Home Blog Page 2

TSMC targets AI acceleration with A14 course of and ‘System on Wafer-X’



Nvidia’s flagship GPUs presently combine two chips, whereas its forthcoming Rubin Extremely platform will join 4.

“The SoW-X delivers wafer-scale compute efficiency and considerably boosts velocity by integrating a number of superior compute SoC dies, stacked HBM reminiscence, and optical interconnects right into a single bundle,” mentioned Neil Shah, accomplice and co-founder at Counterpoint Analysis. “This method reduces latency, improves energy effectivity, and enhances scalability in comparison with conventional multi-chip setups — giving enterprises and hyperscalers AI servers able to dealing with future workloads sooner, extra effectively, and in a smaller footprint.”

This not solely boosts capex financial savings in the long term but in addition opex financial savings by way of power and house.

“Wafer-X expertise isn’t nearly larger chips — it’s a sign that the way forward for AI infrastructure is being redesigned on the silicon degree,” mentioned Abhivyakti Sengar, apply director at Everest Group. “By tightly integrating compute, reminiscence, and optical interconnects inside a single wafer-scale bundle, TSMC targets the core constraints of AI: bandwidth and power. For hyperscale information facilities and frontier mannequin coaching, this may very well be a game-changer.”

Priorities for enterprise prospects

For enterprises investing in customized AI silicon, selecting the best foundry accomplice goes past efficiency benchmarks. It’s about discovering a stability between cutting-edge capabilities, flexibility, and price.

“First, enterprise consumers must assess manufacturing course of applied sciences (reminiscent of TSMC’s 3nm, 2nm, or Intel’s 18A) to find out in the event that they meet AI chip efficiency and energy necessities, together with customization capabilities,” mentioned Galen Zeng, senior analysis supervisor for semiconductor analysis at IDC Asia Pacific. “Second, consumers ought to consider superior packaging skills; TSMC leads in 3D packaging and customised packaging options, appropriate for extremely built-in AI chips, whereas Intel has benefits in x86 structure. Lastly, consumers ought to assess pricing constructions.”

Prioritize media privateness with Android Picture Picker and construct consumer belief



Prioritize media privateness with Android Picture Picker and construct consumer belief

Posted by Tatiana van Maaren – World T&S Partnerships Lead, Privateness & Safety, and Roxanna Aliabadi Walker – Product Supervisor

At Google Play, we’re devoted to constructing consumer belief, particularly in relation to delicate permissions and your information. We perceive that managing recordsdata and media permissions could be complicated, and customers usually fear about which recordsdata apps can entry. Since these recordsdata usually comprise delicate data like household images or monetary paperwork, it is essential that customers really feel in management. That’s why we’re working to supply clearer selections, so customers can confidently grant permissions with out sacrificing app performance or their privateness.

Under are a set of finest practices to think about for enhancing consumer belief within the sharing of broad entry recordsdata, in the end resulting in a extra profitable and sustainable app ecosystem.

Prioritize consumer privateness with information minimization

Constructing consumer belief begins with requesting solely the permissions important on your app’s core capabilities. We perceive that images and movies are delicate information, and broad entry will increase safety dangers. That is why Google Play now restricts READ_MEDIA_IMAGES and READ_MEDIA_VIDEO permissions, permitting builders to request them solely when completely essential, sometimes for apps like photograph/video managers and galleries.

Leverage privacy-friendly options

As a substitute of requesting broad storage entry, we encourage builders to make use of the Android Picture Picker, launched in Android 13. This instrument affords a privacy-centric manner for customers to pick particular media recordsdata with out granting entry to their total library. Android photograph picker offers an intuitive interface, together with entry to cloud-backed images and movies, and permits for personalisation to suit your app’s wants. As well as, this technique picker is backported to Android 4.4, guaranteeing a constant expertise for all customers. By eliminating runtime permissions, Android photograph picker simplifies the consumer expertise and builds belief via transparency.

Construct belief via clear information practices

We perceive that some builders have traditionally used customized photograph pickers for tailor-made consumer experiences. Nonetheless, no matter whether or not you employ a customized or system picker, transparency with customers is essential. Customers wish to know why your app wants entry to their images and movies.

Builders ought to try to supply clear and concise explanations inside their apps, ideally on the level the place the permission is requested. Take the next in consideration whereas crafting your permission request mechanisms as attainable finest practices pointers:

    • When requesting media entry, present clear explanations inside your app. Particularly, inform customers which media your app wants (e.g., all images, profile photos, sharing movies) and clarify the performance that depends on it (e.g., ‘To decide on a profile image,’ ‘To share movies with mates’).
    • Clearly define how consumer information might be used and guarded in your privateness insurance policies. Clarify whether or not information is saved regionally, transmitted to a server, or shared with third events. Reassure customers that their information might be dealt with responsibly and securely.

Find out how Snap has embraced the Android System Picker to prioritize consumer privateness and streamline their media choice expertise. Here is what they need to say about their implementation:

A grid of photos in the photo library is shown on a smartphone screen, including a waterfall and two people smiling and posing for the camera. The Google Photos interface is at the top, with the Photos tab selected, and one photo from the grid is selected for use

“One in all our objectives is to supply a seamless and intuitive communication expertise whereas guaranteeing Snapchatters have management over their content material. The brand new stream of the Android Picture Picker is the proper stability of offering consumer management of the content material they wish to share whereas guaranteeing quick communication with mates on Snapchat.”

Marc Brown, Product Supervisor

Get began

Begin constructing a extra reliable app expertise. Discover the Android Picture Picker and implement privacy-first information practices as we speak.

Acknowledgement

Particular due to: Could Smith – Product Supervisor, and Anita Issagholyan – Senior Coverage Specialist

android – After Capacitor 6 improve and eradicating media permissions, app behaves as if consumer is logged out (however they are not)


A couple of weeks in the past, I upgraded my Ionic/Angular app from Capacitor 5 to Capacitor 6 and adopted Google’s new Play Retailer privateness insurance policies.

Within the earlier model of the app (final up to date in June 2023), we used these Android permissions:

  • READ_MEDIA_IMAGES.
  • READ_MEDIA_VIDEO.
  • READ_EXTERNAL_STORAGE.
  • WRITE_EXTERNAL_STORAGE.

Within the new model, I eliminated all of these permissions and changed them with the Photograph Picker API as required by Google.
For older Android variations, I ensured compatibility by integrating the photograph picker by way of the manifest (Google providers).

App context:

  • We use @ionic/storage to retailer a big appData object domestically.

  • This enables the app to work offline, displaying cached content material when there’s no web.

  • When linked, the app syncs with the backend and updates the native knowledge.

  • Customers have to be registered in our backend system and logged in to make use of the app usually.

  • Consumer tokens additionally saved within the native appData.

The issue after the replace:

After publishing the brand new model (with Capacitor 6 + permission modifications), many customers (largely Android) reported this difficulty:

  • The app opens wonderful, the UI masses, and all buttons work.

  • Nonetheless, the app behaves as if the consumer is logged out, regardless that they’re logged in.

  • The app fails to fetch new knowledge, doesn’t sync with the backend, and solely exhibits outdated or no content material.

  • On iOS, just a few customers had comparable issues (And few customers simply get white display screen), however far fewer than on Android.

Uninstalling and reinstalling the app fixes the difficulty.

It looks as if the app is failing to load legitimate session/token knowledge from native storage, regardless that the login was accomplished earlier than.

What I believe:

  • Some customers might have corrupted or inaccessible native storage after the replace.

  • Probably brought on by modifications in Capacitor 6’s native storage dealing with.

  • Or perhaps the removing of previous storage-related permissions broke one thing on sure gadgets.

What I’ve already accomplished:

My questions:

  1. What may trigger @ionic/storage to fail or return stale/undefined knowledge after upgrading to Capacitor 6?

  2. Might eradicating permissions like READ_MEDIA_* and switching to Photograph Picker not directly have an effect on storage accessibility or trigger corruption?

  3. Has anybody else skilled comparable points with persistent storage after latest Capacitor upgrades?

Any assist or route can be vastly appreciated!

AI and the Way forward for Translation: A New Period of Human-AI Collaboration

0


Synthetic intelligence is remodeling industries at an unprecedented tempo, and the world of translation isn’t any exception. As AI-driven language fashions develop extra subtle, one query continues to floor: Will AI change human translators? At RWS, we imagine the reply is obvious—AI won’t ever change human experience, however it should essentially change how people and AI collaborate.

This perception is rooted in what we name Real Intelligence—the concept true intelligence is not only synthetic, however a mixture of machine effectivity and human experience. AI alone can not perceive nuance, cultural context, or emotion. It may possibly course of language, however it can not really comprehend which means. That’s why the way forward for translation isn’t about AI changing individuals—it’s about AI and other people working collectively in smarter, extra impactful methods.

A Hybrid Strategy: Machine-First, Human-Optimized

Moderately than seeing AI as a competitor, we see it as an enabler—one which enhances productiveness, improves accuracy, and expands the capabilities of language specialists. AI excels at dealing with repetitive, time-consuming duties comparable to pre-translating content material, matching terminology, and analyzing linguistic patterns at scale. Nevertheless, true translation goes far past direct word-for-word substitute. It requires cultural fluency, contextual understanding, and an emotional connection—parts that solely human experience can present.

At RWS, we embrace a “machine-first, human-optimized” strategy, the place AI streamlines workflows whereas language specialists refine high quality, fluency, and cultural nuance. This technique isn’t about automation for automation’s sake. It’s about utilizing AI to release human translators and language specialists to deal with probably the most significant facets of their work—including creativity, important pondering, and strategic perception.

Past Textual content: AI’s Position in Multimedia Localization and Creation

AI isn’t simply altering written translation; it’s reshaping how multimedia content material is produced, localized, and consumed worldwide. In response to our latest research titled “Unlocked 2025: Driving the AI Shockwave,” 70% of world customers report seeing extra AI-generated multimedia content material—movies, photographs, and audio—for the reason that launch of instruments like ChatGPT. This shift has main implications for translation and localization.

As well as, generative AI is quickly being adopted in industries comparable to movie, music, and promoting, significantly in fast-growing digital markets like Sub-Saharan Africa, the place streaming is driving demand. AI-powered instruments are serving to manufacturers scale content material creation whereas sustaining linguistic and cultural relevance. Customers now affiliate main Gen AI instruments like ChatGPT, Gemini by Google, and Microsoft’s CoPilot with enhanced artistic capabilities, whereas rising gamers from France, the UAE, and China are bringing contemporary competitors to AI-generated media.

As this digital content material consumption grows, customers more and more anticipate world manufacturers to supply seamless, localized multimedia experiences. AI-powered speech recognition, artificial voices, and automatic subtitling are actually key to creating video content material accessible throughout languages. The demand for dubbing and subtitling has by no means been larger, significantly in linguistically numerous areas like APAC and Africa, the place customers anticipate manufacturers to talk their language—actually and figuratively.

However localization goes past translation. It’s about making content material really feel native to every viewers. Localized imagery, for instance, performs a important position in establishing authenticity. Many markets, particularly within the International South, desire culturally aligned visuals and narratives in promoting and company communications. AI will help automate this course of, however human oversight stays important to make sure content material is not only translated however really localized.

Generative AI will not be solely remodeling enterprise workflows but additionally fueling a artistic renaissance in rising markets. In Nigeria and India, AI-powered instruments are enabling filmmakers, musicians, and content material creators to scale their attain globally. Streaming platforms are leveraging AI to automate enhancing, optimize translations, and produce regionally related content material, making multimedia extra accessible to numerous audiences.

At RWS, our Evolve linguistic AI answer is revolutionizing multimedia localization. By integrating translation administration (Trados Enterprise), neural machine translation (Language Weaver), and AI-assisted high quality estimation (MTQE), we allow language specialists to refine content material effectively making certain fluency, accuracy, and cultural alignment.

Client Perceptions and Challenges

Regardless of AI’s developments in multimedia localization, customers stay cautious. Whereas Unlocked 2025 discovered that 57% of respondents have observed enhancements in AI-generated multimedia high quality, issues persist round accuracy, cultural relevance, and misinformation. Belief in AI-generated content material is especially low in regulated industries comparable to healthcare and finance, the place errors in translated supplies can have severe penalties.

Transparency can be a rising concern. In response to the report, 81% of customers need AI-generated content material to be clearly labeled, underscoring the necessity for better disclosure in AI-powered multimedia. Moreover, 56% of respondents report an increase in faux multimedia content material, together with deepfakes and manipulated visuals, elevating moral questions on AI’s position in info integrity.

The Way forward for Multimedia Localization with AI

Wanting forward, AI’s affect on multimedia will proceed to evolve, driving new alternatives for immersive, personalised content material experiences. AI is already enabling developments in interactive movies, AR/VR functions, and dynamic ads tailor-made to particular person consumer preferences. Initiatives like Mozilla’s Widespread Voice venture are additionally increasing voice AI capabilities, serving to to generate high-quality voiceovers in underserved languages.

However right here’s what is going to set profitable manufacturers aside: discovering the fitting steadiness between automation and human experience. Hybrid human-AI approaches—the place AI accelerates workflows and people present cultural and artistic oversight—will likely be key to sustaining authenticity, belief, and engagement in multilingual content material.

Ultimate Ideas: The Position of Real Intelligence

The way forward for translation and localization isn’t about AI changing people—it’s about utilizing AI intelligently to reinforce human experience. That is the essence of Real Intelligence: a collaborative strategy the place AI accelerates workflows, and human specialists guarantee accuracy, cultural authenticity, and emotional connection.

Generative AI is unlocking new potentialities for content material creation and localization. Nevertheless, long-term success will depend upon balancing automation with human oversight to construct belief, transparency, and engagement in multilingual content material.

Finally, probably the most impactful manufacturers received’t simply undertake AI—they’ll combine it thoughtfully, utilizing expertise to scale whereas making certain content material stays culturally resonant. However to actually join with numerous audiences, human contribution is crucial. Not simply any human enter—however the nuanced experience of as we speak’s language specialists: professionals who mix area data, linguistic fluency, cultural sensitivity, technical ability, and artistic intuition. It’s this mixture of capabilities that ensures AI-generated content material isn’t simply quick and useful, but additionally fluent, related, and emotionally clever. AI might energy the method—however it’s human specialists who give content material its which means.

Disaster Response Neighborhood volunteers know what it takes to remain related


April is International Volunteer Month. All month lengthy, we’re recognizing among the methods Cisco staff give their time, abilities, and experience of their communities.


In 2005, Hurricane Katrina fashioned over the waters of the Gulf of Mexico and went on to turn out to be a Class 5 storm that will devastate town of New Orleans. In response to the catastrophe, Cisco deployed what was then its Tactical Operations, or TacOps, group to assist reestablish connectivity for first responders and victims.

Now, 20 years later, what was then TacOps has advanced to turn out to be Cisco Disaster Response, a company perform with a mission of offering connectivity in response to disaster, backed by the energy of greater than 800 worker volunteers who comprise the CCR Neighborhood.

CCR Ops Supervisor, Kevin Murphy, exhibits volunteers how one can join a Cisco Meraki-based Mesh Response Equipment to a Starlink satellite tv for pc terminal.

CCR Neighborhood volunteers are able to be deployed to pure disasters and humanitarian crises on brief discover. Their mission is to ship the connectivity wanted to assist first responders assess a group’s wants, coordinate an efficient response, and start the method of restoration whereas enabling victims to contact their family members and entry the essential info and companies they should rebuild their lives.

To make sure they’re prepared when the decision comes, lots of of CCR Neighborhood volunteers full in-person and digital trainings the place they discover ways to shortly set up connectivity in probably the most difficult of circumstances.

CCR Neighborhood volunteers just lately attended coaching in Rome and Munich, and I talked to a few volunteers who participated: Michele Festuccia, a options engineering chief, and Maurizio Cocco, a options engineer, each from Italy, and Anna Increase, a buyer expertise chief in Germany. Listed here are some highlights from our dialog.

Volunteers Answering the Name

I requested Anna why she signed as much as volunteer with the CCR Neighborhood.

“Seeing the totally different areas the place the CCR group has deployed—every little thing from hurricanes to floods and wildfires—and having the ability to assist these missions appeared like such an unimaginable solution to give again to the group,” she stated.

It’s a sentiment shared by many CCR volunteers: a need to contribute to one thing bigger than oneself, alleviate struggling, and stand in solidarity with these affected by crises. CCR permits us to do this within the context of offering Cisco’s distinctive connectivity choices—a strong mixture.

Michele shared , “When there’s a name, coaching positions us to be ready to go within the subject to assist a essential scenario.”

Extra Than Tech: The Human Aspect

Because the volunteers found, volunteering with the CCR Neighborhood is about greater than technical experience.

“Our coaching included disaster simulations. The observe situation stored us all on our toes and made us understand that when a disaster occurs, it’s a must to put together in several methods which might be outdoors of technical and actually come again to connecting with individuals,” defined Anna.

From left: CCR Neighborhood volunteers Martin Hertzum-Larsen, Jan Gahde, Anja Delaquis (NetHope), and Etienne David study to operationalize a CCR Mesh Response Equipment in Munich.

Having the technical chops to construct a wi-fi community when infrastructure is worn out is one factor, however being a CCR Neighborhood volunteer means additionally needing to ship human connection and adapt to evolving conditions within the second.

The empathy, communication, and problem-solving abilities wanted to successfully help communities throughout crises are as a lot part of the CCR volunteer instruments as our fast response kits, and our Community Emergency Response Car (NERV).

Giving Again is in Cisco’s DNA

Giving again as a behavior is a part of what it means to work at Cisco and is a crucial facet of our tradition. Actually, in FY24, 70,000 Cisco staff around the globe acted in service of individuals, planet, and society—the fifth consecutive yr that 80% or extra of Cisco staff did so.

Maurizio talked about how giving again is simply a part of what we do.

“It’s a part of our DNA to assist and to offer again throughout this sort of, let’s say, not so good interval,” he stated. “As a part of Michele’s group in the course of the 2016 earthquake in central Italy, with the assist of Cisco, we have been capable of assist these affected and create a type of hybrid Cisco Meraki package.”

Michele defined why they volunteered for the CCR Neighborhood. “We wished to hitch the CCR group to faucet into the sources of the bigger CCR group. As Maurizio stated, giving again is within the DNA of the individuals on my group.”

By bringing their experience and native expertise to the CCR Neighborhood and leveraging its sources, Maurizio and Michele are capable of amplify their affect and create a extra coordinated response to crises.

Worldwide Volunteers for Native Impression

Talking of native expertise, speaking to Michele, Maurizio, and Anna was a reminder of the worth of getting and coaching volunteers worldwide.

When disaster occurs, there’s nothing like having a deep bench of educated volunteers throughout the globe—one motive why capability constructing is likely one of the key tenets of CCR.

Sending somebody who doesn’t converse Italian to reply to a disaster in Italy is simply not environment friendly or efficient. At 800 volunteers sturdy, the CCR Neighborhood ensures we convey all kinds of experience and abilities to each disaster.

CCR Neighborhood volunteers construct a safe community throughout a coaching train in Rome.

Preparedness is Key

The coaching periods in Munich and Rome underscored the necessity for preparedness and clear processes in responding to crises.

“Being ready and proactive about all of the processes which might be wanted in going through a disaster is essential,” stated Maurizio. “Everybody who’s a volunteer must be managed in a correct manner so as to not create confusion however to supply assist, to supply assist.”

Michele agreed: “All our volunteers must know what they will do and what they will’t do. The potential to plan and to work in a synchronized manner with the opposite individuals is the primary precept that we realized in the course of the coaching.”

Past the Backside Line

In our dialogue, we additionally talked in regards to the perceived enterprise worth of philanthropic work in a company context—what worth humanitarian work brings to the enterprise facet of issues.

Michele defined that constructing relationships and constructing belief “is a unique solution to current Cisco to the market.”

A Name to Motion

We now have the deepest appreciation for the dedication and fervour of CCR volunteers. Their tales function a strong reminder that volunteerism is not only about offering assist, however about constructing connections, fostering resilience, and making a extra compassionate world.

As Anna put it, “Volunteering is a solution to give again to the group and assist Cisco as a result of it’s such an vital factor that we do.”

So, as we have a good time International Volunteer Month, allow us to all be impressed by these people and discover our personal distinctive methods to offer again to our communities. Whether or not it’s via expert volunteerism, direct service, or just providing a serving to hand, each act of kindness makes a distinction.

To study extra in regards to the affect Cisco staff are making, take a look at our “What We Completed” eBook about how we now have supported our staff’ giving again efforts and remodeled our collective contributions within the course of.

Share: