October 8, 2026 · x.com
Local inference: performance and power on an RTX 3090
Tests of eight local LLMs on one RTX 3090 compared throughput and energy efficiency at power limits ranging from 100 W to 450 W.
Technology and security alerts
Each alert sums up the news and says what to do.
Recent alerts
October 8, 2026 · tunbury.org
ONNX inference: comparing CPU and GPU
Published on March 11, 2026, the article compares ONNX Runtime inference on an NVIDIA L4 GPU and an AMD EPYC 9965 CPU. The author reports that OCaml and memory placement with libnuma made CPU inference faster than using scarce GPUs at TESSERA.
October 8, 2026 · arxiv.org
Instability in low-precision Flash Attention
Published on March 15, 2026, the paper examines catastrophic loss spikes in low-precision training with Flash Attention, linking them to similar internal attention data patterns and accumulating biased rounding errors.
October 8, 2026 · users.cs.utah.edu
Integer overflow in C and C++: controls to reduce risk
A 2012 University of Utah article discusses integer overflow in C and C++, a relevant risk when writing and reviewing code in these languages.
October 8, 2026 · videocardz.com
Intel Arc Pro B60: evaluating 24 GB for workstations
An article dated November 6, 2025, says the Intel Arc Pro B60 workstation GPU, with 24 GB of memory, was expected to reach Europe between mid and late November, starting at €769.
October 8, 2026 · cdrdv2-public.intel.com
Intel updates the AVX10.2 specification to version 6.0
In a post published on March 11, 2026, Intel announced version 6.0 of the AVX10.2 specification and the renaming of AVX10_VNNI_INT to AVX10_V1_AUX.
October 8, 2026 · intel.com
Intel documents processor-instruction access through C functions
Published on February 17, 2026, Intel’s documentation presents C-style functions that provide access to processor instructions without writing assembly.
October 8, 2026 · cdrdv2-public.intel.com
Intel publishes HPC tuning guide for Xeon 6
On December 29, 2025, Intel released version 1.2a of its tuning and optimization guide for high-performance computing applications on Xeon 6 systems.
October 8, 2026 · arxiv.org
Interleaved Head Attention explores communication between heads
A post published on April 11, 2026, explores communication between attention heads within a layer, focusing on reasoning with Interleaved Head Attention.
October 8, 2026 · sysdig.com
AWS intrusion: exposed credentials lead from S3 to Bedrock
In an article published on February 4, 2026, Sysdig describes an AWS intrusion that began with credentials exposed in S3, moved through Lambda, and reached Bedrock access. The article highlights an Elastic rule for detecting external layers added to Lambda functions.
October 8, 2026 · ieeexplore.ieee.org
Gradient inversion in Federated Learning: risks and defenses
Published on January 16, 2026, the post presents a study of gradient inversion attacks in Federated Learning. It says the most practical attack method remains unreliable and that the work proposes a three-stage defense.
October 8, 2026 · x.com
IPUs and small batches: cost versus ease of use
A post published on December 29, 2025 describes Graphcore IPUs as massively parallel accelerators suited to small batches, and says used units may cost little but can be difficult to use.
October 8, 2026 · developers.redhat.com
GCC jump threading: evaluating its effects on generated code
A Red Hat Developer article published on April 2, 2026, explains how jump threading aims to reduce dynamically executed jumps across different control-flow paths.
October 8, 2026 · arxiv.org
JustRL tests simple RL on 1.5B reasoning models
Published on January 4, 2026, the paper describes single-stage training with fixed hyperparameters and reports average accuracies of 54.9% and 64.3% across two 1.5B reasoning models.
October 8, 2026 · arxiv.org
KArAt proposes learnable attention without fixed softmax
Published on October 27, 2025, the paper presents KArAt, a learnable alternative to softmax attention. It reports mixed results in vision transformers and low-rank variants intended to reduce memory use.
October 8, 2026 · x.com
Kascade speeds inference with sparse attention
Published on December 24, 2025, the paper describes a training-free method that reuses key patterns across layers. It reports up to 4.1× faster generation and 2.2× faster prefill on H100 GPUs.
October 8, 2026 · arxiv.org
Kascade reduces attention computation in long contexts
Published on February 18, 2026, the article describes a technique that reuses attention results across Transformer layers to reduce inference work without retraining the model or changing its weights.
October 8, 2026 · tensara.org
CUDA array-sum kernels: from example to practice
Published on November 16, 2025, Tensara’s article introduces beginners to writing a simple CUDA kernel for adding arrays. The example can be a starting point for evaluating GPU workloads.
October 8, 2026 · paulwillchan.com
CUDA/PTX matmul kernel reports 6% gain on B200
A post published on April 27, 2026 describes a pure CUDA/PTX matrix multiplication kernel that reportedly outperformed cuBLAS by 6% on a B200 with M=N=K=8192.
October 8, 2026 · popovicu.com
A minimal Zig time-sharing kernel for RISC-V
Published on May 12, 2026, the article describes building a minimal time-sharing kernel in Zig for RISC-V on OpenSBI, including S-mode traps and timer-triggered context switching.
October 8, 2026 · github.com
Triton kernels and PyTorch operators for AttnRes
Published on May 3, 2026, the repository provides Triton kernels and PyTorch operators for Block Attention Residuals (AttnRes).
October 8, 2026 · cnx-software.com
Embedded kit combines modules and AI accelerators
Published on December 30, 2025, the article describes a kit combining computing modules and accelerators to evaluate LLM and VLM workloads on embedded systems.
October 8, 2026 · outcomeschool.com
KV cache: less recomputation in LLM inference
Published on May 9, 2026, the article explains how KV cache reuses previously computed values and why that matters for memory and computation during inference.
October 8, 2026 · x.com
LaSeR explores last-token rewards for LLMs
In a post published on October 19, 2025, Tencent presents LaSeR, a reinforcement learning algorithm that aligns last-token scores with real rewards and uses one extra token inference.
October 8, 2026 · docs.nvidia.com
CuTe layouts help explain shapes and strides
A NVIDIA post published on February 3, 2026 recommends layout algebra tutorials for understanding how CuTe represents tensors.
October 8, 2026 · github.com
Leaker and passive enumeration of exposed credentials
Published on January 12, 2026, the repository describes a tool for finding credential leaks associated with email addresses through passive online sources.
October 8, 2026 · x.com
LeJEPA: SIGReg chunks cut VRAM use and runtime
A report published on March 4, 2026 says splitting 256 SIGReg slices into chunks of 32 used eight times less VRAM and finished faster on a 4090 MaxQ GPU.
October 8, 2026 · turingpost.com
LeJEPA combines latent prediction and regularization
Published on November 22, 2025, the post describes LeJEPA, which combines JEPA's latent-space prediction loss with SIGReg to shape embeddings into an isotropic Gaussian distribution.
October 8, 2026 · lemas-project.github.io
LEMAS publishes a 150,000-hour speech dataset
A January 12, 2026 publication presents an audio dataset with 150,000 hours across 10 languages, word-level timestamps, and the generative models LEMAS-TTS and LEMAS-Edit.
October 8, 2026 · x.com
LeWorldModel: a compact JEPA trained from pixels
A May 9, 2026 publication describes a 15-million-parameter JEPA trained end to end on raw pixels and reports planning up to 48 times faster than foundation-model-based world models on control tasks.
October 8, 2026 · arxiv.org
LightRetriever cuts the cost of LLM-based retrieval
A post published on March 5, 2026 reports query encoding 1,000× faster and 10× higher throughput with LightRetriever, while retaining 95% of benchmark performance.
October 8, 2026 · github.com
Geometric limits on compression in RAG
A NeurIPS 2026 paper studies compression in RAG through embedding geometry and says a spectral quantity defines a minimum limit for compression.
October 8, 2026 · lurklurk.org
Linkers in C and C++: diagnosing build failures
A guide published on May 27, 2026, explains the linking process and how understanding it helps diagnose build and linker problems in C and C++.
October 8, 2026 · belkadan.com
Dynamic or static linking: weigh startup and disk use
Jordan Rose’s article, published on March 2, 2026, discusses the trade-off between startup time and disk space in dynamic and static linking strategies.
October 8, 2026 · x.com
Linux Device Drivers and hardware communication
A post published on May 25, 2026 recommends Linux Device Drivers as an introduction to communication between the Linux kernel and hardware, for readers with some C and Unix system-call knowledge.
October 8, 2026 · kernel.org
Linux documents bus lock detection on x86
Published on April 11, 2026, Linux kernel documentation covers detecting and handling bus locks on x86. Split locks can affect system performance.
October 8, 2026 · videocardz.com
Lisuan begins shipping 6 nm 7G100 GPUs
On December 30, 2025, Lisuan reported that it had begun shipping its 6 nm 7G100 GPUs to its first customers.
October 8, 2026 · diveintosystems.org
Free book covers systems programming in C
Published on March 31, 2026, Dive into Systems covers systems programming, code optimization, and assembly across multiple architectures, using C as its foundation.
October 8, 2026 · github.com
llama.cpp adds Multi-Token Prediction support
On May 17, 2026, a pull request to the llama.cpp repository added support for MTP heads. The report cites 65 tokens per second for Qwen3.6-27B on an RTX 3090, versus 38 without MTP.
October 8, 2026 · github.com
llmfit helps assess models for local hardware
The llmfit repository, published on GitHub on February 21, 2026, analyzes local hardware to identify compatible models and estimate inference speed.
Let's assess your project
I reply with scope and a proposal.