Feed aggregator

Check out CNET experts' recommendations for internet service providers in Connecticut.

An FAA official said the hotline, which connects air traffic control at Washington's DCA with the Pentagon, hasn't worked since 2022. An Army helicopter collided with a passenger jet there in January.

(Image credit: Alex Brandon)

• Adata T7P5 SSD brings 15.36TB and blistering Gen 5 speed to enterprise storage
• Trusta isn’t just fast, it’s built for AI, virtualization, and high-efficiency data environments
• With 13,500MB/s read speeds, the T7P5 crushes most consumer and prosumer storage options

With enterprise demand for AI servers and high-performance storage infrastructure booming, Adata is making a bold move into the data center and AI markets with the launch of its new enterprise brand, Trusta.

Revealed ahead of Computex 2025, Trusta promises to deliver advanced PCIe Gen 5 SSDs in massive capacities, blurring the line between performance and practicality.

Trusta’s flagship model, the T7P5 SSD, leads the new T7 Series and is built to handle demanding workloads such as AI training, vector databases, and virtual desktops.

T7P5 SSD delivers extreme speeds

This SSD offers blazing-fast read and write speeds of up to 13,500 MB/s and 10,400 MB/s, respectively, with capacities ranging from 1.92TB up to a staggering 15.36TB, making it one of the fastest enterprise SSDs introduced to date.

Unlike consumer models, the T7P5 is built in enterprise form factors like U.2, E1.S, and E3.S, ensuring compatibility with server and cloud hardware. However, it’s a large SSD, and it won’t fit inside a typical business desktop.

For enterprises with less intensive needs, Adata also offers the T7P4 PCIe Gen 4 SSD, which delivers up to 7,400 MB/s read and 5,050 MB/s write speeds, in capacities up to 7.68TB.

The entry-level T5 Series, which includes the T5P4B, T5S3B, and T5S3, supports both PCIe Gen 4 and SATA III interfaces. These drives are targeted at system boot operations and applications requiring data reliability over raw speed.

Still, performance isn’t just about headline numbers. Trusta integrates Flexible Data Placement (FDP) technology to optimize data flow, particularly under high-load conditions where latency and efficiency matter most.

For enterprise IT buyers and planners, Trusta’s lineup offers a compelling look at next-gen storage. But for average consumers in search of the best external SSD or a high-capacity Gen 5 upgrade, these drives are out of reach, both in terms of form factor and intended use case.

Via TechPowerUp

You might also like

• We've listed the best portable SSDs right now
• Another 122.88TB SSD just launched from an obscure Chinese startup
• Understanding the differences between enterprise and client SSDs

Badar Khan Suri is the latest scholar targeted by the Trump administration to be released from detention.

(Image credit: Andrew C. Condon)

Nearly half a million people in Gaza now face starvation, according to a new report from the IPC, the international panel of famine experts who advise the United Nations.

For more than ten weeks, Israel has halted the flow of humanitarian aid into Gaza, to pressure Hamas to release Israeli hostages. Israel accuses Hamas of seizing aid, selling it on the black market and using aid distribution to reinforce its control of Gaza.

The UN says hundreds of truckloads of lifesaving supplies are waiting at the border. Meanwhile, inside Gaza, food is scarce. Humanitarian groups like the UN World Food Programme (WFP) exhausted supplies of basic staples weeks ago, forcing them to shut down their kitchens and bakeries, and everyday Palestinians are grinding up pasta and lentils to make flour for bread. Antoine Renard of the WFP says when he was in Gaza last week, wheat flour was selling for $10 a pound.

Juana Summers talks with Renard about what he's seen in Gaza, and what's next for the people there.

For sponsor-free episodes of Consider This, sign up for Consider This+ via Apple Podcasts or at plus.npr.org.

Email us at considerthis@npr.org.

(Image credit: Moiz Salhi)

House Republicans are trying to pass a massive bill full of President Trump's top agenda items by the end of next week. But a new set of hurdles awaits in the Senate.

(Image credit: ANDREW CABALLERO-REYNOLDS/AFP via Getty Images)

Federal rules to reduce the levels of "forever chemicals" in drinking water are getting delayed.

(Image credit: Catherine Falls Commercial/Moment RF)

In a rare break with the president, some inside the GOP are expressing concern over the potential national security risks of accepting a luxury Boeing 747 jet from Qatar to replace Air Force One.

(Image credit: Brendan Smialowski)

It's the second major of the year as the world's top players head to the Quail Hollow Club.

Audiobooks have saved my sanity on many long commutes and have been great company while I'm cleaning or doing other chores. When the performance is good, it's easy to fall into the story. Audible wants authors and their readers to embrace AI as an alternative to human narration, but I am skeptical. Audible is offering publishers access to a fully integrated AI production pipeline. That includes auto-generating entire audiobooks with synthetic voices.

Their pitch is appealing on the surface: there are millions of books out there, and only a sliver of them ever make it into audio. Making audiobooks is expensive, time-consuming, and involves real people who need to be paid fairly for their time. An AI narrator is faster, cheaper, and a lot of people might not even notice it's not a human performing.

But "good enough" shouldn't be the standard for art, and audiobooks are very much an art form. Great narration adds depth, color, rhythm, and even new meaning to a text. It transforms reading aloud from words on a page you can hear to a real performance. Even if AI gets close in a technical sense, and I've heard AI audio that matches a human performance for at least a few minutes, we’ll still know the difference.

Human narration has nuance because it has context. The narrator understands not just the definition of the words they're saying, but the emotion and history behind them. They know the difference between a sigh of relief and a sigh of resignation. AI can approximate those sounds, sometimes amazingly so, but it's like a pet trick. A dog can cover its eyes, but that's not actually the dog feeling embarrassed.

The more AI voices fill our earbuds, the more we risk turning one of the most intimate forms of storytelling into something that feels robotic, flat, and eerily lifeless. It’s like auto-tuning a lullaby. It might hit the right notes, but it doesn’t sing.

AI narration needs

All of that said, I'm not against using AI for audiobooks in the right setting. Like any technology, it's about how AI narration is deployed, not whether it exists. There are so many books and new ones emerging all the time. If you’re an independent author with no budget to hire a narrator, or a publisher with a shelf of titles no one has touched in a decade, AI narration could breathe life into your books.

Synthetic voices don’t replace anything in those contexts; they just provide access. And an AI voice could supplement human readers with a multi-voice performance if you use the self-service version of Audible's AI narration platform. Using AI to supplement rather than replace all human voices feels like a better option to me.

One area I'm all in on for AI voices is translating texts. Audible has a beta test for AI-powered translation tools that could bring books to people unable to understand them in their original language. If there’s anything worse than a great book not having an audiobook, it’s a great book not being accessible in your language. Audible is starting the program by offering to translate English books into Spanish, French, German, and Italian.

The translation service can simply translate text and then give the new work an AI narrator, but what is more interesting to me is the speech-to-speech mode. That means an audiobook performed by a human in English could be replicated in a different language while sounding like the original performer.

The narrator of a bestselling English audiobook could now “speak” fluent Spanish in their own voice, introducing that story to new listeners around the world. That’s my favorite way to think about how to use AI. It can expand the reach of art without diluting its heart.

It's not quite the same as original, human narration, but it's a solution to a problem. That's how Audible should pitch AI audiobooks. We should absolutely use AI narration to make books accessible. But if it's possible to give it a human touch, that should be the first thought.

It's important not to lose sight of how this AI audiobook shift affects the performers who often build careers lending their voices to other people’s stories. If AI starts gobbling up midlist titles, budget-conscious publishers might see no reason to hire real readers anymore. AI doesn’t have to be the enemy. But it shouldn’t be the default.

You might also like

• Your next audiobook's big twist might be that the narrator... is a (voice) clone!
• The AI That Cried AAAAAAHHH!
• How you can get (AI versions of) Judy Garland or Burt Reynolds to read to you
• This app can add AI narration to any site or text – here's how to make it work

Republicans want to use the federal tax code to create a national school voucher even in states where voters have fought such efforts.

(Image credit: Willie J. Allen Jr.)

PlayStation Plus subscribers will also be able to explore the Chornobyl Exclusion Zone in Stalker and play some other great games, too.

• AI data centers overwhelm air cooling with rising power and heat
• Liquid cooling is becoming essential as server density surges with AI growth
• New hybrid cooling cuts power and water but faces adoption hesitance

As AI transforms everything from search engines to logistics, its hidden costs are becoming harder and harder to ignore, especially in the data center. The power needed to run generative AI is pushing infrastructure beyond what traditional air cooling can handle.

To explore the scale of the challenge, I spoke with Daren Shumate, founder of Shumate Engineering, and Stephen Spinazzola, the firm’s Director of Mission Critical Services.

With decades of experience building major data centers, they’re now focused on solving AI’s energy and cooling demands. From failing air systems to the promise of new hybrid cooling, they explained why AI is forcing data centers into a new era.

What are the biggest challenges in cooling a data center?

Stephen Spinazzola: The biggest challenges in cooling data centers are power, water and space. With high-density computing, like the data centers that run artificial intelligence, comes immense heat that cannot be cooled with a conventional air-cooling system.

The typical cabinet loads have doubled and tripled with the deployment of AI. An air-cooling system simply cannot capture the heat generated by the high KW/ cabinet loads generated by AI cabinet clusters.

We have performed computational fluid dynamic (CFD) on numerous data center halls and an air-cooling system shows high temperatures above acceptable levels. The air flows we map with CFD show temperature levels above 115 degrees F. This can result in servers shutting down.

Water cooling can be done in a smaller space with less power, but it requires enormous amount of water. A recent study determined that a single hyper-scaled facility would need 1.5 million liters of water per day to provide cooling and humidification.

These limitations pose great challenges to engineers while planning the new generation of data centers that can support the unprecedented demand we’re seeing for AI.

How is AI changing the norm when it comes to data center heat dissipation?

Stephen Spinazzola: With CFS modeling showing potential servers shutting down with conventional air-cooling within AI cabinet clusters, the need for direct liquid cooling (DLC) is required. AI is typically deployed in 20-30 cabinet clusters at or above 40 KW per cabinet. This represents a fourfold increase in KW/ cabinet with the deployment of AI. The difference is staggering.

A typical Chat-GPT query uses about 10 times more energy than a Google search – and that’s just for a basic generative AI function. More advanced queries require substantially more power that have to go through an AI Cluster Farm to process large-scale computing between multiple machines.

It changes the way we think about power. Consequently, the energy demands are shifting the industry to utilize more liquid-cooling techniques than traditional air cooling.

We talk a lot about cooling, what about delivering actual power?

Daren Shumate: There are two overarching new challenges to deliver power to AI computing: how to move power from UPS output boards to high-density racks, and how to creatively deliver high densities of UPS power from utility.

Power to racks is still accomplished with either branch circuits from distribution PDUs to rack PDUs (plug strips) or with plug-in busway over the racks with the in-rack PDUs plugging into the busway at each rack. The nuance now is what ampacity of busway makes sense with the striping and what is commercially available.

Even with plug-in busway available at an ampacity of 1,200 A, the density of power is forcing the deployment of a larger quantity of separate busway circuits to meet density and the striping requirements. Further complicating power distribution are specific and varying requirement of individual data center end users from branch circuit monitoring or preferences of distribution.

Depending upon site constraints, data center cooling designs can feature medium voltage UPS. Driven by voltage drop concerns, the MV UPS solves concerns for the need to have very large feeder duct banks but also introduces new medium voltage/utilization voltage substations into the program. And when considering medium voltage UPS, another consideration is the applicability of MV rotary UPS systems vs. static MV solutions.

What are the advantages/disadvantages of the various cooling techniques?

Stephen Spinazzola: There are two types of DLC in the market today. Emersion Cooling and cold plate. Emersion Cooling uses large tanks of a non-conducing fluid with the servers positioned vertically and fully emersed in the liquid.

The heat generated by the servers is transferred to the fluid and then transferred to the buildings chilled water system with a closed loop heat exchanger. Emersion tanks take up less space but require servers that are configured for this type of cooling.

Cold-plated cooling uses a heat sink attached to the bottom of the chip stack that transfers the energy from the chip stack to a fluid that is piped throughout the cabinet. The fluid is then piped to an end of row Cooling Distribution Unit (CDU) that transfers the energy to the building chilled water system.

The CDU contains a heat exchanger to transfer energy and 2N pumps on the secondary side of the heat exchanger to ensure continuous fluid flow to the servers. Cold plate cooling is effective at server cooling but it requires a huge amount of fluid pipe connecters that must have disconnect leak stop technology.

Air cooling is proven technique for cooling data centers, which has been around for decades; however, it is inefficient for the high-density racks that are needed to cool AI data centers. As the loads increase, it becomes harder to failure-proof it using CFD modeling.

You're presenting a different cooler, how does it work and what are the current challenges to adoption?

Stephen Spinazzola: Our patent pending Hybrid-Dry/AdiabaticCooling (HDAC) design solution uniquely provides two temperatures of cooling fluid from a single closed loop, allowing for a higher temperature fluid to cool DLC servers and a lower temperature fluid for conventional air cooling.

Because HDAC simultaneously uses 90 percent less water than a chiller-cooling tower system and 50 percent less energy than an air-cooled chiller system, we’ve managed to get the all-important Power Usage Effectiveness (PUE) figure all the way down to about 1.1 annualized for the type of hyperscale data center that is needed to process AI. Typical AI data centers produce a PUE ranging from 1.2 to 1.4.

With the lower PUE, HDAC provides an approximate 12% more usable IT power from the same size utility power sized feed. Both economic and environmental benefits are significant. With a system that provides both an economic and environmental benefit, HDAC requires only “a sip of water”.

The challenge to adoption is simple: nobody wants to go first.

You might also like

• Cooling high-density data centers with coolant distribution units
• New fanless cooling technology enhances energy efficiency for AI workloads
• Liquid cooling gains momentum as AI workloads turn up the heat in data centers

The National Institute for Occupational Safety and Health is restoring several programs and bringing back the staffers who run them, but much of the agency's work is still on the chopping block.

(Image credit: Gene J. Puskar)

Here are some hints and the answers for the NYT Connections: Sports Edition puzzle, No. 234, for May 15.

• Hygon’s C86-5G breaks free from AMD Zen, unleashing 128 cores of homegrown muscle
• SMT4 powers each core to run four threads, stacking up to 512 threads total
• AVX-512 instructions make it a strong fit for AI, analytics, and scientific computing

Hygon, a key player in China’s semiconductor industry, is advancing its server processor lineup with the upcoming C86-5G, a flagship, high-performance CPU featuring 128 cores and 512 threads, positioning it to compete directly with AMD’s EPYC and Intel’s Xeon platforms.

According to TechPowerUp, this marks Hygon’s complete break from AMD’s Zen architecture and the introduction of its first fully homegrown design, the result of five years of domestic R&D in CPU development.

The new lineup is made possible through four-way simultaneous multithreading (SMT4), allowing each core to handle four threads.

Built for parallel workloads and high throughput

While SMT4 is not a new concept - it has appeared in processors like Intel’s Xeon Phi and IBM’s Power8 - its use in a modern, domestically developed Chinese processor is a notable milestone.

The 128-core configuration in the C86-5G represents a major leap from its predecessor, the C86-4G, which had 64 cores and 128 threads using traditional SMT2.

Designed for enterprise and server workloads, the C86-5G features 16 channels of DDR5-5600 memory, potentially supporting up to 1TB using 64GB DDR5 modules. This is a step up from the previous model’s 12 channels of DDR5-4800.

On the connectivity front, while Hygon has not yet disclosed the exact PCIe 5.0 lane count, it has confirmed support for Compute Express Link 2.0 (CXL 2.0), aligning the chip with industry standards used by AMD’s EPYC 9005 (Turin) and Intel’s 5th Gen Xeon (Emerald Rapids). The earlier C86-4G already offered 128 lanes of PCIe 5.0, so similar or better support is expected.

Although the specific microarchitecture has not been detailed, Hygon states it is based on an "enhanced self-developed microarchitecture" that follows the Zen-based Dhyana design of the first generation.

According to the company, the architecture delivers a 17% improvement in instructions per cycle (IPC), though this remains unverified in the absence of benchmark testing.

The chip also supports AVX-512 instructions for high-performance computing tasks and is built to handle physical stress in demanding environments. It is expected to support standard server memory modules like RDIMMs and is intended for large-scale data center deployments.

While Hygon still trails AMD and Intel in overall performance, the C86-5G's technical specifications, including I/O capabilities, memory bandwidth, threading, and core count, place it in a competitive position.

Although there is no official launch date yet, development is likely well underway, given that the C86-4G has been on the market since 2024.

You might also like

• These are the fastest SSDs you buy right now
• Take a look at some of the best external hard drives
• Popular employee monitoring software hijacked to launch ransomware attacks

The nation's aviation infrastructure is again under scrutiny, following a series of paralyzing communications and radar outages at some of the country's busiest airports. Here is a look at the scientific origins of radar.

(Image credit: JULIEN DE ROSA/AFP via Getty Images)

Here are hints and answers for the NYT Strands puzzle No. 438 for May 15.

Here are some hints and the answers for Connections for May 15, #704.

Here are hints and the answer for today's Wordle No. 1,426 for May 15.