TheChek News Flow

Earnings season is in full swing, bringing with it the potential for big gains —or potentially crushing losses. When trying to gauge the risk/reward, pay attention to the starting point, particularly right now.

Experienced traders know that a beat isn’t always a beat, nor a miss a miss. On Thursday, for instance, American Airlines Group slumped 9.6%, IBM dropped 7.7%, and Honeywell International fell 6.2%—all of them after reporting earnings that topped expectations. All three had posted double-digit gains over the previous three months. Conversely, Raymond James Financial rose 3.7% and CenterPoint Energy rose 1.9%, despite reporting profits that fell short.

There’s a lesson in there for investors. Stocks that have gained too much are more susceptible to big drops than companies that dropped heading into their results. “We like stocks that are relatively oversold coming into earnings,” says Fairlead Strategies founder Katie Stockton. “They theoretically have a lower bar set in price terms as investors absorb the earnings data.”

“Oversold” is a technical term that essentially means stocks have gone down a lot, quickly, reflecting that bad news has at least partially been discounted in share prices. Investors don’t have to do a lot of complex math to figure out what to watch out for, though. Stockton recommends looking at shares that have underperformed the S&P 500 since the April lows, reached just after President Donald Trump’s Liberation Day tariff announcements.

Six stocks reporting this coming week that fit the bill include communications infrastructure provider American Tower REIT, Visa, Procter & Gamble, Merck, Oreo maker Mondelez International, and drugstore chain CVS Health, according to Fairlead’s analysis.

Coming into Thursday trading, those six were up an average of just 3% from April lows, about 22 percentage points behind the S&P 500. That’s the set up for a potential post-earnings bump.

Those calls are based on the charts—that is, technical analysis —but fundamentally minded analysts like those stocks too. The average Buy rating ratio for the six is about 72%, according to FactSet, well above the 55% average for stocks in the S&P 500.

Overbought stocks, however, have a higher bar to clear when they report earnings this coming week. Fairlead’s list includes a who’s who of market dominance— Amazon.com, Meta Platforms, and Microsoft —as well as Boeing, cruise operator Royal Caribbean Group, and disk-drive maker Seagate Technology Holdings.

Again, there is nothing wrong with those six. Shares have simply been on fire, up 66% on average since the April lows. Performance like that means it will take a pretty darn good earnings report to send stocks up further. That’s not out of the question— GE Vernova, for instance, surged almost 15% on Wednesday after earnings despite having gained more than 90% since the April lows. And the six stocks on the overbought list are all Wall Street darlings—the Buy-rating ratio for Meta, Microsoft, and Amazon is close to 90%.

Still, it always pays to know where you started.

If you thought the European Union’s top export to the U.S. was cars, think again.

The EU shipped 120 billion euros ($140 billion) worth of “medicinal and pharmaceutical products” across the Atlantic last year, three times the volume of auto exports. The bloc ran a $100 billion surplus in the category. Switzerland sold American patients another $35 billion.

Four of the top 10 holdings in the iShares Europe exchange-traded fund are pharma producers: Roche Holding, Novartis, AstraZeneca, and Novo Nordisk.

Investors have marked down this juggernaut’s future lately. The Stoxx Europe Total Market Pharmaceuticals & Biotechnology index has swooned more than 20% from a peak in September, while broader European stocks are up 10%. “Volatility should lead to a flight to safety in sectors like healthcare,” notes Michael Field, European equity strategist at Morningstar. “That hasn’t been the case.”

U.S. President Donald Trump is clouding the sector’s immediate future with somewhat contradictory initiatives. He has threatened 200% tariffs on imported medicines. He also issued an executive order aimed at making drugs cheaper by “bringing American prices in line with those paid by other nations.” Follow-up in either direction could bite into European pharma’s profits.

Vaccine skepticism emanating from Health and Human Services Secretary Robert F. Kennedy Jr. is another potential threat. Companies like London-headquartered GSK and France’s Sanofi earn nearly 30% of global revenue from vaccines.

Longer term, a venerable industry is resting on rapidly withering laurels, warns Andy Powrie-Smith, communications director at the European Federation of Pharmaceutical Industries and Associations. “Europe absolutely used to be a superpower in pharma,” he says.

The Continent’s global share of “new treatments” has fallen to 20% from half 30 years ago, he notes. Its share of clinical studies for new drugs has dropped by half over the past decade as research migrates to the U.S. and China. The EU is “falling behind in the most dynamic [pharma] sectors and losing market share to U.S. companies,” noted former European Central Bank President Mario Draghi in a monumental review of European competitiveness issued last year.

Trade association EFPIA’s chief remedy—“significantly increasing what Europe spends on medicines and improving access to medicines for patients”–seems an uphill political battle at best, however. How much of all that is in the price? Morningstar’s Field ranks pharma as Europe’s second-cheapest sector, at an average 14% discount to fair value. Both the global pharma pie and Europe’s slice keep growing as the world gets older and sicker. EU exports in the category jumped 14% last year.

Old World companies keep generating new blockbusters, too. Novo Nordisk’s weight-loss breakthrough Ozempic was the No. 2 drug worldwide last year, with $17 billion in sales. Sanofi co-developed the No. 3 seller, anti-inflammatory Dupixent.

Morningstar gives four European pharma stocks its top five-star rating at current prices. Roche and GSK (despite vaccine exposure) are its picks among the big names. Smaller stars include Elekta, a Swedish producer of radiotherapy devices, and Danish equipment maker Coloplast. “The pipelines for a lot of these companies are still there,” Field says.

UBS analysts put Buy ratings on four out of seven companies covered in a recent report: Roche, Novo Nordisk, Sanofi, and AstraZeneca. Any value-driven rally will still wait until the fog from Washington clears a bit, Field expects. “The danger is this could be a 12- to 18-month story, not a three-month story,” he says.

Homework now may pay off later, though.

Shares have dropped 10% since the company’s annual meeting.

erkshire Hathaway stock has been in a funk since the company’s May 3 annual meeting, when CEO Warren Buffett said he would step down as CEO at the end of the year. That has created an opportunity for investors to follow his advice to buy when others are fearful.

Shares of Berkshire have declined over 10% since the meeting, trailing the S&P 500 index by more than 20 percentage points after outperforming at the start of the year. Several factors are behind the flagging performance: an erosion in the “Buffett premium”; concern that the property-and-casualty insurance cycle has peaked; scant new investment activity by the company; no stock buybacks in more than a year; and a recent shift away from defensive stocks like Berkshire.

Berkshire, though, still has a lot going for it. Its main businesses remain among the most dominant in their industries, while its stock portfolio still has a diverse mix of companies that could outperform if the market hits a speed bump. And it has a fortress balance sheet with over $330 billion in cash, a third of the company’s market value of $1 trillion, which could be used for buybacks, a dividend, and even a large deal—the purchase of railroad operator CSX is a real possibility.

“It’s getting to be more compelling now,” says Mac Sykes, a portfolio manager at Gamco Investors.

Berkshire’s shares are no bargain, but they look appealing in a richly valued stock market. The company’s Class A shares, at around $725,000, trade for less than 1.6 times the consensus estimate of the company’s June 30 book value of $461,140—in line with the average of recent years. The company’s Class B shares recently traded at $480.

The stock trades for about 24 times projected 2025 earnings, in line with the S&P 500. Berkshire’s price/earnings ratio is even lower, at around 20, if earnings are adjusted upward to account for the profits of Berkshire’s equity portfolio companies, including Apple, Coca-Cola, and Bank of America. This non-GAAP measure, called look-through earnings, has been endorsed by Buffett.

Berkshire is generating about $45 billion in annual operating earnings. That should help lift book value to perhaps $525,000 per A share by the end of 2026, meaning the stock trades for a reasonable 1.4 times forward book value.

“It’s a great stock to hold in this uncertain environment,” says UBS analyst Brian Meredith, who has a Buy rating and a price target of $892,120 on the A shares, up 23% from current levels. “Berkshire is no different than it was in April, before Buffett said he would step down.”

Buffett, 94, isn’t going away just yet— he plans to remain chairman in 2026. But no matter his role, the company’s three major businesses—insurance, railroads, and electric utilities—are in good shape.

Berkshire is one of the world’s largest property-and-casualty insurers, and while price increases have eased this year, they remain in the 4% to 5% range, auguring well for future results. It also owns Geico, the No. 3 auto insurer. Berkshire’s former insurance problem child, Geico has been largely fixed thanks to a technology overhaul and now is highly profitable and poised to grow.

Berkshire’s utility business is one of the biggest in the country and encompasses regulated electric utilities, a large renewable-energy portfolio, an electricity transmission network, and several natural-gas pipelines. Known as Berkshire Hathaway Energy, it’s deploying $10 billion a year of capital for a slew of projects and stands to benefit from the artificial-intelligence boom.

Berkshire’s Burlington Northern Santa Fe, together with rival Union Pacific, dominate freight railroads in the Western U.S. Union Pacific’s talks with Norfolk Southern about a merger create the possibility that Berkshire would buy CSX, the other major Eastern railroad, and create a transcontinental railroad. Such a deal, which would face considerable regulatory scrutiny, could cost Berkshire $80 billion or more, assuming a 25% premium to the current CSX price, but be 8% accretive to Berkshire’s 2026 earnings, according to UBS’s Meredith.

Berkshire’s other key asset—its $300 billion equity portfolio—is probably lagging behind the market this year, as gains in American Express, Bank of America, Coca-Cola, and Chevron are offset by losses in its largest holding, Apple, which is down almost 15%.

Still, it may take more than solid businesses and hopes for a bounce to get Berkshire stock moving again on its own. Buffett could help investors understand what management will look like after he steps down. Berkshire has said nothing about Greg Abel’s executive team once the head of Berkshire Hathaway Energy takes over as CEO, and it’s unclear whether Ajit Jain, the 73-year-old head of the insurance business, will stay on. The status of Todd Combs and Ted Weschler, who together run about 10% of Berkshire’s equity portfolio, is also unclear.

“Investors are highly sensitive to the changing of the guard, and the Buffett premium is being extracted from the stock,” CFRA analyst Cathy Seifert says. “The lack of clarity is not helping.”

A major acquisition, such as the aforementioned CSX or perhaps Occidental Petroleum, could do the trick. While Buffett says he doesn’t want to own all of Occidental—Berkshire now holds 27%—his successor, Abel, may see it differently, and Occidental CEO Vicki Hollub has said she would love the idea of being owned by Berkshire. Such a deal could cost $45 billion, well within Berkshire’s means.

Another possibility: Berkshire owns 27% of Kraft Heinz stock, and the big food company said in May that it’s considering strategic alternatives. Berkshire could swap its stock for the old Heinz business. Buffett, after all, is a big fan of Heinz ketchup.

A renewed share-repurchase plan would signal that Buffett views the stock as attractively priced, and it’s something the company hasn’t done since May 2024. It has been even longer since Abel, who owns $167 million worth of Berkshire shares, made a large personal stock purchase. His last buy came in March 2023, when the A shares traded around $450,000. Aligning himself more closely with shareholders would be a winning signal as he becomes CEO. Berkshire could even start paying a dividend—given its cash and earnings, it can afford a 2% payout—like most companies its size.

With so many levers left to pull, don’t count Berkshire out yet.

• Theme Parks Are the New Engine: Disney's Experiences division, including theme parks and cruises, now drives nearly 60% of its operating profit, offsetting past declines in traditional TV.
• Streaming Is Growing, But Not Dominant: Disney+ has become profitable and is expanding globally, though it's still far behind Netflix in terms of scale and usage.
• ESPN Shift to Streaming: A full ESPN streaming service ($29.99/month) is launching soon, targeting cord-cutters and potentially reshaping Disney’s sports revenue model.
• Improving Financials: Disney's operating profit has rebounded to near record highs ($15.6B in 2023), with Wall Street projecting $17.5B this year and continued double-digit growth.
• Stock Momentum Has Returned: After a lost decade, Disney stock is up 38% over the past year, outperforming the S&P 500 and showing renewed investor confidence.
• 
  

With TV declines baked in and growth improving, the stock is ready for a new era.

Walt Disney shareholders are coming off a lost decade. The stock, at a recent $121, is trading where it was on Aug. 4, 2015. On that day, the company lowered forecasts for its financial results, citing ongoing subscriber losses at ESPN. The cable sports network was Disney’s biggest moneymaker at the time, even if the public associated the company more with Mickey Mouse, roller coasters, and superhero movies. Shares lost 9% in a day. Discovery, Time Warner, Viacom, Fox, Comcast, and CBS tumbled, too. Consumers had been leaving pricey cable television bundles for streaming for years, but if ESPN was now vulnerable, so was the entire traditional television economy.

Today, Disney appears more resilient—not because ESPN has held the line, but because it is diminished enough to reduce the risk of further declines. There is a new ESPN streaming service launching soon. The core of Disney’s streaming business is profitable and growing. Movies are slumping, but there are dependable box-office performers coming. Theme parks are mostly bustling. Disney, not for the first time in its history, is transforming. “It was a media company that owned theme parks,” says Morgan Stanley analyst Benjamin Swinburne. “Now it’s a theme park company that owns media assets.”

Wall Street is overwhelmingly bullish, with more than three-quarters of analysts who cover Disney stock recommending a purchase. That’s higher than the percentage who like Netflix. Careful, there. Disney was the more popular of the two stocks on Wall Street a decade ago, too. Since then, Netflix has returned 955%, versus 253% for the S&P 500 index, while Disney, including dividends, has eked out 10%.

But those figures include signs of life from Disney shares lately. They have returned 38% in a year, beating the S&P 500 by 19 points. The stock has returned a total of 45% since Barron’s last featured Disney on its cover in July 2023, beating the index by three percentage points over that period.

If that’s the start of a larger revival, it’s not because Disney stock is particularly cheap. It goes for 21 times projected earnings for the company’s current fiscal year running through September, only slightly below the broader market.

But after years of diminished profits, the outlook for growth is now bright enough.

Disney won’t catch up with Netflix in streaming on any timeline that is relevant to investors, and in the long run, both might be vulnerable to Alphabet’s YouTube. But Disney is also in better shape than the rest of Hollywood, thanks to its flourishing theme parks and a shot at success in streaming. For now, continued double-digit yearly stock returns look feasible.

Theme parks got Disney into TV, you might say. Founder Walt Disney dreamed up Disneyland while sitting on a Los Angeles park bench watching his young daughters play on a merry-go-round. But his animated film business needed help to cover what would ultimately be a $17 million construction bill. So in 1954, a year before the park opened, Walt cut a deal with his industry’s rising competitor, TV. He would make a show called Disneyland, and later The Mickey Mouse Club, for ABC, in exchange for funding and on-air promotion. Three decades later, Disney embraced TV’s thriving new pay model when it launched the Disney Channel on cable. But there was a much bigger move coming.

In 1995, Disney surprised Wall Street by announcing that it would buy the parent company of ABC for $19 billion, then the second-largest corporate takeover. A recent regulatory change meant that networks were no longer prohibited from owning their prime time shows. The deal brought ABC’s 80% stake in ESPN, along with ABC head Bob Iger, who would become Disney’s CEO in 2005.

Last fiscal year, linear networks, as traditional broadcast and cable are now called, contributed 22% of Disney’s operating profit. But they once brought in close to half. The economic miracle of cable TV was that customers would pay for programming and endure advertising. ESPN’s audience skews young and watches live, making it particularly attractive to advertisers. And since networks collect fees from cable bundlers, not customers directly, the cost is borne by all households, not just those that watch sports. ESPN became the priciest channel in the cable bundle and easily the biggest contributor to Disney’s TV profits.

“I would sit with portfolio managers…and they’d say, ‘Oh, I love Disney,’ and I’d say, ‘Why do you like Disney?’ and they’re like, ‘Oh, it’s just a fantastic media conglomerate,’” says Citigroup analyst Jason Bazinet. “I used to say back, ‘No, it’s ESPN masquerading as a well-run media conglomerate.’” But a cash cow can fix a lot of problems. Two decades ago, Disney was a broken movie studio with tired theme parks. ESPN’s profits played no small part in funding Iger’s acquisitions of Pixar in 2006, Marvel Entertainment in 2009, and Lucasfilm in 2012, setting up the biggest box-office boom in history, and fueling no end of theme park tie-ins and expansions. “A bad media conglomerate became a good media conglomerate,” says Bazinet.

Now theme parks are the cash cow. Last year, Disney’s Experiences division, which includes its parks, turned in a record operating profit of $9.3 billion. That’s 59% of the company’s total operating profit of $15.6 billion. This year, the Experiences unit is expected to earn 8% more, thanks in part to the launch of two new cruise ships. The company recently announced distant plans for a new park in the United Arab Emirates, which could boost its brand throughout the region.

For now, if Disney has pushed theme park prices too high; or if a new Universal park in Orlando is luring away Magic Kingdom visitors; or if macroeconomic fears are hitting vacationers; or if the “go woke, go broke” movement is keeping visitors away, it isn’t especially evident in park financial results.

Movies, however, have become meager earners. A live-action Snow White remake this year flopped. A Pixar release called Elio did so poorly that moviegoers might not even realize that it came and went in June. Then again, a May live-action remake of the animated 2002 film Lilo & Stitch performed beautifully. Morgan Stanley’s Swinburne has a simple explanation for weak theater results for Disney and others: too few releases.

Disney’s high-water mark for companywide operating profit was set at $15.7 billion back in fiscal 2016, and was nearly matched in 2018, before traditional TV’s decline accelerated. The low point since then was $7.8 billion in 2021, when parks were running pandemic losses and the ramp-up of the Disney+ streaming service was burning through cash. Last year, Disney’s operating profit rose 21% to $15.6 billion, as streaming turned from losses to a negligible profit. This year, Wall Street is predicting a further 12% increase to $17.5 billion, finally blowing past the record. Over the next three years, Disney could tack on an additional $5 billion in yearly operating profit.

This will come mostly from two sources: steady growth from a large base of profit in Experiences, and rapid growth from a small base in Direct to Consumer, as Disney calls streaming. Even so, Disney’s streaming in a few years will produce only perhaps a fifth of the operating profit of Netflix. That is reflected in Netflix having more than double the stock market value of Disney.

“They’re not industry-leading and they’re not industry-lagging,” says Citi’s Bazinet, of Disney’s streaming efforts. “They’re sort of just stuck.” He calls Netflix a true pay TV substitute, because it has high daily customer usage for the price. He compares Peacock TV and HBO Max, which get less usage for the price, to the old HBO; customers might watch a few shows and then cancel. Disney is squarely in the middle. Back when Netflix burned cash for a decade in streaming, interest rates were exceptionally low, and investors cheered.

When Disney went all in on streaming, it briefly basked in the same glow. As profits plunged, the stock hit a record high price above $200. Then rates jumped, and investors demanded profits. Netflix was far enough along on its growth that it could easily fund vast content investments from its cash flow. But most of the others have been left more constrained. “A lot of these streaming services are sort of half-pregnant,” says Bazinet. “They sort of didn’t run the distance to become a pay TV substitute.”

Much of media looks worse off than Disney. Shares of Paramount Global, formed when Viacom merged with CBS, are down 70% over the past decade, and the controlling family is pursuing an unpopular sale. Warner Bros. Discovery, down nearly 50% since it began trading in 2022, is splitting its cable networks from its studios and streaming. Comcast, which has made 43% over the past decade, mostly from dividends, has prospered in theme parks while struggling in TV, and faces rising competition from telecom companies for its cable broadband hookups. It’s expected to produce minimal operating profit gains in the years ahead.

At Disney, Iger’s current CEO contract runs through the end of 2026. Josh D’Amaro, the Experiences chief, is a likely successor. Other speculated-about names include Alan Bergman and Dana Walden, who have leadership positions in entertainment, and James Pitaro, who runs ESPN.

For now, Iger will run Disney to maximize profit and the stock price. “He wants to go out as the hero, right?” says Bazinet.

The next boss will have to decide whether to take another hit to profits to make a serious run at Netflix. The good news is that Disney has already done the hard work of setting up a successful streaming service and stocking it with high-profile intellectual property. What is needed, analysts say, is the endless reservoir of less-differentiated content that can keep viewers around once they have watched their favorite hits. There is also a long runway for Disney to expand its streaming overseas.

For ESPN, too, there’s good news of sorts. Its reach is down to 65 million subscribers from a peak of over 100 million, and its contribution to company operating profit is about 15%, down from a high of perhaps 25%—Disney didn’t begin disclosing ESPN’s financial particulars until 2022. The lower profit base helps to reduce risk from the network’s move to streaming in early fall. Don’t confuse the current ESPN+, which has add-on content but not live ESPN, with the full streaming service, which will be called simply ESPN and cost $29.99 a month. It’s aimed at the 60 million and growing base of cord-cutters and cord-nevers, who haven’t yet had the option to buy true ESPN outside of bundled TV.

ESPN streaming could accelerate cable cancellations. But the network telecasts only about one-third of sports events, so rabid sports fans might prefer to hold on to their full cable bundles to get their other games. They will be entitled to use the new ESPN streaming service at no extra cost, and they might choose to do so, because unlike cable, it is expected to include side features, like betting and fantasy leagues. This would give Disney the ability to both preserve existing cable relationships and to quietly lure sports viewers into its ecosystem, where it can learn from the data they generate.

YouTube, which tops both Netflix and Disney in viewership with its user-created videos, has migrated onto television sets, and is dabbling in live sports and original shows. Hollywood had better keep an eye out. But for Disney, for now, the bar for TV has likely been set low enough to step over. So long as skies remain sunny for the rest of the business, Disney, transformed once again, can move on from its lost decade for good.

The luxury giant has been holding talks with the owner of Reebok, among others

LVMH MC 3.92%increase; green up pointing triangle Moët Hennessy Louis Vuitton is in discussions to sell fashion brand Marc Jacobs in a deal that could fetch around $1 billion, according to people familiar with the matter.

The details

The luxury giant has been discussing deals with multiple parties including Reebok owner Authentic, the people said. Brookstone owner Bluestar Alliance and Vera Wang parent WHP Global are also among the suitors.

A deal could come together soon, assuming talks don’t fall apart, the people said.

LVMH is widely known as an acquirer of assets to fuel growth, such as its big deal for jeweler Tiffany & Co. in 2021. But the company has proven to be a willing seller of brands if they no longer fit its strategy. In 2016, it agreed to unload Donna Karan and DKNY for an enterprise value of $650 million. More recently, LVMH sold its interests in Off-White and Stella McCartney.

“We will not keep brands if we believe they are not a good add-on, or we are not the right operator to operate them,” LVMH Chief Financial Officer Cécile Cabanis said during the company’s earnings investor call on Thursday.

The Marc Jacobs fashion brand was founded in 1984 by the eponymous fashion designer, alongside his business partner Robert Duffy. The label’s most popular products include its tote bags and Marc Jacobs Daisy perfume.

The context

A sale of Marc Jacobs would come as LVMH has been trying to resurrect the brand, in part by trimming the number of items it offers to customers.

The conglomerate, steered by Bernard Arnault, is home to dozens of high-end labels including Dior, Celine, Fendi and Hennessy cognac.

There has been a flurry of dealmaking activity in the luxury retail space. Prada in April said it would acquire Versace from fashion conglomerate Capri Holdings for almost $1.4 billion.

The New Chips Designed to Solve AI’s Energy Problem

Tech giants and startups alike are trying new approaches to what has been a vexing problem

“I can’t wrap my head around it,” says Andrew Wee, who has been a Silicon Valley data-center and hardware guy for 30 years.

The “it” that has him so befuddled—irate, even—is the projected power demands of future AI supercomputers, the ones that are supposed to power humanity’s great leap forward. Wee held senior roles at Apple and Meta, and is now head of hardware for cloud provider Cloudflare. He believes the current growth in energy required for AI—which the World Economic Forum estimates will be 50% a year through 2030—is unsustainable.

“We need to find technical solutions, policy solutions and other solutions that solve this collectively,” he says.

To that end, Wee’s team at Cloudflare is testing a radical new kind of microchip, from a startup founded in 2023, called Positron, which has just announced a fresh round of $51.6 million in investment. These chips have the potential to be much more energy efficient than ones from industry leader Nvidia at the all-important task of inference, which is the process by which AI responses are generated from user prompts. While Nvidia chips will continue to be used to train AI for the foreseeable future, more efficient inference could collectively save companies tens of billions of dollars, and a commensurate amount of energy.

There are at least a dozen chip startups all battling to sell cloud-computing providers the custom-built inference chips of the future. Then there are the well-funded, multiyear efforts by Google, Amazon and Microsoft to build inference-focused chips to power their own internal AI tools, and to sell to others through their cloud services.

The intensity of these efforts, and the scale of the cumulative investment in them, show just how desperate every tech giant—along with many startups—is to provide AI to consumers and businesses without paying the “Nvidia tax.” That’s Nvidia’s approximately 60% gross margin, the price of buying the company’s hardware.

Nvidia is very aware of the growing importance of inference and concerns about AI’s appetite for energy, says Dion Harris, a senior director at Nvidia who sells the company’s biggest customers on the promise of its latest AI hardware. Nvidia’s latest Blackwell systems are between 25 and 30 times as efficient at inference, per watt of energy pumped into them, as the previous generation, he adds.

Getting specialized

To accomplish their goals, makers of novel AI chips are using a strategy that has worked time and again: They are redesigning their chips, from the ground up, expressly for the new class of tasks that is suddenly so important in computing. In the past, that was graphics, and that’s how Nvidia built its fortune. Only later did it become apparent graphics chips could be repurposed for AI, but arguably it’s never been a perfect fit.

Jonathan Ross is chief executive of chip startup Groq, and previously headed Google’s AI chip development program. He says he founded Groq (no relation to Elon Musk’s xAI chatbot) because he believed there was a fundamentally different way of designing chips—solely to run today’s AI models.

Groq claims its chips can deliver AI much faster than Nvidia’s best chips, and for between one-third and one-sixth as much power as Nvidia’s. This is due to their unique design, which has memory embedded in them, rather than being separate. While the specifics of how Groq’s chips perform depends on any number of factors, the company’s claim that it can deliver inference at a lower cost than is possible with Nvidia’s systems is credible, says Jordan Nanos, an analyst at SemiAnalysis who spent a decade working for Hewlett Packard Enterprise.

Positron is taking a different approach to delivering inference more quickly. The company, which has already delivered chips to customers including Cloudflare, has created a simplified chip with a narrower range of abilities, in order to perform those tasks more quickly. The company’s latest funding round came from Valor Equity Partners, Atreides Management and DFJ Growth, and brings the total amount of investment in the company to $75 million.

Positron’s next-generation system will compete with Nvidia’s next-generation system, known as Vera Rubin. Based on Nvidia’s road map, Positron’s chips will have two to three times better performance per dollar, and three to six times better performance per unit of electricity pumped into them, says Positron CEO Mitesh Agrawal.

Competitors’ claims about beating Nvidia at inference often don’t reflect all of the things customers take into account when choosing hardware, says Harris. Flexibility matters, and what companies do with their AI chips can change as new models and use cases become popular. Nvidia’s customers “are not necessarily persuaded by the more niche applications of inference,” he adds.

Cloudflare’s initial tests of Positron’s chips were encouraging enough to convince Wee to put them into the company’s data centers for more long-term tests, which are continuing. It’s something that only one other chip startup’s hardware has warranted, he says. “If they do deliver the advertised metrics, we will open the spigot and allow them to deploy in much larger numbers globally,” he adds.

By commoditizing AI hardware, and allowing Nvidia’s customers to switch to more-efficient systems, the forces of competition might bend the curve of future AI power demand, says Wee. “There is so much FOMO right now, but eventually, I think reason will catch up with reality,” he says.

One truism of the history of computing is that whenever hardware engineers figure out how to do something faster or more efficiently, coders—and consumers—figure out how to use all of the new performance gains, and then some.

Mark Lohmeyer is vice president of AI and computing infrastructure for Google Cloud, where he provides both Google’s own custom AI chips, and Nvidia’s, to Google and its cloud customers. He says that consumer and business adoption of new, more demanding AI models means that no matter how much more efficiently his team can deliver AI, there is no end in sight to growth in demand for it. Like nearly all other big AI providers, Google is making efforts to find radical new ways to produce energy to feed that AI—including both nuclear power and fusion.

The bottom line: While new chips might help individual companies deliver AI more efficiently, the industry as a whole remains on track to consume ever more energy. As a recent report from Anthropic notes, that means energy production, not data centers and chips, could be the real bottleneck for future development of AI.

The task force determines which preventive services insurers must cover at no cost to patients

Health Secretary Kennedy plans to remove all members of the U.S. Preventive Services Task Force.

Kennedy considers the panel too “woke,” following White House targeting of DEI initiatives in different sectors, sources say.

The plan follows a Supreme Court case and separate criticism that the task force embedded left-wing ideology.

Health and Human Services Secretary Robert F. Kennedy Jr. is planning to remove all the members of an advisory panel that determines what cancer screenings and other preventive health measures insurers must cover, people familiar with the matter said.

Kennedy plans to dismiss all 16 panel members of the U.S. Preventive Services Task Force because he views them as too “woke,” the people said.

The White House has made a priority of targeting initiatives that promote diversity equity and inclusion, or DEI, in everything from artificial intelligence to health research grants.

The task force has advised the federal government on preventive health matters since 1984. The Affordable Care Act in 2010 gave it the power to determine which screenings, counseling and preventive medications most insurers are required to cover at no cost to patients. The group, made up of volunteers with medical expertise who are vetted for conflicts of interest, combs through scientific evidence to determine which interventions are proven to work.

The Supreme Court decided a case in June that originally focused on a task-force recommendation to cover certain HIV-prevention drugs. The employer plaintiffs had successfully argued previously that requiring them to cover such drugs for employees violated their religious rights. The group also argued that the task-force members weren’t properly appointed. The high court ruled that the task- force appointments were constitutional, while highlighting that the Health and Human Services Secretary has the authority to remove the members of the panel at will.

A recent essay in The American Conservative magazine called for the removal of the task-force members, arguing that it had embedded “left-wing ideological orthodoxy” in all of its efforts. The essay pointed to the task force highlighting racial discrimination when discussing risk factors for anxiety in older children and teenagers, as well as the task force’s use of terms such as “pregnant persons.” The task force mentioned the “lasting psychological impact and stigma of enslaved Black women being forced to act as wet nurses” in an April publication on breast-feeding, the essay noted.

HHS didn’t offer details on the secretary’s plans for, or views on, the task force.

“The Secretary looks forward to working with the USPSTF to improve public health,” a spokeswoman for the secretary said, referring to the task force.

The secretary’s office this month abruptly postponed the task force’s July meeting, alarming some Democrats and public health leaders.

“In no world should experts be replaced with unqualified anti-science cronies of RFK Jr. who will make preventive healthcare more expensive and harder to get over baseless conspiracy theories or debunked disinformation,” said Sen. Patty Murray (D., Wash.), who sits on the Senate’s health committee.

Kennedy, a longtime vaccine skeptic, in June removed all the members of a separate advisory committee on immunizations. New members appointed by Kennedy pushed forward several of the secretary’s priorities later that month, announcing a new examination of the full schedule of vaccines children receive and recommending flu shots that don’t contain the preservative thimerosal, often used in multi-dose flu-vaccine vials.

Kennedy raised questions on whether Human Immunodeficiency Virus is the sole cause of AIDS in a 2021 book, saying “I take no side in this dispute.” Scientists have for decades considered it a proven fact that HIV causes AIDS.