The new Mini Countryman Untamed is here to explore the path less traveled. The Countryman Untamed, revealed on Tuesday, takes the basic adventure-ready crossover theme and applies it to the brand’s bread-and-butter model. The lifted crossover will be available in the North American market in two trim levels—the Mini Countryman S ALL4 and all-electric Mini Countryman SE ALL4 —with upgrades to ground clearance, specially developed all-terrain tires, and a revised intelligent ALL4 all-wheel drive technology. The Untamed upgrades will be offered as a $1600 package for those two Countryman models. Read our full story on the new Mini Countryman Untamed ➡️ #mini# #offroader# #carsofx# #roadandtrack#

One demonstration closer to delivering VIPER to the Moon. We just completed a risk-reduction test campaign with @NASA: running a full offload with a VIPER Mobility Ground Research Unit (MGRU) before doing so on the lunar surface. We proved out the full concept of operations with our flight-like offloader: the complete offload sequence, nominal and off-nominal, ran end-to-end across the full range of lander tilt angles, with joint mission operations and NASA commanding the rover remotely. Stay tuned for images of our flight offloader, which is being fabricated now!

We offloaded physical labor to machines long ago, and now we are doing the same for intellectual labor. There’s no reason to expect we will retain cognitive agency. Thinking is effortful, so if it can be outsourced, we will happily do that. Post-work society seems inevitable.

Ken Griffin's Citadel has offloaded more than 80% of the aggregate risk associated with the Situational Awareness stock portfolio it acquired from the embattled AI hedge fund. Hema Parmar reports

Breaking news: Ken Griffin’s hedge fund told investors that it had offloaded most of the 'aggregate risk' from Situational Awareness’s book, having completed almost 100 block trades that amounted to $4bn in recent weeks to unload the positions.

For long-context LLM inference, should the KV cache be offloaded to disk or just recomputed on the GPU? There's no universal right answer, and this paper builds a system that decides quantitatively. Title: Building py-kvcache: A Performance Characterization of External KV Caching for vLLM with NVMe SSDs URL: 📝 Overview The paper proposes py-kvcache, an external KV cache system for vLLM. Using io_uring for async I/O, even a Python implementation pulls near-full SSD read bandwidth of 13.5GB/s. ❗ Problem it solves Existing external KV caches like LMCache have no criteria for when to actually use them — they load from disk unconditionally even when a short prefix or a fast GPU makes recomputation cheaper. ⚙️ Methodology It introduces "scheduler-aware preloading," which starts disk reads while a request is still waiting to be scheduled, plus a "break-even gate" that rejects a load whenever it wouldn't improve time-to-first-token. 📊 Results On LongBench multi-document QA it's 6.02-7.43x faster than GPU recomputation and 2.77-3.64x faster than LMCache. On multi-turn SCBench, native vLLM read 3.4TB from disk with completion times over 1200s, while py-kvcache kept disk reads to 85GB and completion time to 480s. 🖥️ Use cases On a high-end H100, requests often don't even clear the break-even point, so skipping external caching is fine — but on a lower-end RTX 4000 Ada, external caching clearly wins, giving concrete hardware-specific guidance. #LLMInference# #vLLM#