Tao Dong

Introducing Nemotron 3 Super, our new open 120B/12B hybrid Mamba‑Transformer MoE model built for real agentic reasoning at scale. 🧠 Latent MoE + multi‑token prediction + hybrid Mamba‑Transformer = up to 5x higher throughput and 1M‑token long‑context reasoning for multi‑agent systems. 🦞 All of this comes together to create a model that is well suited for long-running autonomous agents. Full parameter checkpoints for Nemotron 3 Super are available on Hugging Face and through NVIDIA NIM. 🔗https://bit.ly/4ukdMQl

2

Great engineering does not happen in silos. At Roku, collaboration helps us grow, push boundaries, and deliver experiences that reach millions of streamers every day. #WeAreRoku

1

🚀 Static Timing Analysis (STA) Engineer: The Unsung Heroes of Silicon Success! 🚀 STA engineers play a crucial role in ensuring that complex ASICs and SoCs meet stringent timing requirements across all process, voltage, and temperature corners. 🔍 What makes an STA Engineer critical? Expertise in timing constraint creation and management (SDC) Mastery of STA tools like Synopsys PrimeTime, Cadence Tempus, and TimeVision Deep understanding of setup/hold violations, clock domain crossings (CDC), multi-corner multi-mode (MCMM) timing closure Ability to analyze timing reports and collaborate with RTL, synthesis, and physical design teams to fix timing issues Strong scripting skills in Tcl, Perl, Python for automation and timing data parsing 🎯 Industry Demand & Challenges: Experienced STA profiles are quite rare, making them highly sought-after in semiconductor companies. Many effective paths include transitioning from RTL design verification or physical design backgrounds with focused STA training. STA engineers are key contributors to silicon signoff, power-aware timing closure, functional safety timing checks, and ensuring product reliability. 💡 Career Tips for Aspiring STA Engineers: Build a solid foundation in RTL and digital design fundamentals. Gain hands-on experience with STA tools and constraint scripting. Understand cross-domain timing challenges and low-power design impacts. Continuously collaborate with multi-disciplinary teams (synthesis, PD, verification). I If you’re an STA professional or looking to break into STA, share your journey or tips below! For recruiters seeking STA talent, investing in training can pay off massively by enabling strong internal STA capabilities. #StaticTimingAnalysis #STAEngineer #ASICDesign #PrimeTime #CadenceTempus #TimingClosure #TclScripting #SemiconductorJobs #VLSI #EDAtools #TimingVerification #RTLDesign #PhysicalDesign

9

NVIDIA Reportedly Acquires MediaTek, MediaTek Clarifies Foreign media outlet TechRadar reported that NVIDIA and MediaTek have deeply collaborated on the GB10 Grace Blackwell chip, combining MediaTek’s CPU and memory expertise with NVIDIA’s GPU design capabilities. This partnership reflects advanced technological integration, particularly in system-on-chip design, where NVIDIA has tightly integrated its Blackwell GPU with industry-standard interfaces. The strong collaboration between the two companies has sparked market speculation that NVIDIA might acquire MediaTek for approximately $73 billion (about NT$2.2 trillion). However, MediaTek officially denied these acquisition rumors, stating they are not true. Despite the rumors, MediaTek’s stock rose intraday by NT$40, a 2.89% increase, closing at NT$1,425. Other related market news includes a flat close for TSMC amid active post-market trading and a surge in electronics stocks led by MediaTek. Investor sentiment was boosted by weak U.S. employment data and an increased expectation of rate cuts, benefiting Taiwan’s stock and currency markets, which saw notable gains. #Semiconductors #Taiwan https://lnkd.in/gHh9zFu5

3

#Autonomoustrucking, built for the real world. 🚛 We caught up with Shawn Kerrigan at #CES 2026, COO & co-founder of PlusAI, on how its SuperDrive virtual driver is advancing level 4 #autonomous freight. With #NVIDIA Cosmos and DRIVE Hyperion, #PlusAI is accelerating the path to factory-built, driverless trucks. 🛣️ #Automtive #AutonomousDriving #AutonomousVehicles

2

The strategic intent behind the OpenClaw push is to draw in mainstream users to maintain these agents, helping cloud providers monetize computational power and convert them into a paid user base. However, is this truly a necessity for the general public?

5

Introducing the M5 The M5 chip brings significant advancements in performance and efficiency across various applications. 1. Massive GPU / AI performance boost The M5 features a next-generation 10-core GPU architecture, with each core equipped with an integrated Neural Accelerator. This results in over 4× the peak GPU compute performance for AI tasks compared to the M4. 2. Stronger CPU and enhanced Neural Engine With up to a 10-core CPU (6 efficiency cores + 4 performance cores), the M5 offers approximately 15% faster multithreaded performance over the M4. Additionally, the Neural Engine has been upgraded to 16 cores, enhancing performance for on-device AI and intelligent features. 3. Higher memory bandwidth / unified memory improvements The M5 provides around 153 GB/s of unified memory bandwidth, which is approximately 30% more than the M4. This unified memory architecture allows all components (CPU, GPU, Neural Engine) to access a larger memory pool, enabling larger AI models to run entirely on-device. 4. Graphics & rendering enhancements The improvements in the GPU also boost graphics performance, including ray tracing capabilities. Apple claims up to a ~45% uplift compared to the M4 in ray-traced workloads, enhancing visual fidelity, rendering speed, and smoothness in demanding 3D and graphics applications. 5. Deployment in new hardware + energy efficiency / environmental goals The M5 will power new devices, including the 14-inch MacBook Pro, iPad Pro, and upgraded Vision Pro, all of which are available for pre-order.

7