Jeffrey Welser

Keeping your data close to the GPUs is essential, so we offer VAST Data as part of our #GPUasaService AI Platform. Let us help you #StepontheGaaS and move your #AI strategy forward. #Fast #Secure #Private Dominic Ken Michael Bill Traci

10

1 Comment

In the fiercely competitive semiconductor industry, companies are in a constant battle to retain top talent because in the heart of every good engineer is a passionate problem solver, looking for their next challenge. For them, innovation is the currency of success. Yet companies like Qualcomm, Intel, and Broadcom attract tens of thousands every year by dangling lucrative compensation packages to lure them from competitors. Startup leaders don’t have those resources, so they have to work harder to attract the best talent and then fight like hell to retain them. But some of the most successful startup leaders I know take a different approach. Rather than viewing attraction and retention as a fight, they leverage the inherent qualities of their companies as their advantage. They use their nimbleness, bias toward innovation, and their collaborative cultures like superpowers to attract the right people and those usually end up being the reasons why people stay. Take Richard Fung, CEO of The Six Semiconductor, for example. I had a conversation with Richard and asked him to share his approach to motivating his team. His response was like a mini-masterclass on the topic of talent attraction and retention. Richard emphasized that a culture that encourages engineers to dive deep into the intricacies of their craft, understand the systems they’re building, and connect with the end customer is a successful culture. Richard has been using this approach and it's been paying dividends for The Six Semiconductor so in his honor, here are his six key concepts for creating this type of organization: 1. Foster a Learning Culture: Support ongoing learning and skill development beyond employees' core responsibilities. 2. Prioritize Customer Focus: Embed a customer-centric mindset, and encourage employees to enhance user experiences through their work. 3. Champion Problem-Solving: Empower employees to creatively solve challenges with a focus on customer needs. 4. Promote System-Level Understanding: Encourage employees to grasp the bigger picture of the company's offerings and their role within it. 5. Break Down Silos: Promote collaboration to ensure a comprehensive understanding of projects across teams. 6. Offer Meaningful Challenges: Provide intellectually stimulating projects that make a tangible impact on the company's success. What Richard highlighted is that success goes much deeper than getting great people to join. It's about unleashing the full potential of every individual on your team to drive innovation and help shape the company, and ensuring your team heads in the right direction together. I’m thrilled I got to learn Richard’s perspective and believe other leaders can benefit from hearing it, so I shared the video below. My "challenge question" to other startup leaders is this: what are you doing to foster this type of culture in your organization? Either share below or, feel free to DM me. #semiconductorindustry #startup #leadership

67

2 Comments

📍 Reporting from Denver, Colorado, Graham Morehead discusses Intel's groundbreaking OCI Optical Compute Interconnect. 💡 Why it matters: Speed of Photons: Photons travel significantly faster than electrons, allowing for rapid data transfer. Silicon Photonics: Intel's innovation integrates photons directly into silicon, seamlessly connecting with current AI technologies. Massive Data Access: AI processes require immense data access, often leveraging distributed compute units. 📊 The ability to gather and process data efficiently is crucial, and photons are the fastest way to achieve this. 🔍 Discover why this advancement is pivotal for AI's future! 🚀 #AI #Intel #SiliconPhotonics #OpticalComputeInterconnect #FutureTech #GrahamMorehead #Pangeon #ThePangeon #GPT #AGI #OpenAI

5

Jim Keller, former Intel, AMD, and Tesla engineer, aims to challenge Nvidia's dominance in AI with Tenstorrent's energy-efficient chip design. Tenstorrent's upcoming second-generation AI chip promises superior energy and cost efficiency, presenting a viable alternative to Nvidia's offerings. The innovative design of Tenstorrent's chip features independent decision-making cores, potentially reshaping AI chip technology across diverse applications. #AI #Nvidia #semiconductor #semiconductors #chip #technology #innovation #research #computing Tenstorrent NVIDIA https://lnkd.in/g6a3vAu4

31

It is always awesome when people recognize that what Tenstorrent is doing is really great.

22

Instead of scaling AI training to produce more accurate answers, chain-of-thought reasoning explore different avenues and provide better answers. This "think before answering" approach provides dramatically better performance on demanding tasks like math, science, and code generation, fundamentally boosting the intelligence of AI models without requiring additional training. By running over 70x faster than other solutions, Cerebras Inference allows AI models to "think" far longer and return more accurate results. As agentic AI becomes available and eventually widespread, the demands in ifnerencing hardware will increase by another 10-fold.

4

I am so proud of my students (current and former) who have been selected to receive the Best Student Paper Award for the 2024 IEEE Symposium on VLSI Circuits! Each year, the VLSI Circuits Symposium has presented a Best Student Paper Award to encourage the research of student projects that will lead the next generation. Only one paper is selected each year. We are honored to announce that “SPIRIT: A Seizure Prediction SoC with a 17.2nJ/cls Unsupervised Online-Learning Classifier and Zoom Analog Frontends” by Adelson Chua*, Aviral Pandey*, Ryan Kaveh, Sina Faraji, Justin Doong, and Rikky Muller has been selected for the award. Congratulations to the whole SPIRIT team, and particularly to co-lead authors Adel and Avi who gave an outstanding talk at the conference! #VLSI #SSCS #Brain #EECS #UCBerkeley SPIRIT is an SoC integrating an unsupervised online-learning seizure prediction classifier with eight 14.4μW, 0.057mm2, 90.5dB dynamic range, Zoom Analog Frontends. SPIRIT achieves, on average, 97.5%/96.2% sensitivity/specificity, predicting seizures an average of 8.4 minutes before they occur. Its classifier consumes 17.2μW and occupies 0.14mm2, the lowest reported for a prediction classifier by >134x in power and >5x in area.

217

19 Comments

How does TechSovereign Partners stand out as consultants and contractors to the semiconductor industry? Because we don't bring on experts who "worked" in the industry. We are building a team of technologists who helped build it. Each one of our technologists has 25 years and more developing the tools and processes that the industry runs on. These processes work because we helped get them working in the first place. And we all have something in common. We just. love. solving. problems. We have an incredible amount of firepower for crushing out defects or optimizing processes, or for full integration scheme implementation. And in many cases, the engagement costs less than replacing a filter cartridge. We encounter clients all the time who fought with an issue for weeks (even months) until we engaged and in many cases found the answer within days. Even within hours. Ok, I'm not exaggerating, even within minutes. We say, let the easy stuff be easy. And we can make a LOT of things easy. It's what we're for. #ProcessEngineering #DefectReduction #ManufacturingExcellence #ProblemSolvers #SemiconductorIndustry #SemiconductorManufacturing #SemiconductorSolutions #AdvancedPackaging #PlatingProcesses #SemiconductorConsulting

7

Thinnest gold leaf is just 1 atom thick, and an exciting 2D material. Just far too interesting, not to share, as this is about delocalized bonding, and not a packed & stacked construct. https://lnkd.in/gUxtBpPq

2

✨ Get over 3x perf improvement for #LLM throughput - TensorRT-LLM Multiblock Attention on NVIDIA HGX H200 boosts long-sequence performance without impacting time-to-first-token. Read how ➡️

5

With NVIDIA NIM and NIM Agent Blueprints, medical researchers across the public sector can jump-start their adoption of state-of-the-art, optimized AI models to accelerate their work. 🔗 https://nvda.ws/3NnkMYy Learn how #NVIDIAInception startups Abridge & ConcertAI leverage NVIDIA AI solutions to improve efficiency, reduce administrative burdens, and advance medical research and patient care. 🩺 Abridge ConcertAI

4

1 Comment

AI workflows are driving demand for memory and storage to prepare data, train the model and create useful inference engines.  This is driving announcements of storage and memory products that support the various elements of these AI workflows.

51

4 Comments

Get a deeper look into the results submitted by NVIDIA in #MLPerf Inference v4.1, highlighted by the first ever submission using the NVIDIA Blackwell architecture, increased performance using the NVIDIA H200 Tensor Core GPU, and more.

1

Got a chance to discuss TCG Key Per I/O capabilities and its possible use cases with confidential compute at the Confidential Compute Consoritium's TAC call yesterday. Please see the link below. The topic is covered from around 20 mins into the video and runs till about the 1-hour mark. https://lnkd.in/gvupy8iP

20

Signal processing for SWaP constrained ISR/EW applications in 35mm package, 4K DSP engines. Versal Premium Gen2 https://lnkd.in/gfnqHtDZ #xilinx #dsp #signalprocessing #amd #aerospace #defense

38

At the AI Hardware Summit, I was struck by how the competition is heating up. As we sprint toward the “Age of AI everywhere” it feels like a footrace with a handful of chipmakers vying for first place, unafriad to throw an elbow at their competitors. But one interaction at the Summit seared my consciousness, renewing my faith that as we push toward artificial intelligence everywhere, AI can elicit the best from everyone if it stands for 'Authentic Individuals’. Here’s some background: I’ve followed Enfabrica since I first noticed Rochan Sankar and Shrijeet Mukherjee tackling the efficiency challenges of AI with an approach different than most companies. Building a new type of NIC seemed to be a Herculean task and a journey that could only be undertaken with the right people on board. As I got to know them, I discovered they’re just as concerned with building a world-class company as they are with solving the networking sector’s biggest problems. In five short years, they’ve assembled a team that stands shoulder-to-shoulder, technically, with the most successful companies while fostering a culture that values building things. They’re building groundbreaking technology, strong relationships with customers, and camaraderie with each other. This is only possible when people see excellence in their leaders, and at the Summit, Rochan demonstrated excellence in leadership by being authentic. He presented to a full ballroom with poise and confidence, and as I approached him in the hall after to congratulate him, expecting we’d talk about how it went, he locked eyes with me and asked, “How is your son’s recovery going?” This wasn’t just politeness. He KNOWS the depth of that question. We’ve talked many times about how challenging it is to be the parent of a child who has serious, lifelong disabilities and knew that our entire household was in disarray for months after our son underwent surgery with a painful recovery. During what should have been HIS triumphant post-presentation moment, he asked the question with all the sincerity in the world and it resonated on such a deep, human level. On the flight back to Phoenix, I reflected on how the AI sector has become like a 'gold rush,' with new players entering the market every month. This constant influx makes it challenging for talented individuals who want to make their mark in AI to navigate the ecosystem. The greatest hazard in one’s career is working under the wrong leadership, because people don’t work for companies—they work for people. My interaction with Rochan reminded me that there’s no better indicator of character than authenticity. If you can see how a leader treats people, you’ll know everything you need to know about that person. Regardless of the amount of success Enfabrica achieves, working for leaders with this level of authenticity should be what EVERY engineer and leader should aspire to have in their career. #semiconductorindustry #artificialintelligence #authenticindividual

32

2 Comments

Our President, Justin Kinsey, attended the AI Hardware Summit last week and told our team about the incredible speeds AI chipmakers are pushing their latest processors. It was exciting news, but it also raised a few flags because responsible, efficient AI is one of our core beliefs. AI’s power consumption is a growing issue, projected to reach 1.5%-3% of the world’s energy usage by 2029, roughly equal to the annual energy consumption of Argentina, the Netherlands, and Sweden combined! GPUs, CPUs, and other accelerators often get blamed, but there’s more to the story. Without getting too technical, efficient data center operation relies on voltage regulators to adjust and stabilize the output voltage as current spikes to make sure the processors get the right power to run at higher speeds without disruption. These systems face 'spiky' power demands, especially in AI workloads where chips experience big fluctuations between idle and peak demand, like when training new models, or when millions of users rush to try a generative AI platform. So power delivery and management systems have to handle wide ranges of load which is a difficult equation to balance and often requires building bulky systems that have tradeoffs. But this challenge creates opportunity for innovation and companies like Empower Semiconductor leading the way. Empower aims to replace traditional power-management solutions with an integrated voltage regulator (IVR) that consolidates the entire regulator onto a single chip. The IVR can be positioned close to or even beneath the processor by incorporating key components like magnetics and passives, optimizing power efficiency and reducing energy loss. Their latest family - the EP71xx - which they showcased at APEC, integrates up to four voltage regulators in one package and can be used with CPUs, GPUs, and SoCs. It can even be used with DSPs inside pluggable connectors and active copper cables, which will impact power efficiency in other areas of the data center. As their CTO Trey Roessig put it during the launch, “By looking at power delivery as a whole, we’re able to optimize the power system rather than individual components.” With key governments worldwide prioritizing green energy, and the data center market expected to have a double-digit CAGR for the next few years, solving this problem is more important than ever and This tech wave brings new opportunities for power engineers. I’ve worked with PMIC clients for decades but have never seen such demand for top talent with analog/mixed-signal skills and digital control. Even our clients working on accelerators, networking chips, and interconnect technology look favorably on engineers with at least some power management skills. For engineers and technical leaders who are curious about what other interesting challenges might be out there for you to work on, I’ll be happy to tell you. Just reach out 😉 #semiconductorindustry #powermanagement #artificialintelligence

31