We have the talent, but not a UK LLM company … yet One needs $1B just to get the necessary GPUs for training a top-10 LLM, assuming one has the data centres, and capital to crawl, buy books, OCR, filter, dedupe, parse, quality score, carry out data mixture ablations, etc, and
AI HARDWARE
-
Kimi Proposes Solution to LLM Prefill-Decode KV Cache Transfer Problem
By
–
A new paper just exposed how much money AI labs waste on GPUs. Running LLMs at scale hits a wall when prefill and decode share one datacenter. The KV cache transfer between them is massive. This forces expensive RDMA networks and identical hardware everywhere. Kimi proposes
-
NVIDIA Spatial Computing Examples Built With Latest Technology
By
–
Every Spatial Computing example NVIDIA had at GTC last year was built with it. Totally congrats and well deserved.
-
Tesla Robot Outperforms Rivian in Early Deployment
By
–
Yes, let's. Tesla runs better on my eight year old robot than Rivian will on its robot people can't get until next year.
-
Huawei narrows ASML lithography advantage with improved machine development
By
–
Yes and no. ASML has an advantage, but as yesterday's release showed, less than one might think. Huawei is now also managing to develop good lithography machines, admittedly not as good as ASML, but far better than expected.
-
Chinese VR Companies Pico and Vitur Emerge as Market Leaders
By
–
Watch for new companies, like China's @PICOXR
, or @getVITURE to get there first. -
Israeli Computer Vision Startup’s Unconventional Innovation Culture
By
–
I visited an Israeli computer vision company. With a similarly impressive feature list. After the cameras were off the founder asked: “Where do you think those features came from?” “An offsite at a nice resort?” “Nope.” Then pulled out a bag of weed. “We get stoned in our
-
Robot Bolt Outpaces Usain Bolt With Record Speed
By
–
Bolt by MirrorMe | Claims speeds of 11m/s indoors, 10.09 m/s outdoor so far (Usain Bolt's top speed is 12.42 m/s) Now robots outrun the fastest humans on earth.