Microsoft showed off the Maia 300 recently, the company’s third homegrown AI inference chip. Microsoft says it will pump out 40 percent more tokens per megawatt than the best GPU setups and is expected to ship by 2027. Power is the biggest bottleneck in AI data centers right now. Microsoft shaped the chip to tackle exactly that. Tokens are the bits of text a model takes in and spits out. Tokens per megawatt tells you how much work you get from each unit of power. Scott Guthrie, who heads Cloud + AI at Microsoft, put it this way: the company expects the chip to reach that 40 percent gain. Microsoft stayed quiet on which GPU it measured against and how it ran the tests. So the number comes from Microsoft itself. The company also says one rack of Maia 300 can push up to seven times the token throughput of a Maia 200 rack. The Maia 300 splits its work across several chiplets. TSMC builds the compute chiplets using its N2P process. Memory controllers and input/...
Spending as little as 10 minutes with an AI tool may chip away at a person’s willingness to grind through tough work, according to a newly published study. A team of researchers carried out several experiments and found that people lost focus and gave up more easily on demanding tasks soon after turning to artificial intelligence for help. The team calls the result a striking one. They say it matches what plenty of people already suspect: leaning on AI slowly wears down our mental sharpness. When answers arrive within seconds, working things out on your own starts to feel painfully slow. That robs people of what the researchers call productive struggle, the kind of effort that actually helps us improve. Scam attempts hit 84% of Filipino adults, losses reach Php121.8 billion Artificial Intelligence designs viruses that kill bacteria more effectively Brian Christian of the University of California, Berkeley, who took part in the work, put it this way. These...