Dell Technologies goes all-in on AI factories

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Dell Technologies is going all-in on AI factories and enabling generative AI workloads powered by a wide range of third parties as well as Nvidia-flavored efforts.

What Dell is ultimately trying to do is drive enterprise adoption based on data needs, services, an open ecosystem, infrastructure innovation and use cases. With validated and optimized designs, Dell is looking to accelerate AI adoption by simplifying deployments and improving returns on investment. Dell has seen strong demand in AI-optimized servers and has forged strong partnerships with Nvidia, Hugging Face and a bevy of AI players. 

Speaking at the Dell Technologies World, Chairman Michael Dell said the company is reinventing itself again as it does with multiple technology pivots. 

Dell said the generative AI era will be a "sprint and a marathon." "The uptake and demand for AI is unprecedented," he said. "It will touch every industry in every organization. but each industry and each organization will have its own specific requirements and needs. 

He added:

"If you have engineers and a call center and you're not using an LLM based Assistant, you're already behind. But large language models don't create value on the factory floor. For those you need vision models on diverse models, and real world models. Other models will tackle other problems. We're still in the early stages of model training.  I'm blown away by the progress of open and closed large and small models by the exploding set of enterprise use cases. For inference and for assistance. We're seeing small models that are incredibly capable, that are faster and inferencing and more efficient and lower cost. Eventually the application of AI will be as broad as the internet."

At Nvidia GTC, Dell Technologies was highlighted as a key partner to build out AI factories. At Dell Technologies World, the companies delivered. In addition, Dell is also offering multiple building blocks to build out with packaged systems for use cases, product and consumption model. Sam Grocott, Senior Vice President, Product Marketing at Dell Technologies, said the company is building out AI factory offerings that cover the "broadest ecosystems across applications, infrastructure and partners" as well as the "Nvidia-flavored easy button."

When you boil down Dell Technologies World news, it boils down to the broad AI factory strategy and then the building blocks and packages of systems underneath.

The broad AI factory play will include a bevy of components including AMD and Intel AI accelerators and processors. Grocott said:

"We've got strong partnerships with Intel, AMD and Nvidia. Obviously, we're going to continue to harvest that flexibility and choice for our customers depending on which partnership or solution they want to lean into. We've got them all lined up across, clients, data center and all the way up to the cloud."

"We're going to be choosing best of breed partners to pull in and implement our IP along with our partner IP to simplify, package up and then make it consumable."

But given Nvidia's lead, Dell's AI factory stack built on Nvidia GPUs and software are fully baked. Nvidia CEO Jensen Huang said Dell Technologies AI factory effort will be the largest go-to-market partnership the GPU maker has. "We have to go modernize a trillion dollars of the world's data centers," said Huang.

• Nvidia acquires Run.ai for GPU workload orchestration
• Nvidia Huang lays out big picture: Blackwell GPU platform, NVLink Switch Chip, software, genAI, simulation, ecosystem
• Nvidia today all about bigger GPUs; tomorrow it's software, NIM, AI Enterprise

Dell PowerEdge XE9680L is designed for Nvidia HGX B200, supports 8 GPUs in a dense 4U form factor, has network throughput with 12 PCIe slots and liquid cooling. The system also supports 400G Ethernet as well as Nvidia's InfiniBand while supporting up to 72 GPUs per rack.

According to Dell, the Nvidia-powered AI factory will also make use of Nvidia AI Enterprise and frameworks outlined at Nvidia GTC. Dell will also wrap services around Nvidia deployments of software and infrastructure.

Dell NativeEdge automates delivery of Nvidia frameworks to edge devices.

And here's a look at the building blocks.

Infrastructure

• Dell AI PCs. Why are AI PCs included in the AI factory mix? Dell officials see edge devices as critical parts of AI workloads. Dell announced five new laptops featuring Qualcomm's Snapdragon X Series processors. The Latitude 7455 and 5455, Inspiron 14 Plus and 14 and XPS 13 will be available with Qualcomm chips, a dedicated AI key and the ability to run 13 billion parameters running on a PC.
• Dell PowerScale F910, an all-flash storage system that has 20x performance than PowerScale F900 and 6x performance vs Azure NetApp files. PowerScale F910 is the first ethernet storage for Nvidia DGX SuperPod and is available on-premises or in the cloud.
• Dell Z9864F-ON networking for ethernet fabrics for AI. The systems features Enterprise Sonic 4.4 with SmartFabric Manager and Broadcom's Thor2 network interface card. The system can minimize congestion and supports 8,000 GPU nodes.

Ecosystem

• Dell and Hugging Face have created an Enterprise Hub, an authenticated portal with optimized open-source models for on-premises deployments. The hub includes models optimized for Dell infrastructure, dedicated containers, scripts and technical documents and software dependencies.
• Dell PowerEdge XE9680, XE8640 and R760xa will be optimized for Meta's Llama 3 for on-premises deployments.
• Dell's AI infrastructure will integrate Microsoft Azure AI services with Apex Cloud Platform for Microsoft Azure.

Services

• Dell will launch a set of services to develop Microsoft Copilot in Windows, sales, security and GitHub. The services are designed to plan, design, implement and test and operate at scale.
• The company added Accelerator Services for Dell Enterprise Hub for generative AI prototyping and deploying with Hugging Face models.

Copilot, genAI agent implementations are about to get complicated

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Copilot and generative AI assistant implementations are getting so complicated that the consultants are marching in.

As tech vendors roll out various copilots and assistants across applications sprawl is going to become a real issue in a hurry. Enterprise software vendors all have AI helpers across their suites and platforms often with an per user surcharge.

Implementing these layers of copilots is going to require consultants as enterprises inevitably customize.

It all sounds a bit ERP to me.

Consider:

• IBM Consulting launched IBM Copilot Runway, which will help create, customize, deploy and manage copilots including Copilot for Microsoft 365. Consultants will focus on use cases and various processes.
• SAP and IBM will also collaborate on models for various use cases.
• Accenture has a similar deal to deploy Microsoft Azure OpenAI Service and hone use cases.
• Anthropic has partnered with AWS and Google Cloud with a dash of Accenture.

All of these consultants will come in handy. How do you decide between horizontal copilots and agents vs. use case and industry focused ones? The copilot game will be all about architecture, ModelOps and the ability to swap large language models as needed.

However, I raise an eyebrow when I see consulting firms swarm to the copilot pot of gold. That swarm of consultants mean that enterprises are struggling to execute the generative AI dream and implementation costs are going to go higher. The other axiom of enterprise technology: If there's a way to make projects overly complicated we will.

This post first appeared in the Constellation Insight newsletter, which features bespoke content weekly and is brought to you by Hitachi Vantara.

Here are a few reasons why copilot implementations are going to be a challenge:

Multiple models. Yes, you want model choice. Yes, you want to swap models as they mature and leapfrog each other. But enterprises in copilot implementations can pick a model today that's outdated a month later. Anthropic's Claude 3 was an enterprise fave. Then Meta's Llama 3 became an option. And OpenAI launched GPT-4o which may up the ante even more. LLM development is moving fast enough to give enterprises analysis paralysis.

Abstraction layers are developing, but haven't matured. Amazon Bedrock holds a lot of promise because enterprises can swap models. The platform is developing quickly, but the LLMs are moving faster.

Copilot sprawl. With every vendor, platform and service including a copilot, agent or assistant orchestration is going to be a big issue. Let's just take an average large enterprise stack.

• Multicloud: Google Cloud Gemini and Amazon Q
• ERP: SAP and Joule
• CRM: Salesforce with Einstein Copilot
• Productivity: Microsoft 365 with Copilot
• Productivity part 2: A few departments with Google Workspace and Gemini
• Platform: ServiceNow with Now Assist
• Custom applications: OpenAI ChatGPT, Anthropic Claude, Meta Llama 3, Databricks DBRX

Assuming you take on the AI assistants from just half of your vendors you are going to be crushed with sprawl and expenses.

Costs. The upcharge for various AI assistants was already starting to take a tool. Now you'll be adding consultants to the mix.

ROI. Generative AI returns to date have been judged based on productivity metrics. As these implementations see higher costs those returns will be shaved. Another wrinkle that may hamper returns: A CXO on our most recent BT150 call noted that he was struggling to make his premium copilot work across meetings and applications.

It's always about the data strategy. There's a reason that AI is being deployed successfully in regulated industries: These enterprises have to have their data in order for compliance. Other firms have to develop their data management and governance mojo to even consider genAI.

If all else fails you could just ask the latest LLM how you should manage the sprawl.

See more:

• Generative AI spending will move beyond the IT budget
• Enterprises Must Now Cultivate a Capable and Diverse AI Model Garden
• Secrets to a Successful AI Strategy
• Return on Transformation Investments (RTI)
• Financial services firms see genAI use cases leading to efficiency boom
• Foundation model debate: Choices, small vs. large, commoditization

CoreWeave raises $7.5 billion in debt financing for AI data center buildout

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

CoreWeave, a hyperscale cloud provider focused on AI workloads, raised $7.5 billion in debt financing to build out data centers primarily with Nvidia GPUs to go with $1.1 billion in equity funding earlier this month.

In the last 12 months, CoreWeave has raised more than $12 billion in equity and debt financing. CoreWeave's funding highlights how the generative AI boom has fueled a new category of specialized cloud vendors.

The debt financing, led by Blackstone, Magnetar and Coatue, will be used to develop compute infrastructure for existing contracts. CoreWeave customers are using its infrastructure to train models.

CoreWeave is also building out its European footprint, which includes a headquarters in London and a commitment to invest $1.25 billion in the region.

By the end of 2024, CoreWeave expects to have 28 data centers double its 2023 footprint, which was up from 3 data centers in 2022.

Michael Intrator, CoreWeave CEO and co-founder, said the company is looking to provide specialized GPU cloud infrastructure that can "reshape the cloud landscape, accelerate the AI race, and power the next generation of AI innovation." CoreWeave's infrastructure is designed for AI and machine learning, graphics and rendering and life sciences and other intensive workloads.

Nvidia is a CoreWeave investor, and the cloud provider is adopting its latest technology in its data centers whether it's GPUs, networking systems, applications and architecture. For Nvidia, CoreWeave is a key cog in its plans to scale AI factories. Microsoft is reportedly among CoreWeave’s largest customers.

According to CoreWeave, its infrastructure is purpose built for intensive workloads and can deliver performance 35x times faster and 80% less expensive than general purpose clouds. Here's a look at CoreWeave's Nvidia GPU pricing.

BT15O Zeitgeist: Vendor procurement, cool projects and margin compression

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

CxOs aren't thrilled with their enterprise technology vendors, becoming frustrated with copilots that don't play well together, and spending a good amount of time on procurement and consolidation.

Those were some of the takeaways from Constellation Research's BT150 CXO meetup

Projects from the BT150. We typically start off these calls with a simple question: What are you working on? Here are some of the projects that stood out.

• A virtual customer rep. One of our CxOs worked on a virtual space that combines the website with a physical store and a synthetic employee, which operates in a store and can show products and promotions as well as transact. The virtual employee can leverage the website content, generative AI and multiple language. The pilot is still in the learning stage but looks promising.
• Data centers in Africa. One person in the BT150 is working on data center buildouts in Africa and noted challenges ranging from rain forests, lack of roads and other infrastructure. The upside to these data center challenges is that the jobs involved can form a middle class.
• Mainframe migrations aren't easy. You work through a multi-year transformation and the final step is moving off of the mainframes. The hardest part of moving off mainframes are data migration, risk and cost. Mainframes are stable and can also be the most cost effective. The challenge isn't the compute, but the underlying data.

Security consolidation. Enterprises are looking to consolidate security vendors and platforms and not buying a tool unless it replaces something the company already has. Enterprises have realized more tools don't equate to better results. Vendors are either pitching platforms or forming alliances to work better with the rest of the security stack.

BT150 2024:

• BT150 CXO zeitgeist: Data lakehouses, large models vs. small, genAI hype vs. reality
• BT150 CXO zeitgeist: Low marks for SAP RISE, process automation, change management, AI risk
• BT150 CXO zeitgeist: AI trust, AI pilots to projects, VMware angst, projects ahead
• BT150 nominations | 2023-2024 class

Data breach resiliency and response moves to the forefront. CxOs are tasked with post breach plans and execution. The biggest challenge to responding to a cyberattack is that key parties (security intelligence), analysts and key stakeholders aren't talking.

Copilots aren't working well together even when they're on the same platform. Customers highlighted frustration with copilots breaking and not being well integrated with sibling applications. The problem? Vendors are racing to launch AI and scrimping on product development and integration.

Vendor management, procurement and margin compression. There were multiple takeaways about vendors and some of the comments from our CxOs were less than flattering. Key takeaways include:

• Oracle has a second life as a cloud provider and relevant IT vendor due to its database as well as its Oracle Cloud Infrastructure. The tech stack Oracle has created is strong. What's the problem? Oracle is a pain to deal with. Oracle is great at getting you set up and terrible when you're locked in. Oracle has done a few great things--cutting data egress fees, the Azure partnership and cloud efficiency--but is saddled with a bad reputation due to audits, compliance and sales practices. Oracle's reputation isn't about tech as much as it is the sales approach.
• ServiceNow's platform reduces costs even as customers consolidate on the company. ServiceNow has become a key abstraction layer for enterprise IT. ServiceNow gets the margin compression game and appears to be on the right side to make it work.
• Categorize your vendors based on whether they are trying to preserve margins or enable you to save by compressing them. Salesforce, Microsoft and SAP are playing the margin preservation game, which will require a different negotiation strategy.  The bet is that the vendors that become a lever for margin compression will win out over those trying to preserve margins.
• The bench of enterprise disruptors is thin. The CxO dance with massive enterprise software and cloud providers will continue because there aren't enough new vendors to disrupt them. Disruptive challenges are either acquired or stuck due to a dismal IPO market.

Disruption will have to be home grown. Given the state of the enterprise vendor landscape, companies that want competitive advantage will have to build their own automation and best process practices. You can't wait for an enterprise vendor to innovate.

Palo Alto Networks buys IBM's QRadar security offerings in broad alliance

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Palo Alto Networks will acquire IBM's QRadar Software assets in a broad partnership that gives the security vendor IBM Consulting as a channel.

As security vendors race to consolidate platforms and customer wallet share, IBM and Palo Alto Networks moved to forge an alliance that on paper looks like a win-win.

Here are the moving parts:

• IBM will use Palo Alto Networks as its preferred security platform internally.
• Palo Alto Networks will use watsonx large language models (LLMs) in Cortex XSIAM.
• IBM will offer more security services and feature Palo Alto Networks' platform and train more than 1,000 of its security consultants. These consultants will focus on migrating enterprises to Palo Alto Networks.
• Palo Alto Networks will acquire QRadar's software as a service assets and migrate customers to Cortex XSIAM at no cost. QRadar on-premises customers remain with IBM for features, support and bug fixes, but can also migrate to Cortex XSIAM.
• IBM will receive payments from Palo Alto Networks for QRadar on-prem customers that migrate to Cortex XSIAM.
• Palo Alto Networks to become the lead partner for IBM's hybrid cloud and AI efforts.

With IBM as a partner, Palo Alto Networks is looking to fend off multiple players including CrowdStrike. IBM can focus on its core strengths and leverage security services and AI models. CrowdStrike recently expanded its partnership with AWS. 

Chirag Mehta on the intersection of cybersecurity, design thinking and AI

The companies also said they will form a Joint Security Operations Center. IBM Consulting will be the preferred managed security services provider for Palo Alto Networks customers. Palo Alto Networks will also be accessible on IBM Consulting's Advantage AI services platform.

Financial details of the partnership weren't disclosed.

Cisco Q3, Q4 outlook better as company preps Splunk integration

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Cisco's third quarter was better-than-expected and the company outlined its next steps in the Splunk integration. The acquisition of Splunk means Cisco's subscription revenue is 54% of the total.

The company reported third quarter revenue of $12.7 billion, down 13% from a year ago. Splunk's revenue contribution in a partial quarter was $413 million. Earnings for the third quarter were 46 cents a share (88 cents on a non-GAAP basis).

Wall Street was expecting Cisco to report third quarter non-GAAP earnings of 82 cents a share on revenue of $12.6 billion.

As for the outlook, Cisco projected fourth quarter revenue of $13.4 billion to $13.6 billion, above estimates of $13.54 billion. Non-GAAP earnings in the fourth quarter will be 84 cents a share to 86 cents a share. Wall Street expected fourth quarter non-GAAP earnings of 84 cents a share.

For fiscal 2024, Cisco is projecting revenue of $53.6 billion to $54.6 billion. Non-GAAP earnings will be $3.69 a share to $3.71 a share.

Cisco also said former Splunk CEO Gary Steele will become President of Go-to-Market effective immediately. Steele will lead the Splunk integration process. Jeff Sharritts, Cisco’s Chief Customer and Partner Officer will depart in mid-July.

Here's a breakdown of Cisco's third quarter by product line:

• Networking revenue was down 27% from a year ago, but Cisco did note that customers were installing inventory.
• Security revenue was up 36%.
• Collaboration was flat.
• And observability was up 27%.

Splunk integration

Speaking on a conference call, Cisco CEO Chuck Robbins outlined the Splunk integration. The first effort is focused on sales synergies. Robbins said:

"The closing of the Splunk acquisition in q3 will also enable us to begin driving revenue synergies in our security and observability markets. Upon closing the deal, we identified 5,000 existing Cisco customers who have the potential to become meaningful Splunk customers and our sales teams are already making those connections."

• Cisco closes Splunk purchase, previews integrations ahead
• Cisco to cut 5% of workforce, Q3 outlook weak
• Cisco acquires Splunk in $28 billion observability, AI and cybersecurity play

• Cisco is buying Splunk for $28B in cash. What does it mean for you as a customer?

Robbins added that Splunk will be added to Cisco's partner ecosystem where the acquired company had little presence.

The companies also plan to integrate products and made a first installment at RSA Conference.

"We're working on rapid integration, investing in both product integration and go to market resources, starting with aligning our Cisco and Splunk sales forces and accelerating challenge enablement processes for cross selling and upselling our combined solutions."

Cisco plans to converge its security portfolio with Splunk. For now, the two companies are rolling out new features and integrations. Cisco said Extended Detection & Response (XDR) is now integrated with Splunk Enterprise Security to provide one detection to remediation workflow. Cisco also said its unified AI Assistant for Security is available in Cisco XDR. The company also said its Panoptica application protection platform includes AI and machine learnings for real-time alerts. Cisco also said Hypershield, launched last month, will get tools to isolate attacks from unknown vulnerabilities within the runtime. See: Splunk’s Acquisition by Cisco Accelerates Convergence of Network, Security, and Observability, Fueled by AI | 11 Top Cybersecurity Trends for 2024 and Beyond

AI infrastructure a focus

Robbins added that Cisco designs for AI architecture are doing well in hyperscale cloud deployments. Enterprises are also interested.

He said:

"We continue to see momentum with three of the top four hyperscalers deploying our Ethernet AI fabric, leveraging Cisco validated designs for AI infrastructure in the past two quarters.

Additionally, for those leading-edge enterprise customers who seek to be the early adopters of AI, our partnership with Nvidia will offer easy to deploy cloud based and on prem networking solutions for AI inferencing."

Constellation Research's take

Constellation Research analyst Holger Mueller said Cisco has challenges ahead:

"Chuck Robbins really thinks he can keep the cost structure of Cisco in place, and grow back to a level that justifies the current setup. Revenues were down about $2 billion year over year, but operating expenses are up (incl. restructuring). Splunk's impact was minimal overall, with the executive change of Gary Steele becoming President Go-To-Market for Cisco overall. It is clear that Robbins wants to make up for the networking revenue shrinkage with growth in security and observability, helped of course by Splunk, but with all regions shrinking, this will be a tough 3 to 4 quarters for Cisco ahead – again. The good news is subscription revenue is now 54% of revenue. Can Cisco find a way to gain from cloud and AI growth? So far it has not, but maybe Gary Steele will unlock this challenge."

Will Meta's Workplace shutdown be a boon for Zoom's Workvivo?

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Meta is shuttering its Workplace enterprise platform with a June 1, 2026, sunset and the move could be a boon for Zoom's Workvivo, which was acquired a little more than a year ago.

Workvivo is the sole migration partner for Meta Workplace, which had as many as 7 million subscribers in 2021. Workplace emerged as an employee engagement suite during the COVID-19 pandemic and was an avenue to connect frontline workers. Connecting frontline workers is still a big priority for enterprises. 

The Workplace from Meta timeline goes like this.

• Until Aug. 31, 2025: Workplace functions as usual. Meta will discount Workplace by 50% from Sept. 1, 2024, to Aug. 31, 2025.
• Sept. 1, 2025, to May 31, 2026: Workplace will be only accessible to read and download existing data.
• June 1, 2026: Workplace will be terminated and deleted.

According to Meta, Workplace's Graph API will be able to export information from your instance to Workvivo. However, Meta noted: "we will be working with Zoom to provide additional tools and services to facilitate a better transition for customers from Workplace to Zoom’s Workvivo product. We will follow up with more details for Workplace admins in the coming days and weeks."

The bet here is that most enterprises have Zoom so the migration to Workvivo won't be totally foreign.

What's unclear is how much of a boon the Workspace shutdown will be for Workvivo. For starters, the number of active Workspace subscribers remains a mystery. In addition, there's plenty of time for Workspace customers to consider other options.

The Constellation Research Shortlist for Employee Digital Workspaces includes Workplace competitors such as Microsoft Teams, SAP SuccessFactors Work Zone, Unily and others. Microsoft Teams is likely to be the biggest player looking to court Workplace users.

When Zoom reports its first quarter earnings, Workvivo is likely to get more play. In the fourth quarter, Zoom CEO Eric Yuan said that Workvivo upsold a Fortune 10 company and Zoom customer. A Workvivo customer adopted the broader Zoom platform. "As you can see, adding new products both organically and inorganically creates a virtuous cycle, allowing us to sell more product into a larger base," said Yuan.

If Zoom is successful, the deal as the sole migration partner should give Workvivo some momentum. It remains to be seen how many Workplace customers convert. For now, Workvivo and Meta have to execute and give customers a seamless migration. 

John Goulding, CEO and founder of Workvivo, said: "It's our top priority to support customers through this transition, and our team is working to make this process as frictionless as possible." 

Workvivo and Meta will provide migration tools and free implementation services. 

Google I/O 2024: Multimodal Gemini, Project Astra, AI agents and 'teammates'

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Google I/O 2024 featured a bevy of generative AI advances that will be layered throughout Google's product portfolio--photos, music, Workspace, search, video and other areas--but the real takeaway was that the company laid out its vision for agents, models training models, and creating systems that work for you.

Sundar Pichai, Alphabet CEO, outlined the vision a day after OpenAI laid out GPT-4o.

"With multi-modality soon you'll be able to mix and match inputs and outputs. This is what we mean when we say it's an IO for a new generation. That's one of the opportunities we see with AI agents. I think of them as intelligent systems that show reasoning, planning and memory and are able to think multiple steps ahead. They work across software and systems, all to get something done on your behalf and most importantly, under your supervision.

We are still in the early days, and you will see glimpses of our approach throughout the day."

Search is obviously the most important venue for Google generative AI given that it funds the model buildout. Google rolled out multistep reasoning in search and features such as AI Overview. Google search will ultimately reason, do research for you and tap into real-time data. We assume that this work on your behalf will feature advertising, but the data loop may be more important. 

But search was just the obvious headliner--for developers and Wall Street. Here are the takeaways to note from Google I/O's barrage of news, which served as a sequel to Google Cloud Next.

Project Astra. Astra is a "universal AI agent" designed to be helpful in everyday life. Astra is one of the reasons Gemini is multimodal and has AI training AI models. Google Deepmind Chief Demis Hassabis said Astra makes is a creation that's based on models training models and data feedback loops.

"At any one time, we have many different models in training, and we use our very large and powerful ones to help teach and train our production ready models together with user feedback," said Hassabis.

The token arms race. Pichai outlined how 1.5 million developers are using Gemini models to create next-gen AI applications. Gemini 1.5 Pro will now have a 2 million token context window and aggressive pricing. Gemini 1.5 Pro will be available today in Workspace Labs and be capable of synthesizing meetings, documents and your Gmail quickly.

• Alphabet shows Q1 strength in Google Cloud, initiates dividend

Trillium, a new TPU that will be available on Google Cloud for machine learning and AI workloads. "Trillium delivers a 4.7x improvement in compute performance per chip. Over the previous generation. So most efficient, and performant TPU today will make Trillium available to our cloud customers in late 2024," said Pichai, who noted that Google Cloud will offer Nvidia Blackwell GPUs too. Pichai's message was that Google has the scale to keep building out infrastructure for the AI arms race whether it's in liquid cooling advances, custom processors and fiber cables around the world. 

Gemini 1.5 Flash is a model that's expected to be as powerful as Gemini 1.5 Pro, but be faster. Gemini 1.5 Flash is designed for low latency tasks as well as the small, but fast model trend. Gemini 1.5 Flash, available for public preview, is built for speed and real-time answers such as customer service responses even with an up to 1 million token window.

• Foundation model debate: Choices, small vs. large, commoditization
• Google Cloud Next: The role of genAI agents, enterprise use cases
• Google Cloud Next 2024: Google Cloud aims to be data, AI platform of choice
• Google Unveils Ambitious AI-Driven Security Strategy at Google Cloud Next'24
• Equifax bets on Google Cloud Vertex AI to speed up model, scores, data ingestion

Search will have planning capabilities. Google pitched a vision where search can answer complex questions and plan. When looking for ideas, Google will give you an AI-generated answer powered by Gemini. These answers and suggestions will likely be monetized. AI-organized search results will appear when you're looking for inspiration. Don't be surprised if these planning capabilities turn up in Google Cloud services. Would business and process planning be much of a leap?

Meet your AI teammate. Workspace's Gemini-powered side panel will be available this month. These features across Workspace apps--Gmail, Meet, Docs--were announced at Google Cloud Next. Google demonstrated AI Teammate, an agent that has a Workspace account and participates in tasks and projects. The demo featured "Chip" configured by the business. The timeline? Stay tuned. Google is also rolling out Gems, which is a personalized Gemini assistant.

Gemini Nano will be multi-modal on Android devices. This AI on device announcement front runs Apple WWDC, but the move highlights how smartphones and edge devices will run models with low latency and create new experiences. Android 15 will feature "Gemini at the core."

AWS names Garman CEO effective June 3

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Matt Garman will become the new CEO of Amazon Web Services as Adam Selipsky is stepping down. Garman takes over June 3.

The Verge first published the memo from Jassy to AWS. Sources confirmed the report is accurate with more to come from the company shortly.

Amazon published the memos from Jassy, Selipsky and Garman.

Garman currently is Senior Vice President of AWS Sales, Marketing and Global Services. Previously, he was Vice President of AWS Compute Services.

Selipsky was one of the first vice presidents at AWS in 2005, spent 11 years there and then left to become CEO of Tableau. When Andy Jassy took over as CEO of Amazon, Selipsky returned to run AWS.

In his staff email, Jassy said when Selipsky returned it was expected that he'd lead for a few years and develop the bench of executive talent.

AWS has seen customers optimize their cloud spending and manage through generative AI. With the launch of Amazon Q and the uptake of Amazon Bedrock, AWS is on a good trajectory and $100 billion revenue run rate.

AWS annual revenue run rate hits $100 billion as growth accelerates | Constellation ShortList™ Global IaaS for Next-Gen Applications | Constellation ShortList™ Artificial Intelligence and Machine Learning Cloud Platforms

Constellation Research CEO Ray Wang said:

"Adam played a key role in Andy Jassy’s transition to CEO of overall Amazon.  The current morale has not been the best, and customers’ expectations remain high as we enter an Age of AI. This sets Matt Garman up to be in a good place to take AWS to the next level."

More:

• Amazon Bedrock gets custom model import, evaluation tools, new Titan models
• Amazon CEO Jassy's shareholder letter talks AWS' approach to generative AI
• AWS' Matt Wood on model choice, orchestration, Q and evaluating LLMs
• AWS ups its investment in Anthropic as giants form spheres of LLM influence