Hopper
PulseAugur coverage of Hopper — every cluster mentioning Hopper across labs, papers, and developer communities, ranked by signal.
- 2026-05-12 product_launch Hypercubic launched Hopper, an agentic development environment for mainframes and COBOL. source
7 day(s) with sentiment data
-
HOPPER framework enhances graph sequence models with learnable hop extraction
Researchers have introduced HOPPER, a novel framework designed to enhance Linearized Graph Sequence Models (LGSMs). Unlike previous LGSMs that rely on fixed graph operators, HOPPER enables end-to-end learning of hop seq…
-
PhyAI engine unifies physical AI inference across edge and cloud
Researchers have developed PhyAI, a unified inference engine designed to streamline the deployment of physical AI models across various platforms, including edge devices and cloud environments. This single runtime aims …
-
NVIDIA GPU Warp Divergence Behavior Stable Across Architectures
A new paper analyzes warp divergence across NVIDIA GPU architectures, from Pascal to Blackwell. The research found that divergent paths serialize linearly with the number of paths, a behavior consistent since the Pascal…
-
AMD's MI500X Architecture Details Emerge, Featuring Shared MoE Descriptor
AMD has reportedly released details about its MI500X integrated system architecture, which features a TDM Mixture-of-Experts (MoE) descriptor shared across all experts. This architecture appears to be an evolution of th…
-
Nota AI releases 4-bit quantized Solar Open2 250B model for NVIDIA Blackwell
Nota AI has released a 4-bit quantized version of Upstage's Solar Open2 250B model, named Solar Open2 250B — Nota NVFP4. This new version utilizes Nota AI's proprietary quantization technology, specifically designed for…
-
SenseTime unveils "Future Lab" at WAIC 2026 with embodied AI and advanced infrastructure
At the 2026 WAIC, SenseTime showcased its "Future Lab" with four themed zones: Future Life, Future Intelligent Computing, Future City, and Future Assistant. The "Future Life" zone highlighted AI-native products like the…
-
US allows ZTE to buy Nvidia H200 AI chips, joining tech elite
The U.S. government has granted approval for Chinese telecommunications company ZTE to purchase Nvidia's H200 AI chips. This decision allows ZTE to join a select group of Chinese tech firms, including Alibaba, Tencent, …
-
NVIDIA touts Blackwell platform's performance per watt for AI infrastructure
NVIDIA is emphasizing performance per watt as the critical metric for AI infrastructure, especially with the rise of agentic AI and Mixture-of-Experts (MoE) architectures. The company highlights its Blackwell NVL72 plat…
-
TurboQuant AI compression sees community adoption, but hype cools
Four months after its announcement, Google's TurboQuant algorithm for compressing AI model KV caches has seen significant community adoption but with a more nuanced understanding of its capabilities. While Google has no…
-
Meituan trains 1.6T LongCat-2.0 AI model without NVIDIA hardware
Meituan, a Chinese company, has developed LongCat-2.0, a 1.6 trillion parameter AI model. Notably, the model was trained entirely on domestic AI chips, avoiding NVIDIA hardware. The model's efficiency stems from its Mix…
-
Jet-Long method boosts LLM long-context performance without retraining
Researchers have introduced Jet-Long, a novel method for extending the context window of large language models without requiring retraining. This tuning-free, zero-shot approach dynamically adjusts rescaling factors to …
-
Why GPUs Dominate AI Hardware Despite Specialized Alternatives
Despite the existence of specialized hardware like Groq's LPU and Cerebras' WSE, GPUs continue to dominate the AI landscape due to significant economic and structural barriers. These include the immense capital required…
-
MiniMax AI releases open-weight M3 model with 1M context
MiniMax AI has released its new open-weight model, MiniMax M3, featuring a 1 million token context window and advanced capabilities. The model utilizes a novel sparse attention architecture called MSA, which includes de…
-
NVIDIA Blackwell Systems Lead New Agentic AI Benchmarks
NVIDIA has set new performance records on the first agentic AI benchmarks, AgentPerf and Agentic AI Benchmark. The company's GB300 NVL72 system, powered by Blackwell architecture, demonstrated up to a 20x performance le…
-
Tokens per Watt to Dictate 2026 GPU and Cooling Decisions
The primary constraint for AI compute in 2026 will shift from raw processing power to efficiency, specifically tokens per watt. This is because inference, which now accounts for the majority of AI compute spend, is fund…
-
User optimizes DeepSeek v4 Flash for 200 tok/s on Hopper
A user shared optimization tips for running the DeepSeek v4 Flash model locally, achieving nearly 200 tokens per second on a Hopper system. By utilizing specific quants from Canada-Quant and patching the MTP code in vLL…
-
Stanford's ThunderKittens DSL optimizes AI kernel performance
A new article details ThunderKittens, a compact domain-specific language (DSL) developed at Stanford's Hazy Research Lab for creating high-performance AI kernels. The DSL aims to strike a balance between research produc…
-
Taiwan cracks down on AI chip smuggling to China
Taiwan has initiated its first formal crackdown on the illicit export of AI chips, raiding 12 locations and seeking three fugitives accused of document forgery and fraudulent declarations. This action is part of a broad…
-
Hypercubic launches Hopper agentic interface for mainframes
Hypercubic has launched Hopper, an agentic development environment designed for mainframes and COBOL. This tool allows AI agents to interact with legacy systems, enabling tasks like navigating TN3270, inspecting dataset…
-
Perplexity research shows NVIDIA GB200 excels at LLM inference
Perplexity has published research detailing how they serve large language models, specifically Qwen3 235B, on NVIDIA's GB200 NVL72 Blackwell racks. The findings indicate that the GB200 platform offers significant improv…