Welcome to
the Forum

Day one of Edge Field Day 3 kicked off with a presentation from VMware. This sage of enterprise cloud and expert in workload automation entered the presentation room with a challenge for us to forget about all of that and focus on the edge.

Chris Taylor, Product Marketing Manager and Alan Renouf, Technology Product Manager for the Software Defined Edge group, bring fantastic technical chops to the table. Alan has an extensive background in workload automation from the data center and Chris brings a unique perspective in manufacturing as a mechanical engineer - as well as seasoning in the security arena - to the table.

So what is the state of edge from a VMware perspective? VMware has created an edge compute stack leveraging an intelligent overlay of VMware VeloCloud SDWAN and SASE Security software and the VMware Telco Cloud Platform driving 5G, Fixed and LEOS network support. This software stack provides a horizontal foundation for deployment across edge environments and vertical use cases. Top markets extend across manufacturing floors, retail, energy generation and distribution, logistics operations and more. What drives customers to edge deployments? Too much data sitting at the edge, unreliable or inefficient networking from edge to cloud, and challenges in moving data due to regulatory requirements like data sovereignty are key customer pain points met by VMware customers.

And with this vast market and clear challenge driving deployment, you’d expect that edge implementations would be scaling like a hockey stick. But Chris and Alan pointed to key challenges for deployment sharing a slightly stunning stat from World Economic Forum that 70% of companies investing in industrial 4.0 projects get stuck in the pilot phase.

The challenges that many of these companies face include oversight of the scale and differentiation of myriad edge devices, use cases, and support required. Underpinning all of this is the noted difference in edge as compute that cannot be managed with the exact approach as cloud. After all, data centers are centralized…they’re rich with IT support, and they are, well, clean and orderly.

So what is VMware’s view on management of this wild west? Alan and Chris stated the approach needs to start with a change in workload management away from push architectures of cloud to pull mode. They signaled that the world of mobile app updates provide a parallel to how they’ve architected their edge stack to perform where management checks for updates and pulls updates when there is network capacity and compute cycles to load.

So how is VMware differentiated? Their edge optmization accounts for real time apps and simplified operations and is based on a flexible platform that grows with edge implementation and breadth of apps deployed. They’ve also built in support for both VMs and K8s with zero-touch orchestration delivering provisioning via preconfigured edge HW authorization and profiles established by an edge admin residing in a central location. This allows for plug-and-play deployment, which allows for an edge device to pull desired apps once plugged in without on-the-ground IT support. Edge admins maintain all workloads across edge locations with monitoring and downloads of components to match desired end state, all delivered by VMware’s Edge Cloud Orchestrator (VECO).

So who is tapping VMware for edge management? This is one area that I wanted to hear more as VMware KNOWs the enterprise. While a breadth of customers was not provided, the one customer example referenced was a big one – Audi. Audi is looking to transform their factory floors for Industrial 4.0 – details here. 

What’s the TechArena take? With solutions like ESX and VMware’s SDWAN configurations, we expect those IT admins who have trusted VMware in the cloud may also trust VMware at the edge, all questions about the current state of the company now as part of the Broadcom empire notwithstanding. The solutions discussed provide a practical approach to automated edge oversight, and I’d hoped the speakers would take more credit for what is VMware’s ace in the hole: understanding of customers at a very deep level. I would expect that customer discussions on this edge solution would reflect bringing the entire heritage of VMware to the table, and I expect to see other name brands like Audi to join as notable customer deployments for the edge stack solution.

Michael Maxey, Manny Calero and Jason Grimm from Zededa shared their vision for the edge today at Edge Field Day 3. Zededa was founded in 2016 with a mission to deliver a cloudlike experience for the edge and most notably began delivering Edge Application Services last year.

Their investors have guided their initial use cases, and with Chevron and Porsche on board, they’ve targeted oil and gas implementation and automotive as well as agriculture, renewable energy, manufacturing and more.

Unpacking the solution, Zededa delivers an open-source operating system based on Linux, a cloud control for edge app management, and a global marketplace for partner applications. They also have a broad ecosystem of supported infrastructure they support from the industry from the usual suspects. Expounding further, their cloud gives central management of all apps at the edge with built-in security, policy-based management, and pull updates for edge to address the unique environmental challenges of managing this fleet of devices.

The LF Edge operating system supports workloads across virtual machines, containers, K8s clusters, VNFs, CNFs, and custom runtimes, giving incredible flexibility to organizations. It is uniquely designed for the edge, providing unique security control given the challenges of maintaining data protection with remote devices. It’s an API-only system - meaning there is no log-in at the edge, which seems like the right choice given relative technical skillsets in edge environments.

The configuration has flexibility for Linux distributions as well as embedded hypervisors and supports partitions to allow for staging of workloads prior to activation. This also means that configurations can be rolled back to ensure reliability and continued uptime.

Zededa has invested in unique edge control in the automotive arena, and Manny walked us through details on a massive deployment they’ve driven with the second largest automotive manufacturer on the planet.

The challenge within automotive? We’ve covered the electrification of automobiles on the TechArena through Robert Bielby’s series, and there are many targets for “data center on wheels” deployment across system control, infotainment, autonomous control and more. Zededa has focused on system control aspects of the automobile to drive secure transfer of data for service activation at the dealership. They’ve delivered a platform that offers the scale required to support millions of vehicles across 70,000 dealerships with an open-source foundation enabling the customer to avoid lock-in to one specific application.

Zededa delivers this configuration with support for all major platforms across x86, ARM and RISC-V. With this configuration, you can expect dealers to improve customer support while increasing high-margin service revenue and grow likelihood for repeat sales in the process.

Zededa also walked through a rail freight use case, and Jason walked us through the details in this space. Their solution has achieved 12 fortune 500 deployments with tens of thousands of nodes in production with zero trust oversight and a host of connectivity options for delivering the solution. The solution was deployed as a bungalow connected via cellular and satellite connected to edge cameras capturing images of train movement. This resilient solution protects against connectivity failures and provides central control via the Zededa cloud controller.

So what’s the TechArena take? What marked Zededa’s session was the scale of deployments they’ve been able to achieve, the real world use cases that they described, and the broad scale of options customers have at every layer of the stack. I’m impressed with their team’s deep knowledge of customer challenges and ability to transform this knowledge into deployed solutions, solving real world business challenges. If companies are evaluating edge orchestration and application management alternatives, Zededa should be on the short list.

‍

OnLogic capped Edge Field Day 3, and they brought a fantastic story to the event. Lisa Groeneveld and her team described how OnLogic is not the typical VC-funded startup. They are a 20+ year old company headquartered in Burlington, Vermont and led by a husband and wife team that are co-founders of the firm.

They drive infrastructure anywhere – delivering industrial computers, rugged computers, panel PCs, and edge servers to a wide variety of edge deployment use cases.

They specialize in expert technical consultation and built-to-order solutions. They listen to their customers and work to deliver solutions that meet what customers require most. The team shared details of the actual infrastructure. Starting with the far edge, they break down infrastructure requirements as devices and edge servers, and they look at landing infrastructure in these environments to meet temp, dust, space, and vibration realities that exist in these environments.

To tell the story of their infrastructure, the team utilized a series of deployment examples including agriculture. OnLogic’s customer sought foundational compute for robotic food harvesting. Enter the Karbon 800 series computer offering a fanless design, extended temperature range, and military grade shock and vibration tolerance. This computer runs an embedded Intel i9 processor and GPU expansion ports for AI acceleration and comes in three configuration options. The Karbon fueled the robotic solution the customer required to improve crop yields and save costs.

The team shared an example in the mining sector with a partner called Flasheye. Flasheye was seeking to tap visualization to reduce belt downtime to mining operations in horrifically rugged environments. Their customer operated a mine in Northern Sweden whose belt experienced spillage and required human intervention increasing safety risk. With the Flasheye solution, powered by an OnLogic Karbon 400 series platform. This computer can drive uptime up to -40-degree temps with an efficient Atom processor and flexible I/O that supported the visualization integration desired by the customer. 

The conversation next turned to a steel mill use case where intense heat and grime and other particulate are a constant challenge. The customer, Steel Dynamics, sought to deliver monitoring of the fabrication environment. Enter the OnLogic Tacton TC401, an all in one panel PC, PCs that are built into the panels of industrial equipment. This PC operates from temperatures between -20C to 70C and features Intel 12^th Gen CPU with a ModBay I/O expander supporting wifi and cellular connectivity.

The team rounded out use cases with a residential application, more specifically in hi-rise elevator control. Nantum AI brought OnLogic a challenge for an interesting use case – how to deliver AI control to elevators to predict a power outage and avoid stranding people on elevators. Enter the Helix Industrial Edge Computer featuring a 10^th-12^th Gen Intel embedded Core processor, solid state design, and operating temps between 0C-50C. The Helix offers slightly less ruggedness than the other solutions mentioned making it perfect for a building deployment.

Finally, the team walked us through their partners. This started with Intel, which was no surprise given the continuous refrain of Intel processors across their fleet. While processors begin the collaboration, OnLogic is also tapping the OpenVino toolkit to optimize software for their targeted solutions. The depth of this partnership, given my own personal history, comes as no surprise. Intel has invested deeply in embedded, IoT, and edge optimization including ruggedization of long-life CPUs. OnLogic’s depth of collaboration extends to AWS. OnLogic is providing infrastructure for AWS Greengrass. The company also extended their collaboration discussion with software providers RedHat, Avassa, and GuiseAI.

What’s the TechArena take? If I ever head back to industry, OnLogic represents the type of culture I’d want to be a part of. This company is led with integrity, and my guess is that extends to customer and partner relationships. But beyond that, they deliver the meat and potatoes of what is required at the edge – computing tailor designed for rugged environments with the core capabilities needed to ensure long life, continued uptime, and feature options to fuel desired workloads. They’ve built a reputation for both broad product configurability as well as trusted reliability. We expect their business to continue its growth path as edge use cases advance.

‍

The all-too-familiar hexagonal red STOP sign with white trim around the border (at least in the United States) instructs the driver to come to a complete stop, look in both directions, and yield to traffic as required before proceeding. If the stop sign had the misfortune of being vandalized, perhaps a casualty of graffiti, the human driver would most likely be able to look beyond the graffiti to recognize that it is a STOP sign, and take the appropriate action.

It could be said that this behavior is analogous to “inductive priors,” a term used in the science of neural networks, where logical assumptions are made based upon the input data – even though some of the input data may be corrupted or occluded in some manner. In this case, the driver has seen enough stop signs and has enough cues in this sign to recognize it as a stop sign, despite the fact that it has been defaced.

Convolutional Neural Networks (CNNs) typically employ an equivalent type of inductive priors. This class of neural networks assumes that nearby pixels from a camera image are related (typically a reasonable assumption). This class of neural net also assumes that the importance of the different parts of the image is weighted similarly (no one part of the image is any more important than another part). The result is that for CNNs, a relatively limited amount of training is required before the network starts to demonstrate significant levels of accuracy. However, even when fully trained, the inherent accuracy of the CNN is limited when compared to other classes of neural nets, which means it might “jump to conclusions too quickly” to make a judgment and make mistakes.

Slightly off topic for a second, the intelligence of different breeds of dogs is typically measured by how many times a dog of a certain breed typically has to hear a command before that breed of dog can recognize a command, or is trained. I am going to draw the analogy that different classes of neural networks are like different breeds of dogs in both the manner as to how long it takes before they are trained and the accuracy by which they are repeatedly and accurately able to recognize a command from different voices.

While CNNs have been the mainstay deep neural network (DNN) employed to address perception in ADAS applications, as of late, Vision Transformers (ViT) have been gaining significant interest within the ADAS design communities, primarily due to their improved accuracy in object classification over CNNs. In the case of the ViT, however, the increased accuracy comes with the need for a very extensive set of training data. The ViT also demands increased compute performance over that of the CNN during deployment.

ViTs have very low inductive biases. While this is one of the factors leading to improved accuracy (this form of neural networks is less prone to “jump to conclusions”) a significantly larger training set is needed before the ViT can achieve a level of accuracy equivalent to that of the CNN. Also, because the ViT does not assume the importance of adjacent pixels, the full frame with all pixels present must be in place before ViT calculations are performed. This, in turn, is one of several factors that drive up requisite compute performance during deployment/inference.

Increasingly, image sensors with higher resolutions are being employed in the automobile as they can provide a greater amount of information about the surrounding environment of the vehicle as compared to their lower resolution (Mpixel) counterparts. Sensors that provide more information enable objects to be more readily detected from greater distances. This, in turn, enables the vehicle to operate autonomously at higher speeds. As an example, a lower resolution camera may provide only 5 pixels, which reflects an object that has been detected off in the distance. This, however, is an insufficient amount of information to accurately recognize the image and plan the appropriate action. The higher resolution camera, on the other hand, may provide 100 pixels of that same object, enabling the ADAS perception engine to readily recognize that object off in the distance as a pedestrian, cyclist, or stationary object providing more time to the autonomous vehicle to take the right action.

While the CNN doesn’t require the full frame of the image to accurately detect the image, the ViT does. This further exacerbates the computational requirements of the ViT due to the greater amount of data within the frame that must be processed from the higher-resolution camera image in the same time period.

Just as some breeds of dogs have different propensities towards being trained, neural network architectures can be viewed as being on a spectrum of inductive biases from strong to weak. CNNs are on one end of the spectrum while ViTs are on the other end. 

As typically seems to be the case, there is no free lunch. The increased power consumption of the ViT driven by the inherent higher computation and the required increased training datasets and training times represent meaningful recurring and non-recurring costs. Hybrid architectures are emerging that combine CNNs and ViTs into one architecture, leveraging the inherent strengths of each neural network. These architectures strike a middle ground between CNNs’ and ViTs' inductive biases, finding a balance between the learning flexibility and accuracy of the ViT while reducing the amount of training required.

While the focus has been on comparing CNN vs. ViT, similar debates and tradeoffs are ongoing in other areas wherever AI is being employed. To that end, this is why it is paramount that upgradable AI architectures are deployed in the field to ensure that the optimal neural network can be deployed as it is unearthed and thus avoiding the wrong response to a STOB sign.

The co-founders from TSecond, Sahil Chawla CEO, and Manavalan Krishnan, CTO, were on hand at Edge Field Day 3 to showcase Bryck and Bryck AI, black box edge devices chock full of storage, and in the case of the AI flavor, AI accelerators.

These solutions are created due to large stores of data being generated across the edge, and in rising cases, driving AI inference at the edge eliminating the need to move large data stores to the cloud.

We have covered rugged storage platforms for the edge on TechArena, and some of the vertical targets sounded very familiar – media, content delivery, defense to name a few. However, what made my ears perk up was a strong focus on aerospace and deep investment from Boeing in this growing organization. 

Let’s go under the covers. Manavalan took us on a detailed tour of the hardware, describing that the standard Bryck is a storage platform with up to one PB of storage running 40GB/sec throughput with 256 parallel PCIe Gen 4 data lanes and integrated security capabilities.

A Bryck requires 240-800W to run, bringing it in as energy efficient vs. storage alternatives. The platform promises to be faster and lighter than competitive offerings as well, while promising rugged resiliency for edge environments, hot pluggability, fault tolerance, and - due to low weight - cost effective to ship to edge locations. TSecond claims less than $100 to ship anywhere in the US – which matters when you consider the scale of edge device delivery in this growing market and you consider that competitive solutions weigh in at up to 500 lbs vs. 20 lbs for Bryck. Bryck can also scale with a multi-Bryck configuration as well as Bryck Mini with various capacities.

Bryck integrates the expected storage software including support for AWS, GCP and Azure for easy integration. They also offer Data Dart, a collaboration with Equinix to accelerate data transport to the cloud utilizing physical Bryck upload with movement of data from Equinix to the cloud provider of choice.

I was delighted to see Avassa back at Edge Field Day showcasing the latest technology that the company is delivering for edge application and orchestration oversight.

Avassa is a software only, cloud based, Control Tower platform for edge that is container first in orientation. Container first is a mantra at the company as Avassa called out the trend that “cloud out” is trending towards VMs eventually being sunset out of existence and “IoT in” seeking containers eventually sunrise as the de facto standard.

Carl Moberg, CTO of Avassa and a TechArena innovation voice, said that full elimination of VMs will be difficult due to pesky Windows-based applications that continue to require support and are unfriendly in native containerized environments.

So what is Avassa doing to support this continuum? They noted that open-industry standard frameworks for integrated VMs and containers is required that supports containerization of VMs for legacy apps.

Avassa delivers encapsulation through a stack containing a container image encapsulating a VM image and start script sitting on top of a container runtime using KVM with integrated health probes. While their container orchestrator cannot natively manage VM updates, they have developed means for this management and demonstrated this live in the session.

Their management extends to leading edge deployments of AI models or web assembly (WASM) containerized alongside Linux-based modern applications. This natural containerization of AI models and web assembly makes sense given that the latter have been primarily developed by born-on-the-cloud organizations by the very same engineers in many cases that built cloud native computing models.

The Avassa Control Tower demonstration highlighted ease of control of edge applications as well as depth of information collected by the Avassa oversight including the aforementioned health probes. They demonstrated migration of a VM-based service to a containerized alternative without disruption of service.

So what about Avassa’s decision to control with Docker runtime with KVM vs. K8s? This was a point of contention in last Edge Field Day’s presentation from the company, and Carl provided a helpful insight on market traction, stating that many companies from mid-market to hyperscalers have embraced this configuration for far or leaf edge environments where Avassa is targeted. He admitted that when edge starts looking like a data center in control and complexity, K8s may make sense, but that this is not Avassa’s target. This makes sense given the leaner stack offered by the Avassa solution for the constraints of edge infrastructure.

Carl and Fredrik moved on to management of applications in offline edge scenarios. Once edge devices are up and running with Avassa’s Control Tower, how are applications managed, and what happens when facing a scenario of upstream outages? The Avassa team provided a demo with a literal cut cord which made the Edge Field Day delegates quite twitchy. After all, many are former or current IT administrators. Carl pointed out that edge requires a different approach to cloud environments, which relies on failover in a situation like this.

How Avassa has managed this is to rely on a downstream admin with unique control of nodes at the site. This control creates an updated stack at the edge disconnected from central Control Tower, and customers choose how they manage the complexity of central workloads vs. local deployments once edge resources come back online.

So what is the TechArena take? Avassa has delivered a compelling central control across device orchestration and application management. I think they’ve made some tough choices on how to manage edge environments with the efficiency and control required for these environments, and I think they’ve chosen wisely for delivery of services within the far edge. For customers looking to deploy containerized environments, they’re a terrific option and should be evaluated by enterprises across industries. To learn more about Avassa, visit their site and check out TechArena’s content featuring the company.

Check out our in-depth look at key takeaways from AI Hardware and Edge AI Summit, highlighting how advancements in AI infrastructure, acceleration, and connectivity are transforming the future of computing.

I love talking to Letizia Guiliano. Letizia drives product marketing and management at Alphawave, but her passion for silicon design and standards innovation is infectious. In our latest chat on my In the Arena podcast, I spoke with her about the game-changing innovations required for AI advancement across connectivity and broader silicon foundations.

Letizia shared a fascinating view on how Alphawave is pushing the boundaries of high-speed interconnects to support the ever-growing demands of AI infrastructure, and how new competition is coming to the connectivity arena in this growing market.

Letizia and I have been discussing chiplets since last year, and last week’s chat highlighted how the designs are revolutionizing semiconductor implementation by improving both performance and efficiency. Unlike traditional monolithic chips, chiplets allow designers to break up complex components into smaller, interconnected pieces, optimizing for power, performance, and cost.

I challenged Letizia on progress towards an open chiplet industry, and she explained that while Alphawave is focusing on open standards to ensure interoperability between different chiplets and systems, current designs focus on bespoke customer solutions. There is still challenge in the market for standards-setting as well as tension that some players benefit more from an open industry than others.

One area that is advancing rapidly with standards delivery is scale-up fabric technology. With AI workloads growing at an unprecedented rate, moving massive amounts of data quickly and efficiently between processing units is a key challenge. Letizia was particularly excited about the industry’s progress in this space, highlighting that the new Ultra Acceleration Consortium is advancing standards that will enable interoperability across multiple company’s accelerator technology. Today, this area of technology is largely controlled by NVIDIA GPUs and NVlink serving as a barrier to entry to other accelerator alternatives.

We also touched on the importance of scalability. As AI models like GPT-4 and LLaMA continue to grow, the ability to scale up infrastructure to support these models is becoming increasingly critical. Alphawave is aiming to provide scalable solutions that not only meet today’s needs but are also future-proof, ensuring that data centers can easily expand to handle tomorrow’s AI workloads.

Letizia shared that the future of AI isn’t just about more powerful models, but about smarter infrastructure. And with Alphawave’s commitment to driving forward innovation in both connectivity and standards embracing design, the company is on track to be an important player in the future shaping of AI silicon.

For those who want to dive deeper into the details of Alphawave Semi’s solutions in this space, I highly recommend tuning into the full interview and visiting their website.

For those of you who follow the TechArena – or the technology landscape – you’re likely familiar with the name Lisa Spelman. Former head of all things Xeon at Intel, Lisa has driven tens of billions in business creation in her career as well as gaining real-world experience in the IT realm at the company.

Her knowledge of both the data center arena and of the large hyperscalers is unquestioned, and we at TechArena were excited by her recent move to take on leadership as CEO of Cornelis Networks as the company seeks to drive their technology into the AI landscape. Cornelis is the provider of Omni-Path fabric technology, a competitive fabric to InfiniBand that heretofore has had great success in HPC clusters across the globe.

I finally got to catch up with Lisa on her vision for the company at the AI Hardware and Edge Summit about how she is going to harness Cornelis technology to compete squarely with NVIDIA’s InfiniBand technology as a fabric foundation for AI training clusters. Lisa did not disappoint. We dove into the future of AI scale-out and how Cornelis is leading the charge with their Omni-Path solutions with an approach to maximize performance, scalability, and interoperability for data centers, ensuring that AI models can scale effortlessly across different environments.

Lisa highlighted how Omni-Path provides the foundation for building out next-gen AI infrastructure, making it easier for businesses to scale their AI workloads without running into bottlenecks of other technologies like Ethernet. She shared that Cornelis has already seen substantial traction with customers who are looking to optimize their data center operations for AI. One of the key takeaways for me from our conversation was how AI scale-out is moving beyond just a cloud discussion. Companies are increasingly interested in hybrid models that balance the power of cloud computing with the control and security of on-prem solutions.

What’s the TechArena take? If you’re keeping an eye on AI infrastructure, you’ll want to closely watch what Cornelis Networks is doing. Cornelis is helping companies deliver to the hybrid vision, and their goal of future-proofing AI infrastructure and providing an alternative technology to an all-NVIDIA solution is something that businesses can't afford to ignore. Cornelis’ focus on true interoperability with Ethernet and Ultra Ethernet extends Omni-Path into, from my view, a very viable and competitive solution in the market addressing both technology and business needs for new AI cluster deployments.

Their innovative approach to fabric delivery, and frankly their mere existence as a competitor to NVIDIA, will be a game changer for companies looking to optimize their near-term AI operations while maintaining the flexibility to grow in a constantly evolving tech landscape. Cornelis is indeed in a perfect position to disrupt a market, and with Lisa’s extensive knowledge of the data center landscape, key players, and business savvy, I expect a bright future for Omni-Path.

In my recent chat with Sascha Buehrle from Uptime Industries, we dove deep into the exciting world of AI at the edge, powered by their innovative platform Lemony. What stood out to me is how Lemony is designed to bring scalable and secure generative AI right to your desk. Instead of relying on massive, cloud-based infrastructure, Lemony offers an on-premise solution that is not only easy to deploy, but also incredibly accessible for small and mid-sized businesses.

One of the key highlights from Sascha was Lemony’s model support. It comes preconfigured to support several generative AI models with an emphasis on Meta’s Llama, the predominant open source tool de jour, which means you can hit the ground running with everything from natural language processing to image generation tasks and take advantage of a variety of tools to fit the job. It's a seamless integration for businesses that want to start using AI but don't have the resources to maintain large AI clusters or hire a specialized data science team.

Storage is another critical aspect when considering AI training at the deskside, and Lemony comes with options for up to 10TB of storage, making it perfect for businesses that need to manage substantial amounts of data securely. Sascha mentioned that Lemony’s edge AI solution can handle complex datasets while maintaining data sovereignty — a huge selling point for companies that prioritize privacy and data control. And since it's all on-premise, there’s no need to worry about the potential risks of sending sensitive information to the cloud.

One thing I’d called out in my pre-AIHW Summit show blog was a question on customer traction, and Sascha delivered insight into the early success the team has had with Lemony. It’s impressive to see the traction the company is gaining with customers. In just under a year, Uptime Industries has secured over 100 early adopters across a variety of industries — from finance to healthcare — who are using Lemony to streamline their operations with AI. This is a massive vote of confidence in the platform’s ability to scale and deliver real business value. Businesses are particularly drawn to the security features and the fact that Lemony is an all-in-one box solution, which means you can deploy it right in your office and start seeing the benefits immediately.

What’s the TechArena take? I don’t think Lemony represents the alpha and omega of the future of generative AI, but it does bring a fantastic product to the market that could tap a sweet spot of early adoption. Lemony is about bringing AI to where it’s most needed today: closer to your data, closer to your operations, and in a way that fits many companies’ specific usage model. I can’t wait to see where Uptime Industries goes from here, but with the momentum they’re building, I have no doubt they’ll play a key role in making AI practical and accessible for the rest of us. Watch this space as Lemony continues to evolve, and don’t miss out on the opportunities AI at the edge can bring to your business.

During our latest Data Insights podcast, sponsored by Solidigm, Ian McClarty of PhoenixNAP shares how AI is shaping data centers, discusses the rise of Bare Metal Cloud solutions, and more.

Sean Lie of Cerebras Systems shares insights on cutting-edge AI hardware, including their game-changing wafer-scale chips, Llama model performance, and innovations in inference and efficiency.

Letizia Giuliano of Alphawave Semi discusses advancements in AI connectivity, chiplet designs, and the path toward open standards at the AI Hardware Summit with host Allyson Klein.

Lisa Spelman, CEO of Cornelis Networks, discusses the future of AI scale-out, Omni-Path architecture, and how their innovative solutions drive performance, scalability, and interoperability in data centers.

Kelley Osburn gets storage. As an industry veteran and leader at Graid Technology, Kelley recently shared his insights on how the storage arena is rapidly transforming to fuel AI workloads and how his company’s SupremeRAID™ solution – a revolutionary approach to tackling modern data storage challenges – is hitting a sweet spot in the market.

So why is traditional RAID no longer sufficient? Kelley explained how these configurations struggle with high-performance computing demands, especially in data-intensive environments. He emphasized the need for innovation in data storage as the exponential growth in data continues to challenge existing systems, explaining that RAID's original purpose was to provide redundancy and protection against disk failures. While this redundancy is still valued, it lacks the performance desired by many customers.

As data stores grow and speed-of-delivery of data becomes more urgent, innovation to the approach helps extend RAID solution viability while meeting customer demand. Graid's SupremeRAID™ solution, for example, optimizes storage performance by offloading RAID tasks to a dedicated hardware device, enhancing speed and efficiency without compromising data integrity. This makes it an ideal solution for customers managing massive amounts of data for applications like AI, machine learning, and big data analytics.

Kelley detailed the core value of Graid’s solution, describing how SupremeRAID™ addresses critical bottlenecks in traditional storage systems by offering unprecedented performance gains while reducing the computational load on CPU and system resources. The innovation lies in its architecture, which integrates both hardware and software in a way that eliminates RAID-specific processing burdens from the host server, thus allowing server resources to focus on other tasks. The result is a solution that dramatically improves throughput and reduces latency, creating a more balanced and efficient data environment.

In addition to AI and ML, SupremeRAID™ also proves to be a valuable tool in applications in media and entertainment, where high-resolution content creation, editing, and rendering demand significant data processing power. Its ability to handle these workloads without compromising performance makes it a game-changer for companies managing large data sets.

Industry Implications and Future Outlook 

So what are the broader implications of Graid's innovations for the storage industry? Kelley explained that as companies continue to generate vast quantities of data, the demand for more efficient and performant storage solutions will only grow. Graid’s SupremeRAID™ is positioned to address these challenges head-on, providing enterprises with the tools they need to manage, protect, and access their data faster and more reliably than ever before. This will rely on underlying storage media delivering the performance and density required for these tasks, and Kelley pointed to Graid’s strategic collaboration with Solidigm as an example of how high performance QLC memory delivers unique value to customers.

Looking to the future, Graid plans to continue evolving its technology to meet the ever-increasing demands of the data economy. As data volumes grow, so too will the need for innovative storage solutions that can handle not just the size, but the speed and complexity of modern workloads. SupremeRAID™ represents a critical step in that direction, offering a glimpse into the future of RAID technology and its role in addressing the data challenges of tomorrow.

Want to learn more? Check the episode here.

‍

Join Sascha Buehrle of Uptime Industries as he reveals how Lemony AI offers scalable, secure, on-premise solutions, speeding adoption of genAI.

In the latest episode of The Tech Arena, I had a fantastic conversation with Neeraj Kumar, the Chief Data Scientist at Pacific Northwest National Laboratory (PNNL), about how AI is shaping the future, especially around energy efficiency and large-scale data processing.

Neeraj is a dynamic thinker who’s deeply invested in how AI can revolutionize everything from scientific discovery to sustainable technology. Together, we explored some of the key innovations and challenges in the AI space, with a special focus on energy consumption and the balance between advancing technology and environmental responsibility.

Neeraj kicked off by discussing one of the most pressing topics in AI today — how to manage the skyrocketing demand for computing power without exacerbating the climate crisis. AI is, after all, a double-edged sword. While it brings incredible breakthroughs in fields like medicine, climate modeling, and autonomous systems, it also consumes a huge amount of energy. Neeraj highlighted this tension between innovation and sustainability and stressed that AI's future hinges on solving this dilemma.

He shared insights into his work at PNNL, where AI is driving cutting-edge research. One area he focused on was the collaboration between PNNL and Micron to improve memory systems for AI and high-performance computing (HPC). As AI systems evolve, they require ever-greater amounts of data to be processed in real-time. That’s where memory plays a critical role. Neeraj pointed out that memory bottlenecks can cripple even the most advanced AI models, and optimizing these systems is crucial for enabling more energy-efficient AI. He believes that advancements in memory technology will be a game-changer in achieving the balance between computational power and energy efficiency.

I loved how Neeraj dove into the technical aspects of how PNNL and Micron are innovating in this space. They're working to enhance memory architectures to meet the needs of next-generation AI systems. The aim is to reduce the energy footprint of AI while improving its capacity to handle massive datasets. This kind of optimization is vital for industries like healthcare, where AI can sift through oceans of data to discover new treatment methods or predict patient outcomes, but only if the computational infrastructure can keep up.

We also discussed how intertwined the worlds of AI and compute efficiency have become. It’s no longer enough to simply innovate; those innovations must also be energy-conscious. Neeraj made a great point about how the energy demands of AI systems are growing exponentially, and we need to rethink how we design hardware and software. He highlighted how PNNL’s research is pushing the boundaries on what's possible, from creating AI models that are more efficient in their data use to developing entirely new computing architectures that reduce energy consumption without compromising performance.

Neeraj shared his vision that AI itself can be a powerful tool for combating climate change on a broader scale. For instance, AI can help optimize energy grids, improve the efficiency of wind and solar farms, and even model climate patterns more accurately. But to truly leverage AI in the fight against climate change, we must first ensure that the technology is sustainable. Neeraj was candid about the fact that we’re not there yet, but with ongoing research and innovation, we’re heading in the right direction.

Climate change is not the only area where AI is contributing to advancement in the scientific community. Neeraj shared some compelling examples of how AI is transforming research at PNNL. One standout example was how AI is being used to accelerate scientific discovery by identifying patterns in data that humans might miss. This is particularly important in fields like materials science and chemistry, where AI can analyze vast datasets from experiments and simulations to uncover new materials with potential applications in everything from energy storage to quantum computing.

The role of AI in scientific discovery is something Neeraj is particularly passionate about, and it was evident in the way he spoke about its potential. He emphasized that AI doesn’t replace scientists; rather, it amplifies their ability to make new discoveries. In many ways, AI acts as a partner to researchers, sifting through enormous amounts of data and generating insights that can lead to breakthroughs much faster than traditional methods. This is where AI’s promise really lies — in its ability to accelerate progress in some of the most complex and pressing challenges of our time. 

Towards the end of our conversation, we touched on the future of AI and what it might look like in the next decade. Neeraj is optimistic about where the field is headed but also realistic about the challenges that lie ahead. The next big leap, according to him, will come from integrating AI with other emerging technologies like quantum computing. This, he believes, will open up new frontiers in both AI capabilities and energy efficiency.

We cover a lot on the TechArena about enterprise applications of AI, and the podcast with Neeraj was a welcome reminder of how AI is contributing positively to society at large, advancing science in ways we wouldn’t otherwise accomplish. While there are significant challenges ahead to tap the full potential of AI to these use cases, especially when it comes to energy consumption, with leaders like PNNL and other US labs driving innovation, I’m confident that we’re on the path to a more sustainable, AI-driven future.

“I.N.T.E.L.L.I.G.E.N.C.E. is down! I repeat, we have no I.N.T.E.L.L.I.G.E.N.C.E.!”

-  Lisa, Team America: World Police*

Numbers and dates sometimes seem to take on an oversized significance. One could posit that the random patterns AI sometimes finds in a stream are the equivalent of the random patterns burned into the human noggins by repetition, creating reputation. This could lead to a topic for another time: does human bias lead to implicit programmatic bias by accident? Only the ones and zeroes know, and those got buried in the compiler eons ago in computer years. 

Anyway, some date combinations evoke visceral reactions among the subsets of populaces. Drop a 12/25 into a Western Culture discussion, and the twinkly lights and a sudden introduction of forestry into the household takes mental hold. In China, mention of 8/8 sends the wedding planner industry into a tizzy. Some dates are less about happiness. In former days, a mention of 12/7 (though they called it December 7th back then) brought out the somber reactions of those who remember when our nation was lured into war by the attack on Pearl Harbor. Though fading, that pattern still evokes a response.

These days, the date of today’s publication seems to resonate in the minds of many, 9/11. It’s fresh in the cultural mind, and the global implications of the day a bit more than two decades ago keep washing ashore like some geopolitical flotsam. Among we, the tech weenies, a lot of the response drifted to the feeling that technology needed to create change to protect the innocent. We’re only accidental technocrats, after all. Two subjects drift to mind, proving that patterns evoke patterns.

One that we can address later in detail was the realization that redundancy and continuity involved more than putting the backup tapes in the basement of the same building. The needed push of number-crunching close to the trading floor (someone wrote about distance and latency recently, can’t remember where…) was fighting the pull of real-estate values and commute times. One firm’s bright techie idea of building some redundancy across the Hudson in New Jersey led to copycats that did the same. That migration alone likely shortened the closure of the largest markets to six calendar days from the months it would have otherwise taken.

On another vector, though, the fairly new power the computer industry was feeling resonated off the political firewalls: “We can use technology to make us all safer against global threats.” The bright engineering minds educated in the 30-year trough of relative international peace saw the opportunity for sea change. Even as we said it, the creep of horror at what technology as a tool might unleash sowed the seeds of its own inertia.

Yer Humble Author (YHA) remembers not long after the infamous date visiting one of the largest database vendors in the world with a shiny new concept server. Said system would throw all of its 32 threads across eight processors at a then-incredible 8TB of DRAM. It was 20 years ago, you stand on the shoulders of ants, friends. Rhetorically, YHA asked the team of think-biggers what Homeland Security could do with an 8TB in-memory database. The conversation that followed was interesting, and a little scary, though the implications seemed to point to the positive…if you squinted.

Now that a quick trip to the open-source Gits and the virtual card swipe at Amazon provides everyone similar capability, how has technology made the globe safer against the shadowy baddies? Well, many of the apparent changes in safety and security seem to be fractional, in the diddly over squat range. Technology purchasing in the hopes of a more secure globe certainly absorbs the tax dollars of many, but our tools don’t appear to have created that vital change that moves us into the realm of 60’s Star Trek unity. Some of us want the transporter, airports really suck post-9/11.

And therein lies the conundrum. Technology is a tool. The redundant tools that saved the backsides of the finance industry aren’t all that different from those analytics tools that get tossed at global metadata for security. And yet, the conflicting thought processes – political of all stripes, apolitical liberty, the need for consensus of committee, trying to explain all this to leadership who stopped at algebra – grind some tools to a nub of inefficiency. The tool is not to blame, it’s the users. Accidental technocrats, untie (sic). 

Hope should never be lost where technology is available. The leaps of innovation repeatedly outperform the bloat of human decision making. And maybe the human bias projecting into machine bias is a re-reflection of the same human bias. We’d probably just feel better on some digitally-significant days if we didn’t always stand in our own way.

*(Seriously, that might have been the only quote from the movie that could be used in a family-friendly space. But now that you’re at the end, you know that’s the only movie we could have used.)

Mark Wade, CEO of Ayar Labs, explains how optical I/O technology is enhancing AI infrastructure, improving data movement, reducing bottlenecks, and driving efficiency in large-scale AI systems.

Gayathri Radhakrishnan knows how to spot differentiated value and understands well the difference between innovative technology and well-designed PowerPoint.

As a partner at Hitachi Ventures with a deep pedigree in venture capital, she’s a savvy expert in valuation of what comes next. This is why I was so excited to catch up with her at this week’s AIHW & Edge AI Summit in the valley. The AIHW conference is many things – a landscape of silicon innovation, a trend source for AI model development, and also a hub for venture deals in-the-making as startups showcase their new tech and silicon valley investors evaluate respective offerings. I wanted to check in with Gayathri on the state of the AI market, what areas of innovation will be garnering venture investment in the future, and how she sees the landscape solidifying as we enter the second half of the decade.

Hitachi Ventures has made a few moves of late in the AI space, most recently announcing investment in Archetype AI, Ema, Strikeready and Trustwise. Their portfolio shows an acute focus on mindful investment in use cases that will drive real enterprise value, fitting with Hitachi’s overarching stable approach to the tech arena. Gayathri shared that evaluation for real innovation is nothing new, harkening back to the cloud era rife with startups that were delivering virtualization solutions and branding them cloud. AI is no different. There may be fantastic analytics solutions under development, but ferreting out what is truly using machine learning, what is truly automated, is the key. That metric, Gayathri explains, is based on what layer of the stack is in play and what unique value is required for that layer of innovation to disrupt in the market.

Gayathri and her team are also surveying the infrastructure landscape for innovation in efficient delivery of computing to fuel AI. With US data centers forecast to consume 9% of the world’s energy by 2030 alone, it’s no surprise that this focus on efficiency is a priority. She underscored the push and pull in the market between customer demand for more efficient IT and industry engineering investment to deliver. This starts at the boardroom and C-level and trickles down to IT vendor conversations. influencing both disruptive tech evaluation as well as potential M&A moves from large infrastructure and cloud players with startup infrastructure solutions.s

All of this innovation and investment is ultimately focused on delivering value to the enterprise. Gayathri stressed Hitachi Ventures’ broader purview across industrial markets and the need for enterprises to adopt this new powerful technology responsibly, with focus on safety and reliability required for the solutions enabled by its use in the marketplace. While she sees continued advancement of large AI models and breakthrough technologies like AI agents, smaller model integration into reliable use cases will be a feature for 2025.

If you’re at AIHW Summit this week, connect with Gayathri to learn more about her perspectives and be sure to check out her panel, Industry Landscape Overview: Funding Trends & Emerging Business Models in Generative AI Start-Ups, & Enterprise Adoption Patterns, featuring insights from Applied Ventures, Silicon Catalyst Ventures, and NGP Capital.

It’s Oracle CloudWorld week, and I was excited to hear from industry legend, Larry Ellison, on his view of the future of computing and the AI era. Oracle holds a distinct view of the enterprise and adoption of technology, so there are likely very few people who can project enterprise adoption curves better than Larry.

The keynote did not disappoint. Larry emphasized Oracle’s vision of AI for the enterprise, where practical AI solutions are designed to address real business needs. He highlighted the integration of generative AI across Oracle’s cloud applications, with tools supporting functions like finance, HR, marketing, and customer service.

This was lack of splash of the Open.AI crowd and more meat and potatoes for grizzled IT veterans. Larry emphasized that Oracle’s AI focus is on transforming business processes, not just hype-driven tools, and delivering AI-driven automation at scale.

Larry also spotlighted Oracle Cloud Infrastructure (OCI) as a trusted platform, built to handle sovereign cloud needs and ensure secure, region-specific data processing. This commitment to AI readiness has positioned Oracle as a trusted provider for industries like healthcare, finance, and government. Oracle’s approach underscores its reliability and adaptability in helping enterprises deploy AI efficiently, combining performance with the security enterprises demand.

One of the most compelling aspects of Larry’s speech was the focus on vector search, a method Oracle uses to process large volumes of unstructured data, seamlessly embedding AI into database environments like Oracle Database 23ai. This ensures that enterprises don’t need to move their data to leverage AI — AI can operate directly within the database environment, optimizing workflows for speed and scalability.

Oracle’s generative AI technology is geared to simplify complex tasks, automate workflows, and enhance decision-making. For example, Oracle Fusion Applications integrate AI for tasks like project proposal generation, strategy development, and content summarization, tailored to support business functions directly.

One thing I thought was interesting was Oracle’s ambition to follow a similar path with AI as it did with the cloud — building a trusted ecosystem that delivers real value for enterprise applications. Larry noted that while Oracle’s focus isn't on flashy AI tools for consumers, its enterprise-first strategy offers the reliability and scale required for mission-critical applications, very similar to its path with delivery of cloud services that met specific enterprise requirements.

What’s TechArena’s take? I think of a lot of companies before Oracle in terms of pure AI innovation. Those born on the cloud who lean forward into the tip of innovation will likely deliver the most impactful transformative tech in the near term. But…for those meat-and-potatoes workload enhancements with AI automation and agent control, I think Oracle will make an outsized impact on adoption. Time will tell if Oracle can once again re-imagine itself for this latest era of computing and prove to continue its centrality on enterprise application and service delivery.

Neeraj Kumar, Chief Data Scientist at PNNL, discusses AI's role in scientific discovery, energy-efficient computing, and collaboration with Micron to advance memory systems for AI and high-performance computing.

I believe that AI at the edge will start dominating AI chatter in 2025 as enterprises large and small seek to integrate generative AI tools into operational applications.

This is what we think of at the TechArena as the 2nd wave of gen AI, the first being the training of behemoth models by brute force of massive AI clusters in the largest clouds on the planet. While the hyperscalers continue on their quest for AGI supremacy, the rest of us are looking for real world integration into business operations to ramp the efficiency and insight at our fingertips with this powerful technology. Of course, whenever we discuss enterprise adoption, we need to consider words like reliability, trust, security and more. Hallucinations and other risks of too careless of AI adoption can represent brand reputation loss or worse.

The team at Uptime Industries has designed a solution to enable smaller businesses and work teams to harness the power of generative AI for unique workloads with that trust and reliability in mind.

Meet Lemony, a secure, on-premise generative AI platform that delivers an all-in-a-box solution for on-prem deployment at the edge. A single Lemony box will run you $499 for up to five users, and boxes can scale to address larger team requirements. What’s best, your data, combined with generative AI model power, can work in tandem in a transparent and secure fashion, to capture the application advancements you need to fuel your business.

After reading about Lemony, I’m intrigued to learn more! I love innovative solutions that take the heady and complicated of new tech innovation into a consumable tool that even companies the size of TechArena can use. I love the focus on data security and on- prem deployment for organizations that don’t want to compromise on data location or control. I love the speed that this company has been moving as well, going from formation in March of 2023 to seed round in January of 2024 complete with a core team of experts from across the companies you’d expect to help make them a success.

Frankly, I have a lot of questions for the Uptime Industries team, which I’ll get answered at next week’s AIHW and Edge Summit. The Uptime Industries team will be showcasing their wares at the event. While most of the companies assembled will be discussing the latest innovations to pave the path to AGI, they will be discussing how those who aren’t running the world’s largest data centers can tap AI’s power to transform our businesses. Watch this space for more next week.

‍

Arun Nandi, VP of Global Data and Analytics at Unilever, knows his way around data and how to apply it for positive business outcomes.

In my recent conversation with Arun in advance of this week’s AIHW and Edge AI Summit, Arun shared how his company is leveraging AI and data analytics to enhance business operations, drive sustainability, foster innovation and more. 

Arun’s insights centered on the transformative potential of AI in improving corporate decision-making, optimizing supply chains, and refining product offerings. What’s better, Arun believes all of this can be delivered while maintaining a focus on environmental responsibility.

AI and Analytics Sit at the Core of Innovation

Unilever is one of the world's largest consumer goods companies. They handle vast amounts of data daily from across their business portfolio, ranging from information on supply chain logistics to culled insights on consumer preferences. Arun sees AI and advanced analytics as crucial in handling this data to extract meaningful insights that inform decisions at all levels of the organization, and that this extraction is critical for competitive advantage in today’s market.

From product development to marketing strategies, AI is being used to refine and predict consumer needs, ensuring that Unilever can keep ahead in an ever-changing market.

Arun shared some details about customer behavior modeling taken on by businesses like his. By analyzing consumer behavior data, organizations can personalize marketing and product recommendations, providing a more tailored and relevant experience for customers. This not only improves customer satisfaction, but also increases loyalty and retention, and this is widely deployed today utilizing existing analytics and AI capabilities. With new generative AI models, the ability to forecast consumer behavior will only grow more accurate and impactful.

Sustainability Through Data

Our interview also covered the challenge with efficient IT to fuel AI, and Arun highlighted that this is a core value at Unilever. In fact, AI is playing a pivotal role in helping the company achieve its environmental goals. Arun shared that Unilever is committed to reducing its carbon footprint and creating more eco-friendly products. Through data analytics, the company can identify areas where it can reduce waste, optimize resource use, and create sustainable products without compromising on quality or performance.

For example, AI helps Unilever track the environmental impact of its supply chain, ensuring that the company can source raw materials more responsibly and efficiently. AI also allows the company to forecast demand more accurately, reducing the chances of overproduction and excess waste. These efforts contribute to Unilever’s broader mission of achieving net-zero emissions and promoting a circular economy. And, they are part of a larger trend in corporate use of the powerful technology to enhance energy use and implement sustainable business practices.

The Future of Enterprise Data Architecture 

Arun also provided his vision for the future of enterprise data architecture as one where AI and machine learning will become even more integral where continued investing in AI advancement is required to stay ahead of the curve. This includes building robust data infrastructures that can handle the increasing complexity and volume of data Unilever manages while ensuring privacy and security for consumers.

One of the challenges that Arun touched upon is the need for a cultural shift within organizations to embrace AI and data-driven decision-making fully. He stressed the importance of upskilling employees and creating a data-driven mindset across all departments to ensure the successful integration of AI technologies. This requires new avenues of collaboration, not just across Unilever business groups, but across the industry. 

The solution for this collaboration is new partnerships between businesses, academia, and governments to drive innovation and tackle global challenges such as climate change and resource scarcity. By sharing data and working together, companies can amplify their positive impact on the world and achieve shared sustainability goals.

What’s the TechArena take? Our conversation with Arun was a fantastic reminder that AI has been implemented in IT organizations for years, automating critical functions in relation to advanced analytics. These powerful tools are opening new business opportunity and driving efficiency to corporate processes. While much of the chatter on AI is focused on advancement of large language models, the enterprise is happily deploying core AI-enabled applications across business functions with an eye to improve these functions with new model capability. As we look ahead to 2025 and expected deployments of gen AI in the enterprise, they likely will be built atop what’s already been done in many organizations.

Our second take? We see an important trendline on AI’s positive impact to overall corporate sustainability efforts and mindful use of energy and resources. Expect more stories from corporations in the months ahead on this theme as companies look to counter set the energy consumption utilized for deploying these models from data center to edge.

If you’re at AIHW Summit this week, be sure to check out Emerging Architectures for Applications Using LLMs, the Transition to LLM Agents, featuring Arun alongside experts from Stanford, Union.ai and Nava Ventures as well as his talk, Revolutionizing Language Models: Innovative Designs in Database Layers for Retrieval Augmented Generation, both within Tuesday’s lineup.