With support for Windows 10 ending in just over a year, Microsoft is nudging users to update to Windows 11, as many still appear to be in no hurry to make the switch.
Windows 10 has more than 68% share among Windows users worldwide as of May 2024, according to Stat Counter. However, Microsoft has said it will no longer provide technical support or updates for Windows 10 after Oct. 14, 2025, a decade after its launch.
To encourage users to upgrade to Windows 11 — released back in October 2021 —the company is displaying a banner on users’ screens when they make updates to Windows 10 that urge them to take “a new journey with Windows.” Microsoft-focused news source Windows Latest posted a screenshot of the banner, which the article’s writer said appeared on a Windows device when he installed Microsoft’s May round of OS updates.
Migration complications
Migration to new versions of Windows among both enterprise and home users has often presented a challenge for Microsoft and users alike. Sometimes users or organizations skip an update because of a hardware compatibility issue on older PCs. If a new version Windows also requires hardware to be updated as well, it can get expensive, particularly for an enterprise.
Some versions of Windows have had a harder time catching on than others for other reasons, such not having new feature sets interesting enough to users to warrant an update, or just plain bad timing.
“Generally, people are resistant to change, and when there is no perceived additional benefit for migration, they are not always willing to migrate to the newer versions of software, noted Pareekh Jain, CEO of EIIRTrend & Pareekh Consulting.
Perfect storm creating update delay?
Windows 11’s lackluster adoption may be the result of a perfect storm of conditions that have made updating to it unattractive for many users. Though the update itself is free, for corporate customers especially, Ranjit Atwal, senior director analyst at Gartner said there are always hidden costs that “lie in assessing application compatibility, migration, and support.”
The system requirements of Windows 11 were more demanding than Windows 10, and PCs sold even in the last several years before its release had trouble meeting them. This may have given many users pause when it came time to adopt the new OS.
Moreover, the transition is being further delayed by current budget constraints within IT departments, and market changes as “PC vendors are attempting to lure businesses towards higher-priced AI PCs,” Atwal noted. This latter scenario is “further risking the Microsoft Win10 support deadline,” he said.
With that support deadline looming, however, both corporate and personal users alike may finally have an incentive to migrate to Windows 11, given that “regular updates, including security updates, are essential for the trouble-free operation of computers,” Jain observed.
Asana is developing an AI “teammate” feature that embeds generative AI (genAI) into the flow of work for users of its work management app. The teammates — created using “trigger and action” format in Asana’s existing workflow automation tools — can perform tasks such as carrying out research and interact with co-workers to help coordinate work.
The new feature accesses information on the Asana work graph — data on the work a team does, and the people doing the work — and relies on large language models [LLMs] from the likes of Anthropic and OpenAI.
“When people think about working with AI like working with the teammate, they get better value out of it, because they’re better able to give it instructions, give it feedback, and effectively partner with it,” said Paige Costello, head of AI at Asana.
There are a variety of ways to customize a teammate, said Costello. One example (from a customer testing the feature) involves the work request process for a design team. Here, a “request triage” assistant can check an incoming work request form to determine whether it contains sufficient detail about the scope of the request. If not, the AI assistant sends a message back offering suggestions for more information to include.
That process can save back-and-forth messages between coworkers trying to get the correct information, Paige said.
Another customer — a marketing firm — set up an AI teammate to conduct research around clients. In that case, the teammate can generate a first draft brief with information on clients’ previous brand campaigns and what they’re known for.
The Asana app already has workflow automation tools, but the addition of LLMs allows for more powerful automations. “What’s really exciting about embedding LLMs into work management is that there’s judgement, there’s nuance,” said Costello. “Instead of it being a black and white rule — ‘if these things are missing, ask this’ — it’s directional, like, ‘here’s what we typically look for; use your judgment.’”
That said, given the unreliability of LLMs, there’s potential for the teammates to introduce errors into the work management process.
Costello said it’s important to have a “human in the loop,” when the AI assistant takes action. “We want to make sure that humans are the ultimate deciders,” she said. “They’re the ones that are accountable, they’re the ones completing the work: AI is just helping. AI shouldn’t be moving work forward in a way that is not permissioned effectively by the team or the person driving that work.”
The AI teammate feature could potentially be accessed by a broad range of Asana users, said Chris Marsh, research director at 451 Research, part of S&P Global Market Intelligence.
“I believe the AI teammate is designed to appeal in a general sense, in that the aim is to have the teammate enhance general team workflows,” he said, noting it could improve withaccess to more data in Asana’s work graph over time.
“It’ll inevitably be a work in progress in terms of how helpful it is from the launch, but what it does represent is AI being applied to the next tranche of use cases in work management,” Marsh said.
He expects “market interest” in the capabilities of a tool such as Asana’s AI teammate.
A recent 451 Research survey indicated that 44% of employees see “low or no impact” of AI on their workforce tools. Only 24% said AI has had a moderate impact. But when survey participants were asked about the types of technologies that could enhance how they work, the popular response was an “an AI assistant that can help with automating basic tasks and more strategic kinds of work design,” said Marsh.
“So, few are seeing AI’s impact so far, but a lot want to see it and want to see it in the form of an AI assistant to do the kinds of things Asana’s AI teammate is focused on doing,” he said.
The AI teammate feature is currently in beta trial with select customers. Asana is evaluating different pricing models with more specifics available closer to a general availability launch.
The buoyant Apple-in-the-enterprise space is fascinating to watch as it grows and evolves, in part because not so many years ago it didn’t really exist. Now that it does, this part of the device management and services industry is remarkable in that vendors are working hard to diversify what they provide.
Take a look, for example at how Jamf is blending device management with tough security, Mosyle’s focus on MSPs, and the many other players who are becoming increasingly focused on specific verticals. I see this diversification of approaches as an incredibly positive sign that represents the health of this part of Apple-related industry.
The latest such move comes from Kandji, which has optimized the device management experience it provides with something it calls “Assignment Maps.”
What are Assignment Maps?
Assignment Maps automate the assignment of settings and apps to Apple devices. This basically means IT can implement configurations remotely and automatically. The idea is that it streamlines IT workflows.
How do Assignment Maps work?
To some extent, Assignment Maps reflect Kandji’s core ‘Blueprints” concept. Those consist of collections of Library Items that deploy profiles, settings, scripts, security controls, and apps to devices. With these, devices are configured at scale in response to the Blueprint to which they’ve been assigned. You could have a different Blueprint (and different apps and permissions) for the sales people than for the marketing department, for example.
Assignment Maps work via a concept of assignment nodes. A node is a collection of settings and configurations assigned to a group of devices. The idea is that you might have one node that applies to all company devices, and a sequence of other nodes that are applied to reflect particular roles, tasks, or locations. For IT teams, it means configurations can easily be grouped together and nested, ideally optimizing deployment workloads while enabling devices to be configured precisely in line with business needs.
There are other features, including a new tool to identify why specific items were installed or configured on a device, and systems to prevent conflicting settings from being set on the same device. You’ll find out more about the service here.
Eat to the beat
As we wait on Apple’s announcements at WWDC next week, Apple admins are beginning to consider what fresh challenges they’ll need to overcome to manage their fleets. For many, perhaps the biggest obstacle they may need to be overcome will be how Apple deploys generative AI (genAI) across its platforms.
While for many that will be of huge benefit in terms of getting things done, IT will need to ensure it has mitigation strategies in place to constrain sharing of confidential data — particularly if Apple takes a hybrid approach with both on-device and cloud support, from Apple and third party services.
Solutions that enable granular control — enabling the use of genAI at the edge and forbidding access to not-yet-trusted cloud providers — will inevitably enter currency, and the “one screen manages all” approach championed by most device management vendors will be part of the response to this.
There will be other changes within Apple’s platforms that concern the enterprise, of course, (constraining use of unverified third-party app stores, for example). And yet the beauty of the ecosystem surrounding Apple’s products is that when it comes to large enterprise deployments of new technologies at scale and at device level, a major industry now exists to help your business through.
Microsoft’s Copilot+ PC vision revolves around Recall — a photographic memory for your PC that allows you to search all the web pages, documents, and conversations you’ve seen in a single place.
Contemporary as that concept may seem, it’s the same vision Google once offered on Windows PCs — with Google Desktop 20 years ago. The first release of Google Chrome for Windows PCs, launched in 2008, included a similar, now long-forgotten feature.
What’s old is new again. Let’s take a look at some classic Windows desktop tools from Google — which Google hasn’t offered in over a decade, unfortunately — and how incredibly similar they are to Recall.
The similarities aren’t all about searching your PC, either. Amusingly enough, Google Desktop also delivered widgets — another blast-from-the-past feature that Microsoft is getting more excited about these days.
Want expert insights on the future of Windows and tips for making the most of your PC? Sign up for my free Windows Intelligence newsletter — I’ll send you three things to try every Friday along with free copies of Paul Thurrott’s Windows 11 and 10 Field Guides as a special welcome bonus.
Microsoft Recall connections: Google Chrome 1.0
Remember Google Chrome 1.0? While Chrome is the most popular web browser on Windows PCs today, it was a quirky tool with an unusual interface when it launched back in 2008. The initial version had one long-forgotten (and long-ago-removed) feature when it launched: it could search your entire web browsing history.
Google even advertised the feature in the official comic the company created to promote Google Chrome. As a Chrome software engineer in the comic explains: “You have full text search over your history. If you found a good site for digital cameras yesterday, you don’t have to bookmark that page. Just type ‘digital camera’ and quickly get back to it.”
Yes, just like Recall, 2008’s Chrome would keep the full text of every web page you visited. Then when you opened the browser’s History page, you could search for words that appeared on web pages, giving you a full-text search of all the pages you’d visited.
As with Recall, the pitch is obvious: If you could quickly find any web page you vaguely remember with a quick search, that could boost productivity — whether you were an office worker, a student, or anyone else.
The original version of Google Chrome remembered all the words on the web pages you visited, too.
Chris Hoffman, IDG
Microsoft’s Recall plans may sound concerning at first, but Recall isn’t the privacy threat it appears to be. Chrome’s initial history-tracking feature had some similar privacy implications, storing a lot more data than a browser normally would. Yes, you could tell Chrome not to save your history or clear the history — but you have that same control over Windows Recall.
Chrome’s full-text-history-search feature apparently didn’t get much use, and it was quietly removed at some point. In a Hacker News comment, a former Google engineer who worked on Chrome explained that most people never found the feature, it used up a lot of disk space, and the search ranking algorithm wasn’t particularly great.
Perhaps there’s a lesson here and many people won’t use Recall to dig through their browsing history, either.
Windows Recall feels like Google Desktop 2.0
But Recall reminds me most of Google Desktop, a once-popular desktop-search-and-widgets tool for Windows PCs that Google launched in 2004 and axed in 2011. (Google even offered Google Desktop for Macs and Linux PCs!)
Google Desktop was designed to bring the company’s search features to everything on your PC — and in your Google account. Here’s what Google Desktop could search:
The files on your computer, including words inside Word, Excel, PowerPoint, PDF, and text files;
Emails stored in desktop email programs like Outlook, Thunderbird, and Lotus Notes;
Emails from Gmail in your Google account;
Web pages you previously viewed in your web browser (Google Desktop would log your browsing history and make it searchable, too — just like Windows Recall and Google Chrome 1.0);
Chats from Google Talk and AOL Instant Messenger;
Media files, calendar events, tasks, notes, contacts, and more.
After installing it on your PC, it sat in the background, watching your web browsing, chats, and the files you downloaded and created, building a searchable index of them. It even supported indexing plug-ins, so developers could extend Google Desktop and index whatever else they wanted to.
Back in the 2004-2008 era, that was a lot — maybe everything you did on your computer. I remember a lot of people who loved Google Desktop and treated it as their PC’s “photographic memory” in that era.
Google Desktop could have provided more comprehensive search results for office or school work than the tools we have today. I frequently find myself searching piles of different apps I use for work — was that piece of information in my workplace chat app, email, notes, or a shared document? Today, there’s nothing quite like Google Desktop was back in its day.
Google Desktop offered a web-based interface that put “Desktop” alongside traditional “Web,” “News,” and “Images” searches.
Chris Hoffman, IDG
Google Desktop was very web-centric in its own way. When you searched with it, you saw a Google-style search page in your web browser. But, like Recall, it stored your data on your PC and performed the searches locally.
That being said, Google did include some optional extra features that would raise an eyebrow today. For example, it could “search across computers” if you signed in with your Google account, which involved uploading your search index to Google’s servers. Not even Microsoft’s Recall can do that — Microsoft is taking pains to say your Recall data will always stay on your PC.
Sure, Google Desktop didn’t have plain-language search powered by generative AI tools — and it didn’t take screenshots of your computer every five seconds to extract information from them. But it did everything it could back in those days.
Google sent Google Desktop to the Google Graveyard as part of a “spring clean,” explaining that there was a big shift toward cloud computing — and that modern operating systems offered better built-in search features.
That’s true, but it also meant that the dream of searching everything in one place — every web page you viewed, email you worked with it, chat message you received, and document you had — was gone. Search became scattered across various different applications and tools — until now, when Microsoft is trying to bring it back together with Recall.
Google Desktop’s widget-style “gadgets”
Interestingly enough, the parallels with the Microsoft of 2024 and the Google of 2008 don’t end there. Google Desktop also had a whole desktop widget platform, which Google called “gadgets.” You could see information like the weather, news, and photos right on your desktop without opening another window.
When I look at Windows 11 and see the big search plans and the eagerness to place windows everywhere, I can’t stop thinking about Google Desktop and how far ahead of its time it was.
A forgotten era of Google-powered Windows software
The Google of that era feels like a different company than the Google of today. Google was a company with a mission of organizing all the world’s information, so why wouldn’t it offer superpowered desktop search tools that helped you dig through all the web pages you ever viewed, chats you had, emails you received, and Office documents you worked with?
Google really delivered a lot of excellent Windows software back then. Case in point: the Google Toolbar released in 2000. When you installed that free toolbar in Internet Explorer, it added a pop-up blocker to your web browser. (Yes, there was a time when Internet Explorer didn’t even have a pop-up blocker! What a mess that was.)
But Google Desktop is long gone, and Google is more focused on web-based tools and mobile apps than anything else. After so long without a big unified desktop search tool, I’m excited to put Recall through its paces. I’m hopeful it will speed up my work — just as Google Desktop did back in its day.
If you enjoyed this article, I know you’ll enjoy my free newsletter! Sign up for Windows Intelligencetoday — you’ll get three things to try in your inbox each Friday and free Windows 11 and Windows 10 Field Guides, too.
The battle royale for the future of workplace culture continues as companies and employees square off over the right to work at home and the return to office (RTO) movement.
Companies that think they’ve won the battle by issuing strict return-to-office (RTO) mandates could lose the larger war. RTO mandates are causing senior-level employees, women, and millennials among others to push the eject button and find other jobs with companies that have more liberal remote-work polices. If, by some miracle, the Federal Trade Commission’s move to strike down most existing and future non-compete agreements takes effect this year, an increasing number of employees looking for new work could wind up at competitors.
Finding AI talent, which more and more companies are going to need, is already a search for the proverbial needle in a haystack. Adding strict RTO policies to that scenario won’t make the talent search easier.
And while RTO policy pronouncements have slowed in recent months, there have been numerous reports of affected employees searching for new jobs, refusing to comply, and even taking legal action.
A new study by ResumeBuilder.com found that eight out of 10 companies have lost talent because of their RTO mandates. And yet, the same study found that 70% of companies plan to increase or maintain the number of days employees are required to be in the office in 2025.
Who will blink first?
Keeping track of employees
It hasn’t helped that some companies, including Amazon, Dell, Google, JPMorgan, Meta and TikTok, have reportedly begun tracking employees’ badge swipes. Dell is also using VPN (virtual private network) monitoring, and has created a badge color-coding system to categorize employee onsite presence. According to Business Insider, Dell also told workers who work remotely full time that they won’t be eligible for promotions.
It’s difficult to comprehend why some top execs at companies with RTO mandates seem more focused on making their employees suffer with commutes, added costs, lost productivity, and in some cases invasive tracking than they are with making a profit. At some organizations, there’s almost a vindictive quality to the RTO policies and the tone of communications. What possesses CEOs and other leaders to create such a toxic environment at their companies? What is this obsession with making employees work in the office when research has shown that employees are more productive when working where they choose?
The cost of doing business in the office
One motivation behind employees’ desire for the freedom to work at home involves the financial costs of going into the office. RTO means some sort of commute for employees, which means time and money, especially if you’re a two-income family and two adults have to commute by car. Gas, maintenance, insurance, parking, public transportation, and toll roads are just some of the possible added commute expenses.
Working at home could lead to potentially less expensive, less stressful, and more flexible childcare options. (Many parents with older children are able to drop before- and after-school daycare programs when they work remotely, for example.)
With the likelihood of buying more lunches in the office, and less time to cook dinner after a long commute home, food costs go up, too.
All of these considerations add up to a significant drag on work-life balance. The desire to improve that balance is the single most important reason why remote work is so highly prized by employees. The COVID-19 pandemic gave many workers a good long taste of it — and they don’t want to let it go. It makes employees more productive on the job and gives them more control over their personal lives.
This is playing out now in companies around the globe. And with neither side willing to give in, the fight will continue to bubble up this year and perhaps even next, leaving employees frustrated and on the hunt for new jobs — and companies in danger of undermining their best assets: their workers.
Goodness, this is an unseasonably chilly June morning. Maybe that’s because the metaphorical sensation of hell freezing over is quietly lingering in this damp virtual air we share.
All right — I’ll stop waxing poetic and get to it: For years now, the common narrative in the tech universe has been that Apple’s iPhones lead the way when it comes to longevity and the knowledge that the device you’re buying will continue to receive current software updates for many, many months to come.
Also for years now, we’ve talked about how that narrative is actually rather misleading — because comparing software updates on Apple and Android is akin to comparing apples and oranges, given the fact that nearly every single element that’s considered a significant part of an annual Apple iOS update is handled in an a la carte manner on Android, with multiple monthly updates that impact close to every still-functioning Android phone.
But that’s a complicated narrative to explain to the tech-totin’ masses — something far less neatly packaged and cohesive to wrap your noodle around than the simple “this phone gets operating system updates longer than that one” story Apple’s had in its corner all this time.
Today, my fellow Android-appreciating animal, that story is changing. For the first time in tech history, Apple is admitting — albeit quietly — the limits of its iPhone support setup. And it’s telling us, at least implicitly, that Google’s Pixel phones now hold the software support advantage and unambiguously offer a longer shelf life and better value.
That, suffice it to say, is quite the weighty twist in this high-stakes tech soap opera we’re all immersed in.
Let me explain.
Google Pixel phones and the evolving state of software support
Real quick, first, a pertinent pinch of perspective:
When we talk about the promise of ongoing software support — as in, official operating system and also security patch updates provided by a device’s manufacturer — Android phones have come a long way in a relatively short time.
Early on, most Android devices were given a mere year-and-a-half window of such support. It wasn’t until 2014, following a certain overly obsessed writer’s public outcries, that one Android-associated hardware-maker decided to take a stand and increase that standard up to a full two years. Then, sure enough, other Android-associated companies slowly but surely followed suit.
(It’s worth noting that Samsung followed suit with a similar seven-year support promise, though its version seems to apply only to its flagship-level phones and not the midrange or budget models. Samsung also makes no guarantees around the timeliness of rollouts and often does less than admirably in that regard, especially when it comes to its older models — and the company shifts to sending security updates only quarterly or sometimes even biannually once a device is more than a few years old. All of those reasons combined are why I’m focusing on Google and its Pixel promise here, as it truly is in a league of its own in this area.)
Throughout all this time, the narrative of Apple offering an unmatched support experience for the iPhone has remained — despite the fact that Apple itself has actually never made any explicit or specific promise about how long it’ll support any given iPhone model. Analyses have suggested the typical window is somewhere around six years, but Apple notably has never said a peep or offered up any concrete assurances.
Until now.
As a result of a new tech product security regulation that came into effect in the UK this spring, Apple was forced to put its money where its mouth is and put out a specific promise for how long it’ll support each iPhone with ongoing software rollouts. The official info is published on Apple’s own website, albeit in a place where no normal person would ever look or find it. But — oh, yes — it’s there, all right.
And here’s the actual iPhone software support promise for the company’s current top-of-the-line iPhone 15 Pro Max model — black and white and as plain as can be:
Apple’s official software support promise: a “minimum five years” for the top-of-the-line iPhone 15 Pro Max.Apple
Five years. That’s it — and that may be the biggest, most echoing admission Apple’s ever made.
Pixel vs. iPhone lifespan math
Now that Apple’s moved past its magically vague thinking and into specific numbers around its software support promise, we can put all of this into proper perspective.
A top-of-the-line iPhone 15 Pro Max will cost you $1,200. And with five years of guaranteed software support, that comes out to $240 for every year of fully advisable use that device provides.
A top-of-the-line Pixel 8 Pro, in contrast, starts at $999. With seven years of guaranteed software support, that comes out to roughly $143 a year for the device’s entire advisable lifespan. (And that’s not even taking into account all the ongoing system-like updates the device will continue to receive even after that seven-year period, as we were discussing a moment ago — a meaningful advantage the iPhone also can’t match.)
All value calculations aside, what’s truly striking here is the seismic shift in such a long-standing mobile market perception. Plain and simple, Apple’s iPhones are no longer leading the way when it comes to longevity and shelf life. Google’s Pixel devices have taken that crown — and while Tim Cook and company certainly won’t be crowing about that at their upcoming adjective-slinging supershow, Apple is essentially admitting that it’s fallen behind and allowed Android to lead the way.
And you’d better believe that matters, in more ways than one.
The true value of Google’s Pixel support pledge
Ever since Google announced its seven-year support promise for Pixel phones, I’ve seen more than a few fly-by commentaries about how that doesn’t amount to much — ’cause as we’re seeing with features like the recent Gemini AI elements (for better or maybe for worse), older Pixel models don’t always get every new feature that’s announced for the platform.
That’s true. But it’s also missing the point about this update promise and what it’s actually all about.
When we’re talking about operating system updates and the monthly security patches that support ’em, it’s not so much the device-specific, often separately bundled surface-level features that are important. It’s the underlying OS-connected updates related to privacy, security, and performance that make the most meaningful difference.
And look, I get it: We regular Hominidae tend to focus on the features we can see when assessing a new software update on our devices. That’s understandable. But in terms of a long-term support promise and its impact on a product, it’s more the invisible under-the-hood stuff that counts the most.
New Android versions always contain a smorgasbord of such improvements — things that go beyond the little fixes provided in those separate monthly security patches, even, and make a phone safer, more protective of privacy, and more efficient in the way it runs. They also introduce both expansions and restrictions to APIs, which are what permit third-party apps to interact with your phone and data and perform a variety of advanced functions.
These sorts of improvements aren’t the most titillating topics to talk about, but they’re immensely important to a phone’s day-to-day operation — arguably more so than the surface-level feature additions that tend to command our attention. They’re the reason why most people in the know will tell you it isn’t a good idea to keep using a phone after it’s no longer actively receiving such updates.
Ultimately, if having all the latest and greatest features is important to you, then yeah: At a certain point, you’re gonna have to buy new hardware to support that. That’s par for the course.
The software support promise isn’t intended to change that or to tell you any phone you buy today will support every single still-theoretical new feature and capability that may come up for the next seven years. It’s meant to assure you that the device you buy today will remain up to date and current foundationally, in terms of those core privacy-, security-, and performance-oriented areas — and, sure, it’ll keep getting new features and interface improvements, too, more often than not. But the real value is the fact that the phone will remain up-to-date in the areas that matter most and will remain viable and advisable to use because of that.
The same is true even more so on the Apple side of things, within its five-year framework. There, new features are frequently held back from even relatively recent iPhone models, but the critical system-level components are updated in order to keep the phone current and safe to use.
But now, Android — and specifically Google, with its Pixel phones — is leading the way and setting the standard in that area. And while most average phone-owners are bound to remain unaware of that shuffle, it’s a significant shift and one that’s absolutely worth emphasizing.
And here’s the best part: If typical trends hold true, it’ll probably only be a matter of time before Apple feels compelled to follow suit and step up its own standard to avoid being outdone. This isn’t a competition between us, as people who prefer different types of technology. It’s a competition between the companies that create those devices.
And when one company pulls ahead like this, it’s people like us who ultimately win by reaping the rewards of that competitive pressure. Google may be taking the lead in the software support race for now, but with any luck, all of us at the ground level will be the beneficiaries of that change sooner or later — whether we use Google’s Pixel products or not.
No matter who you are or what type of device you like, that’s something we can all celebrate.
Don’t let yourself miss an ounce of Pixel magic. Sign up for my free Pixel Academy e-course to discover tons of hidden features and time-saving tricks for your favorite Pixel phone.
Intel is aiming its next-generation X86 processors at artificial intelligence (AI) tasks, even though the chips won’t actually run AI workloads themselves.
At Computex this week, Intel announced its Xeon 6 processor line, talking up what it calls Efficient-cores (E-cores) that it said will deliver up to 4.2 times the performance of Xeon 5 processors. The first Xeon 6 CPU is the Sierra Forest version; a more performance-oriented line, Granite Rapids, will be released next quarter.
The upgraded Xeon processors can enable 3:1 data center rack consolidation while delivering the same performance and up to 2.6 times performance-per-watt gains over their predecessors, according to Intel.
“The data center AI market is hyper focused on the impact of AI power consumption with increasing concerns around the environmental impact and impact on the power grid,” said Reece Hayden, a principal analyst for ABI Research. “Intel Xeon 6 will be used as the CPU head node within Gaudi-powered AI systems. Improved performance per watt and density will reduce the AI systems’ power consumption, which will be positive for AI’s total energy footprint.”
Greater rack density allows for data center consolidation, freeing up room to deploy AI-focused hardware to support training or inferencing, Hayden said.
A worker in the Intel Assembly Test facility in Kulim, Malasia, inspects Intel Xeon 6, Sierra Forrest processors with E-Cores.
Intel Corp.
Intel also took the wraps off its Lunar Lake line of client processors, which are aimed at the AI PC industry. The x86 chips use up to 40% lower system-on-chip (SoC) power compared with the previous generation, according to Intel.
The Lunar Lake Core Ultra processor line is expected to be available in the third quarter of this year; with neural processing units (NPUs) on board, the chips will have more than 100 platform tera operations per second (TOPS) and more than 45 NPU TOPS and are aimed at a new generation of PCs enabled for generative AI (genAI) tasks.
“Intel is one of the only companies in the world innovating across the full spectrum of the AI market opportunity — from semiconductor manufacturing to PC, network, edge and data center systems,” Intel CEO Pat Gelsinger said in a statement from the Computex Conference in Taiwan this week.
Intel also announced pricing for its Gaudi 2 and Intel Gaudi 3 AI accelerator kits — deep learning accelerators aimed at supporting training and inference of artificial intelligence large language models (LLMs). The Gaudi 3 accelerator kit, which includes eight of the AI chips, sells for about $125,000; the earlier generation Gaudi 2 has a list price of $65,000.
Accelerator microprocessors handle two primary purposes for genAI: training and inference. Chips that handle AI training use vast amounts of data to train neural network algorithms that then are expected to make accurate predictions, such as the next word or phrase in a sentence or the next image, for example. So chips are required to speedily infer what that answer to a prompt (query) will be.
But LLMs must be trained before they can begin to infer a useful answer to a query. The most popular LLMs provide answers based on massive data sets ingested from the Internet, but can sometimes be inaccurate or offer downright bizarre results, as is the case with genAI hallucinations.
Intel Corp.
Intel vs. Nvidia; who wins this fight?
Intel is hoping its Gaudi line of accelerators can rival Nvidia’s more expensive GPUs, which have rocketed the chipmaker into the status of leader in the AI processor marketplace. Nvidia has reported soaring revenue as its chips continue to dominate AI cloud services and data center rollouts of genAI applications.
Last year, Nvidia controlled about 83% of the data center chip market, with much of the remaining 17% claimed by Google’s custom tensor processing units (TPUs).
“Most estimates suggest these [Intel’s] prices are between one-third and two-thirds [the price] of competitors. This is highly indicative of their approach to the AI data center market — focusing on affordability and undercutting competitors,” Hayden said.
AMD and Nvidia do not discuss pricing of their chips, but according to custom server vendor Thinkmate, a comparable HGX server system with eight Nvidia H100 AI chips can cost more than $300,000.
Intel claims its Gaudi AI accelerators are a third less expensive compared to “competitive platforms” — namely Nvidia’s GPUs.
“Intel is certainly a direct competitor against NVIDIA in the AI market with products across data center/cloud, edge, devices,” Hayden said. “However, NVIDIA is hyper focused on data center accelerators and supporting training and inference at scale. They dominate this market with very high market share. Increasingly, as this market grows, Intel will grow its overall share, but NVIDIA is still likely to dominate.”
Conversely, Intel’s goal is to enable “AI everywhere” with a major focus on edge and device, especially PC AI (Intel Core Ultra). NVIDIA has been less focused in that area as data center/cloud GPUs are still a fast-expanding market, Heyden noted.
Smaller LLMs — an opening for Intel?
The LLMs used for genAI tools can consume vast amounts of processor cycles and be costly to use. Smaller, more industry- or business-focused models can often provide better results tailored to business needs, and many end-user organizations and vendors have signaled this is their future direction.
“Intel [is] very bullish around the opportunity of smaller LLMs and [is] looking to embed this within their ‘AI everywhere’ strategy. ABI Research agrees that enabling AI everywhere requires lower power consumption, less cost-intensive genAI models, coupled with power efficient, low TCO hardware,” Hayden said.
While its x86 line of chips won’t actually run AI processes, the combination of Xeon 6 processors with Gaudi AI accelerators in a system can make AI operations faster, cheaper and more accessible, according to Intel.
Forrester Senior Analyst Alvin Nguyen agreed that Intel’s strategy of targeting smaller LLMs and edge devices is a smarter bet than attempting to go head-to-head with Nvidia in cloud data centers, where increasingly larger LLMs are prevalent.
“The approach Intel is taking with AI gives them a reasonable chance of competing with Nvidia: they are not trying to replicate what Nvidia is doing,” Nguyen said. “Instead, [Intel is] taking advantage of their breadth of technology coverage and current generative AI approaches to provide an alternative approach that will be more appealing to the enterprise.
“Inference is where enterprises are at and Intel is well positioned with the supply chain issues to entrench themselves here: ‘good enough’ performance, lower costs, and ubiquity of products help them here,” Nguyen said.
The server and storage infrastructure needed for training extremely large LLMs is expected to take up an increasing portion of the AI infrastructure market, according to IDC. The research firm projects that the worldwide AI hardware market (server and storage) will grow from $18.8 billion in 2021 to $41.8 billion in 2026, representing close to 20% of the broader server and storage infrastructure market.
More than a dozen current and former AI industry employees have signed an open letter warning that the technology’s dangers could result in “human extinction.”
The letter was written by 13 people who have worked at OpenAI, Google DeepMind, and Anthropic, all leading providers of generative AI (genAI) technology. Specifically, they raised alarms about a series of risks from AI, ranging from “further entrenchment of existing inequalities, to manipulation and misinformation, to the loss of control of autonomous AI systems potentially resulting in human extinction.
“AI companies themselves have acknowledged these risks, as have governments across the world and other AI experts,” the letter says. It also calls for assurances that employees that do raise concerns will not be retaliated against by their companies.
The letter also received the endorsement of AI scientist Yoshua Bengio, British-Canadian computer scientist and psychologist Geoffrey Hinton, and University of California Computer Science professor Stuart Russell.
The latest missive echoes an open letter released In March 2023, when more than 150 leading AI researchers and others called on genAI companies to submit to independent evaluations of their systems, the lack of which has led to concerns about basic protections. Later that same month, more than 1,000 signatories including industry experts, scientists, ethicists and others, posted an open letter warning about a possible “loss of control of our civilization” from unchecked AI.
The new letter noted there is no effective government oversight of corporations creating and selling AI solutions. “Current and former employees are among the few people who can hold them accountable to the public. Yet broad confidentiality agreements block us from voicing our concerns, except to the very companies that may be failing to address these issues,” the signatories said.
“Ordinary whistleblower protections are insufficient because they focus on illegal activity, whereas many of the risks we are concerned about are not yet regulated. Some of us reasonably fear various forms of retaliation, given the history of such cases across the industry. We are not the first to encounter or speak about these issues.”
The employees laid down four specific measures they want from companies to ensure the safety of genAI technology:
The company will not enter into or enforce any agreement that prohibits “disparagement” or criticism of the company for risk-related concerns, nor retaliate for risk-related criticism by hindering any vested economic benefit.
The company will facilitate a verifiably anonymous process for current and former employees to raise risk-related concerns to the company’s board, to regulators, and to an appropriate independent organization with relevant expertise.
The company will support a culture of open criticism and allow its current and former employees to raise risk-related concerns about its technologies to the public, to the company’s board, to regulators, or to an appropriate independent organization with relevant expertise, so long as trade secrets and other intellectual property interests are appropriately protected.
The company will not retaliate against current and former employees who publicly share risk-related confidential information after other processes have failed.
“We accept that concerns should be raised through such a process initially. However, as long as such a process does not exist, current and former employees should retain their freedom to report their concerns to the public,” the letter said.
The EU AI Act required genAI systems to meet transparency standards to help regulators and others distinguish deep-fake images from real ones. The measure also prohibited social scoring systems and manipulative AI.
In the United States, there have been several efforts to curb AI, but no meaningful legislation from Congress. For example, in October 2023, US President Joseph R. Biden Jr. issued an executive order that hammered out clear rules and oversight measures to ensure AI is kept in check, while providing paths for it to grow. Among more than two dozen initiatives, Biden’s “Safe, Secure, and Trustworthy Artificial Intelligence” order was a long time coming, according to AI industry experts who’ve been watching the rise of genAI tools and platforms since late 2022.
One day when we look back at the last few decades in tech, we will realize that one of the most influential products ever released was, in fact, one that failed. Years ahead of its time, the Apple product unleashed many of the technologies that define consumer electronics today.
What is it? The Apple Newton.
Now, hear me out.
What was the Apple Newton?
Developed by Apple, the Newton was one of the first personal digital assistants ever made. A handheld gadget capable of understanding handwriting, it was years ahead of its time when it appeared in 1993. But cost and limited functionality meant it was eventually canned when Apple co-founder Steve Jobs returned to the company and closed the loss-making Newton product division down in 1998 (as he struggled to rescue Apple from extinction).
Part of the original Newton team, Albert Chu, spoke with me about the Newton in 2003. “Newton was good technology,” said the then-vice president of business development at Palm. “It had a lot of great features, but when we launched it, it was not launched as part of Apple.
“Yes, it was ahead of its time and was a great exploration, but it was just not ready for primetime: handwriting recognition did not work, for example.”
It also suffered from a lack of connectivity, which blunted its potential.
While it didn’t succeed, Newton still casts a shadow over today’s tech industry, not least because its designers included many key members of the iPhone design team, including Jony Ive and a who’s who of luminaries who helped craft the Mac. (Newton even throws a little shade across generative AI as it had its own intelligent assistant tech that attempted to let you perform tasks using natural language.)
But perhaps the most important contribution to today’s tech world is something you can’t easily see.
From the acorn grows
While developing Newton, Apple was forced to search high and low for a suitable processor. That story had a few twists and a couple of turns, but in the end a small UK firm called Acorn was selected. That company had created a processor that delivered excellent computational performance at low power. Apple invested $3 million in Acorn to help it build a new revision of its Acorn RISC Machine chip. That chip ended up powering Newton, and also became the name of the company: Arm.
You should be in no doubt about the importance of the chip reference designs Arm continues to create. These reference designs are then tweaked and adopted by other companies, including Apple, which uses Arm-based chips across iPhones, Macs, and iPads. Arm CEO Rene Haas, recently explained how Apple’s move to use its reference designs in Macs, “woke up the industry on the art of the possible.”
This acorn grew.
…And fell a little further from the tree
Arm’s designs are also the processor reference designs emerging in Apple Silicon competitors, including Qualcomm’s current hot hope, the Snapdragon X series, which is arguably of fundamental importance to the future evolution of the wider PC industry.
What matters about these processors is that they deliver computational performance at low energy — precisely the challenge Apple had to solve with the Newton. It’s a challenge that seems even more relevant today, as both energy security and climate change demand a reduction in energy use, even while the ongoing AI revolution poses ever larger strains on energy supply.
Computational power at lower performance was relevant for the world’s first PDA at the close of the last century; these days it is becoming a more complex and existential problem that impacts every industry and every person.
A moral debt
To this day, the PC industry carries a moral debt to Apple for the development of the graphical user interface (GUI), and adoption of keyboard and mouse. That tomorrow’s PC industry now seems to rely on Arm chip designs shows us the influence of Apple’s failed product was — and remains — incredibly important.
It is hard to ignore that if robust mobile internet had been available at the time the Newton appeared, we would almost certainly still be using these devices today. To some degree, we do.
Think of the many Newton technologies adapted and improved by Apple for use in iPhone, and the extent to which competitors also chose to emulate that device in their own smartphones.
Think of the processor inside these devices and the importance of their progeny to the future of the big PC firms. Think of the user interface and ideas pertaining to mobile productivity the Newton first explored (albeit replacing the stylus with a finger).
Think of all these things and it should be clear that this important and influential product deserves a very special place in tomorrow’s history of today’s technology. Now, spend a few minutes watching Apple’s early 90’s promo video for Newton and think how many of the technologies it championed are driving your day today.
Developers and DevOps teams must manage many moving parts to deliver an exceptional product on time. Developers have adopted effective tooling such as Jira and Asana to manage and facilitate structured collaboration. But when teams need to solve problems that require creative thinking and don’t have a predetermined outcome — such as determining the goals for the next sprint, understanding the user flow, or brainstorming ideas for a new product — standard conferencing and messaging tools don’t provide enough support.
Visual collaboration is a new category of tools that facilitates and enhances fluid, roll-up-your-sleeves collaboration between team members, whether they are physically together or connecting virtually. During the course of a discussion, it’s simple for team members to visually represent and connect systems, practices, and next steps. Through diagramming and whiteboarding, participants produce a record that’s easily accessible to everyone, and as the conversation continues, changing these diagrams is just as easy. Even more powerful, in some tools, AI can pitch in to create diagrams or help sort, filter, summarize, and refine ideas.
Since visual collaboration tools are an emerging category, however, it may not be immediately obvious how development teams could take advantage of them. Here are four examples:
Ranking and visualizing aggregate responses: Visual collaboration tools can identify conflicts that are hard to ferret out in other mediums. Let’s say the team is determining which features to develop in the next sprint for a new product. Two of the participants said that building shopping cart capability should be the aim of the next sprint, while the rest of the team ranks it near the bottom, an insight that can spark an important discussion.
Starting from scratch: Starting from a blank page can be very hard. AI-powered visual collaboration tools can create a diagram from a simple prompt, such as “Create a user flow diagram that has the following characteristics.” From there, the team can easily adjust, expand, and edit the diagram until the team creates a user flow that works for the project.
Clarifying complex processes: Sometimes, the problem isn’t staring at a blank page but rather having far too much information to process. In this case, intelligent systems can connect to other data sources and visually represent the content, giving the team a much more intuitive grasp of processes.
Visualizing infrastructures: Let’s say the team is about to start working on a mobile app, and they need a diagram of the data flows and what’s nested within AWS. Intelligent systems can analyze the cloud infrastructure and produce a diagram automatically. But this capability is not limited to IT infrastructures. You could just as easily import an org chart and view it by scrum teams to highlight all the senior engineers or people who possess specific skill sets.
Lucid provides advanced and intuitive visual collaboration tools to nearly every Fortune 500 company, enabling them to increase productivity, become more efficient, and collaborate more effectively in hybrid teams and with other departments. In fact, an ROI study conducted by Forrester showed a 410% ROI over three years and more than three-quarters of a million dollars of savings in operating expenses.
For more information on how Lucid can help your teams collaborate more effectively, visit lucid.co.
