiency# #ClimateTech# #SustainableDesign#

To meet the new energy-efficiency requirements, we introduced the GANSPIN611 and GANSPIN612, ST’s first GaN motor drivers. Both include two GaN power transistors in a half-bridge topology. 🔍 Discover more on the #STBlog#:

Getting to 1000x energy efficiency in AI isn’t about one breakthrough. It’s about solving two hard constraints: 1. Data movement dominates energy 2. Amdahl’s Law caps system-level gains Which means you have to rethink everything: models, hardware, and how they’re designed together. If this kind of problem excites you, you’ll enjoy our latest blog:

FULL INTERVIEW: @beffjezos says Z1T hits up to 140x the energy efficiency of a GPU, and 100x at the data center level. The end game, he says, is chips in space powered by the sun. @beffjezos is founder and CEO of @extropic. He joined @schisofrenia to launch Z1T, the first family of transformer-like models built for sparse probabilistic hardware: 01:12 what Z1T is, and how long they've been cooking it 02:32 why models on GPUs are dense and his chip isn't 03:15 a new scaling law where you scale connectivity, not just flops 04:08 how you get 140x, and why a pbit uses 10,000 fewer transistors 07:12 where Z1 sits next to GPUs, and why it's decode not prefill 08:37 100x energy efficiency at the data center level 09:14 you might see Extropic chips in space sooner rather than later 09:45 open weights vs an open training recipe 10:24 Z1T is barely GPT-2, and they started a run last night for GPT-3 11:05 the chips are cooking right now at a major fab overseas 12:58 gated convolutional attention, and the layer they haven't cracked 14:03 the Hardware Lottery, and why he wants people gambling again 15:20 what happens if inference gets 1,000x more efficient 16:59 why you can do this research without a basement full of GPUs

Unconventional AI CEO @NaveenGRao says Moore's Law let computers ignore energy efficiency, but AI scaling has made that impossible: "We've never really cared about energy efficiency in a computing system. Moore's Law was kind of working and we didn't really need things to be efficient, we just needed them to be fast." "Now we're in this new place where we do need them to be efficient. We're running out of energy at the global level in several years. If we don't think about efficiency, we can't keep scaling." "The kind of almost basic thing we've seen out of computers is they achieve more things when we scale. If we wanna keep building bigger models, better models, and more of it used in the world, we have to solve this problem. It's not gonna be something that we're just gonna keep generating more and more energy." "There are thermal limits to certain form factors. If you wanna build a robot, you're just not gonna do it with an orbital data center." @unconvAI

TPU 8 will be a massive leap forward in energy efficiency—with up to 2x better PPW than Ironwood. But for us, sustainability isn’t about waiting for the next best chip; it’s a daily effort to make our entire production fleet more efficient → #8daysofTPU8#