Amazon Web Services researchers published a method called 'priming' that converts pretrained Transformer models into hybrid architectures mixing attention and state space model (SSM) layers, reusing existing weights instead of retraining from scratch. The approach cuts pretraining costs by over 99% and lets a 32-billion-parameter hybrid model match its source Transformer's quality while decoding up to twice as fast. Amazon tested three SSM types and released an open-source model zoo, training code, and serving tools under Apache 2.0.
NVIDIA, working with Princeton's Dao AI Lab, built ReplaySSM, a new way to store state in hybrid models that mix state-space (SSM) layers with attention layers. Instead of rewriting the entire state on every token, the method keeps a periodic checkpoint plus a short list of recent inputs, splitting the job of producing output from the job of updating memory. The result is roughly half the memory traffic, decode kernels running up to 1.80x faster at scale, and speculative decoding throughput more than doubling on 120B-class models, all while keeping accuracy close to full precision. The method already ships inside TensorRT-LLM, vLLM, and FlashInfer, and is running in production for GLM-5.3-Flash and Kimi-K3.
At the OCP Global Summit, NVIDIA contributed its Vera Rubin rack and compute tray designs to the open MGX architecture and announced a series of partnerships covering AI factory power, cooling and networking standards. Partners include hyperscalers such as Meta, Google, Microsoft and AMD, plus Cisco and Visionbay. The announcements cover rack design specs (MGX 1.2), new open-source software for managing AI data centers (DSX OS components), grid resilience standards, an 800VDC power architecture, new Ethernet switching (Spectrum-6) and a new optical transceiver standard (USFP) with Cisco.
AWS published a joint engineering write-up with Postman describing how Agent Mode, Postman's AI-native interface for API testing and documentation, was built to run in production at scale. Rather than model quality, the team found that tool sprawl and missing context were the biggest obstacles to reliable agent behavior. The post details architectural patterns including dynamic tool scoping, schema-based query access, dedicated context handlers, and how Amazon Bedrock provides model flexibility, cross-Region inference, data residency controls, and prompt caching.
Qwen published Qwen-Image-2.1-Turbo on Hugging Face, a distilled version of its Qwen-Image-2.1 text-to-image and image-editing model. It uses the same 7B visual generation architecture but runs with only 8 denoising steps instead of the standard schedule, using CFG=1 and prefix KV caching to reuse text and reference-image context across steps. It loads directly via the QwenImage21Pipeline in Diffusers and supports the same resolution presets and aspect ratios as the base model, including editing tasks like turning sketches into photorealistic images.
Google released ML Drift, an open-source, cross-platform GPU compute engine for on-device ML inference, replacing the legacy TFLite GPU delegate. It unifies shader handling across OpenGL ES, OpenCL, Metal and WebGPU, adds 5D tensor support, and introduces stage-aware optimizations for LLM prefill/decode. It is already running in production in Chrome, YouTube Shorts, Google Photos, Meet and partner apps from Adobe and Snap.
AWS published a case study on Amazon Bedrock AgentCore payments, a managed feature that lets AI agents autonomously pay for services like model inference, one request at a time, using the x402 payment protocol. Incarna used it to let its agents pay BlockRun, a multi-provider inference router, for each individual model call, settling in USDC on the Base blockchain. AWS says Incarna cut integration time from a scoped two to three months down to three days and about 200 lines of code. During the beta, agents processed over 1,000 payments ranging from $0.001 to $0.05 per call.
Cloudflare built an AI agent system for its Managed Defense security service to handle the flood of alerts that overwhelm human analysts. Instead of one general-purpose agent analyzing everything, the system separates deterministic data collection from model-based reasoning and runs four specialist agents in parallel before a synthesis agent combines findings. The company says this reduces hallucinations and makes investigations auditable and reproducible.
Z.ai's GLM 5.3, a 753-billion-parameter mixture-of-experts model optimized for coding and long-horizon agentic tasks, is now available as a fully managed model on Amazon Bedrock. It builds on GLM 5 with improved coding benchmark results, standout cybersecurity capabilities (CyberGym score of 84.5), and deeper Bedrock integration including OpenAI-compatible APIs, prompt caching, cross-Region inference and service tiers. AWS also demonstrates using GLM 5.3 with the open-source Strix agent for authorized penetration testing of your own applications.
Google engineers describe how they implemented sparse spatio-temporal attention for video diffusion models on TPU v6e chips, converting the theoretical sparsity of the Sparse VideoGen (SVG) approach into real hardware speedups. Through a progression of kernel optimizations (full/boundary tile specialization, tile-size tuning, and mask-tile alignment), they reduced attention kernel latency from 96.37ms (naive sparse) to 32.76ms, a 2.40x speedup over dense Splash Attention on a single TPU v6e chip, using 75.6K tokens, 10 heads, and head dimension 128.
NVIDIA is releasing a new 64GB unified memory configuration of its DGX Spark personal AI supercomputer, available from partners like Acer, ASUS, Dell, HP, Gigabyte and MSI starting Oct. 23 at $4,999. The system runs up to 100-billion-parameter models locally on device, and two units can be clustered via NVIDIA Sync Cluster Assistant to pool memory to 128GB and support models up to 200 billion parameters, with up to 1.7x performance gains in NVIDIA's Qwen 3.8 27B test. It ships with the NVIDIA Agent Toolkit, CUDA-X AI libraries, Nemotron models and support for Ollama, vLLM and PyTorch out of the box.