#GPU-based
I think this is a slight mischaracterization. I don’t know for sure how this Slack-based mulit-model interface works, but it sounds like all of GPT workers and GPU processes involved are presented to Claude by the harness as things it can control
nitter.tiekoetter.com/SohanSarabu/...
seeing cases where Claude code checks the behavior of other software on the computer and killing processes if it thinks you’re violating Anthropic’s policy for what work is allowed to be done. dire fucking stuff
November 17, 2026 at 1:39 PM
Intel Talks Up GPU D270 Performance & Open-Source Software Support - https://www.phoronix.com/news/Intel-GPU-270-Open-Performance
Intel Talks Up GPU D270 Performance & Open-Source Software Support
For Supercomputing 2026 week, Intel published a blog post talking up their Intel GPU D Series "Crescent Island" hardware based on Xe3P...
www.phoronix.com
November 16, 2026 at 3:12 PM
This isn't me being glib, I am just genuinely surprised people are surprised by this

Nvidia's model for at least 4 years:

- Nvidia invests/loans money for its GPUs
- Shortages made GPUs appreciate in value
- Business got loans based on GPU appreciation to buy more GPUs
- Nvidia pre-sells more GPUs
Sigh.

WSJ reports that NVIDIA allegedly artificially inflated the ongoing GPU and memory shortage by investing in AI startups that would go on to buy NVIDIA GPUs and build data centers.

https://www.techpowerup.com/354814/nvidia-allegedly-engineered-scarcity-by-funding-ai-startups-to-buy-its-gpus
November 14, 2026 at 10:49 AM
you can actually see when the Attention is All you Need paper was published based on when "GPU" overtook "graphics card" in google search trends
November 15, 2026 at 6:13 PM
New GPUs such as Intel Xe, AMD RX 7000, RX 9000 series, and NVIDIA RTX 40, 50 series now works in Windows 7 with this unofficial WDDM 2 backport!
DirectX 12 is also working too!
This isn't Windows 10/11 larping as Windows 7, this is ACTUALLY Windows 7! 😎
w7revived.chefkiss.dev/WDDM2Win7/
November 17, 2026 at 11:27 PM
- Split big terrains into chunks, then divide these into separate scenes and additively load/unload them based on player position;
- Reduce alphamap and detail map resolution, increase Pixel Error to 5 or higher
- Assets: GPU Instancer for mass tree/grass rendering, or Microsplat for terrain shaders
November 18, 2026 at 1:59 PM
Success story🏆Researchers used GPU-accelerated SOD2D simulations and Bayesian optimisation to improve high-lift wing aerodynamics.

🔹10.9% higher lift-to-drag efficiency
🔹9.7% lower drag
🔹CFD–reinforcement learning workflow demonstrated on 120 GPUs of MareNostrum 5

www.excellerat.eu/enhancing_hi...
Enhancing High-Lift Wing Performance through GPU-accelerated Simulations and Active Flow Control
Within EXCELLERAT, the CFD code CODA, developed by DLR, ONERA and Airbus, was successfully optimised for modern GPU-based High-Performance Computing (HPC) systems. By significantly improving the…
www.excellerat.eu
November 16, 2026 at 8:55 AM
New Python package combines gradient-ascent pulse engineering with measurement feedback for quantum control. GPU-accelerated implementation uses neural networks and lookup tables to optimize state preparation and stabilization.

#QuantumControl #QuantumSoftware #Research
GPU-Accelerated Feedback-GRAPE: JAX-Based Python Package for Quantum Optimal Control
iq.fp2.dev
November 17, 2026 at 5:28 AM
(Wired and WiFi pretty common on motherboards now.)

Then the GPU, it's really just preference, even with an AMD CPU an Nvidia GPU is fine. Personally I go AMD+AMD cause screw Intel and screw Nvidia but eh.

Then based on your build you gotta estimate power needs and pick a power supply (PSU)
November 18, 2026 at 10:53 PM
Noticed I was getting low GPU memory alerts in FH6, apparently when you post images to Bsky GPU memory usage in Firefox explodes. Also the amount it increases is based on the number of images you post. What the fuck!?
November 13, 2026 at 9:53 AM
- Power profile based effects toggles for minimizing power usage when on battery or low power profiles and a separate "game mode" for when you want to cut off GPU based effects for game performance
- Support for packaging and sharing themes and dynamic wallpapers
November 17, 2026 at 8:49 AM
I have a question.

How big is the difference between an open-source local LLM that runs on a somewhat fancy PC with a nice GPU or two, and the cloud based subscription AI models?

My vibe is that it's like 70-80% as good... Thoughts?
November 19, 2026 at 6:25 AM
SANSAR Linux Cryptomining Malware Uses Modified NVIDIA Library to Hide GPU Activity

A financially motivated threat activity cluster tracked as REF3364 has been attributed to a Linux-focused cryptocurrency mining operation that tampers with the readings used to detect high…
#hackernews #news #nvidia
SANSAR Linux Cryptomining Malware Uses Modified NVIDIA Library to Hide GPU Activity
A financially motivated threat activity cluster tracked as REF3364 has been attributed to a Linux-focused cryptocurrency mining operation that tampers with the readings used to detect high GPU workloads. Central to the intrusion is a toolkit called SANSAR, a Rust-based orchestrator that receives commands over the TOR anonymity network to manage CPU and GPU miners. The activity was observed in
thehackernews.com
November 17, 2026 at 9:08 AM
Just starred colibri. It's a C-based inference engine for MoE models that manages memory by moving experts between GPU, RAM, and NVMe storage based on their usage. It's a clever way to run large models on local hardware.

https://github.com/JustVugg/colibri
JustVugg/colibri
⭐ 51,988 stars • C A hierarchical inference engine for frontier MoE models: hot experts on the GPU (CUDA, Vulkan), warm ones in RAM, cold ones streamed from NVMe. Written in C, runs on the hardware you already own.
github.com
November 19, 2026 at 2:01 AM
The people complaining about GTA VI only being 30fps crack me up. The graphics hardware in the PS5 is based on 7 year old technology, similar to an AMD RX6700 (a mid range gpu at the time). It just doesn't have the grunt to push 4k/60. Especially not in an open world game like GTA VI with the /1
November 19, 2026 at 7:41 AM
Can an AI agent fix a broken physics simulation overnight? KTH built a Nemotron-based agent that repairs broken GROMACS molecular dynamics setups. In one overnight run on a single DGX Spark, it fixed 81% of test cases, each fix checked by GROMACS itself. #MachineLearning #Agents
Nvidia Unveils Rubin HPC GPU and Nemotron for Science at SC26
At SC26, Nvidia introduced Nemotron for Science, the Rubin HPC GPU, Vera CPU racks and new CUDA-X libraries for scientific research.
techstrong.ai
November 17, 2026 at 12:34 PM
📰 **Intel Talks Up GPU D270 Performance & Open-Source Software Support**

For Supercomputing 2026 week, Intel published a blog post talking up their Intel GPU D Series "Crescent Island" hardware based on Xe3P...

📰 Source: Phoronix
🔗 Link: www.phoronix.com/news/Intel-G…
Intel Talks Up GPU D270 Performance & Open-Source Software Support
For Supercomputing 2026 week, Intel published a blog post talking up their Intel GPU D Series "Crescent Island" hardware based on Xe3P...
www.phoronix.com
November 18, 2026 at 3:35 PM
November 18, 2026 at 10:01 AM
"While Jensen can’t be faulted for making these moves because he was keeping the best interests of the company in mind" This is false based on the writer's own reporting lmao

wccftech.com/jensen-huang...
Instead Of Lowering GPU Prices, Former NVIDIA CEO Jensen Huang Reportedly Financed AI Startups To Absorb Excess Silicon, Shattering The Shortage Perception
New details reveal that former NVIDIA CEO Jensen Huang used circular financing to fund AI startups and absorb excess TSMC wafers and memory rather than letting GPU prices drop
wccftech.com
November 14, 2026 at 1:59 PM
not today/Monday based on the info I have seen. I bought Micron instead. They will sell GPU memory to whoever's buying.
November 15, 2026 at 2:14 AM
Hyperscalers List GPUs at Nearly 2× the GPU-Cloud Price, Fluence Research Finds

Fluence's AI Compute Race report reveals GPU compute pricing disparities for companies outside of frontier labs and hyperscalers. Hyperscalers list GPUs at nearly double the median price found elsewhere, …
#crypto #news
Hyperscalers List GPUs at Nearly 2× the GPU-Cloud Price, Fluence Research Finds
Fluence's AI Compute Race report reveals GPU compute pricing disparities for companies outside of frontier labs and hyperscalers. Hyperscalers list GPUs at nearly double the median price found elsewhere, with significant variation based on the specific GPU model. Longer commitment periods drastically reduce GPU-hour costs, with three-year terms offering a median price cut of 48%. Spot capacity provides a cheaper option for interruptible workloads, costing 53% of on-demand rates. Despite expectations, 17 out of 22 tracked GPU models have increased in price over the past year. This rise is driven by the rapid adoption of open-weight AI models. These models allow any company with GPUs to serve advanced AI on their own hardware, increasing demand. The largest buyers have secured significant future GPU supply through massive capital spending and large contracts with neocloud providers. Consequently, the GPU-hour is emerging as a financial asset, with futures markets and lending options now available. The report advises companies to pay less for compute by comparing providers, committing to steady workloads, matching GPUs to models, and looking beyond hyperscalers for capacity.
www.newsbtc.com
November 15, 2026 at 8:38 AM
Could work, harder part is probably to make a usable language model that is lightweight enough to be practical to distribute with the game and cheap enough to be used "on the fly" during gameplay (doesn't eat all the RAM or tank the CPU/GPU). Cloud based LLMs would seem like a bad idea here...
November 16, 2026 at 8:31 PM
The new Xe3P-based hardware outperforms NVIDIA's Blackwell Server Edition in financial and AI workloads while using significantly less power.

https://hardwarebrief.com/article/intel-unveils-d270-crescent-island-gpu-at-s...

#IntelD270 #CrescentIsland #SC26 #GPU #ArtificialIntelligence
November 17, 2026 at 11:35 AM