Connect with us

Tech

Next Silicon’s Dataflow Chip Could Disrupt The Processor Landscape

Published

4 days ago

on


Every so often, a semiconductor startup emerges claiming to have cracked a problem the industry’s biggest players have wrestled with for decades. Most fade quietly, but occasionally, one arrives that turns heads as a potential disruptor of the industry. Next Silicon, an Israel-based compute architecture firm founded by CEO Elad Raz back in 2017, appears to be one of those rare cases.

With its new Maverick-2 accelerator, built on what the company calls an Intelligent Compute Architecture, Next Silicon is betting on a long-pursued but rarely realized approach to accelerating HPC and data center workloads, known as dataflow computing. A dataflow architecture is designed such that the data itself, not instruction sequences, drives computation. The company believes it’s finally solved the twin barriers that kept dataflow architectures confined to research labs: programmability and practicality.

The Limits Of Conventional Compute And Next Silicon’s Solution

Modern computing remains largely defined by a concept called the Von Neumann model, a design that has served faithfully for over 80 years. CPUs and GPUs alike spend enormous amounts of silicon area shuffling instructions, managing branch predictions, and juggling cache coherency — doing an inordinate amount work over and above the actual math that a workload requires and developers care most about. CPUs still dominate for their flexibility, but they’re relatively inefficient. GPUs deliver significant parallel processing horsepower but require specialized programming and tight ecosystem dependence. ASICs typically offer extraordinary efficiency yet lock customers into single-purpose hardware at a huge cost impact.

Next Silicon’s proposition is simple but perhaps audacious: what if there was a fourth way — a compute engine as efficient as an ASIC, as parallel as a GPU, and as flexible as a CPU? At least that’s the goal of the company and it’s claiming a significant milestone has been achieved in that effort with its Maverick-2 accelerator.

Inside Next Silicon’s Maverick-2: When Data, Not Instructions, Rule

At the heart of Next Silicon’s new Maverick-2 accelerator is a dataflow execution fabric. Instead of relying on a program counter to step through instructions, the processor’s arithmetic logic grid, which is proprietary custom logic, activates whenever input data becomes available. Imagine an automated factory where every station starts work the moment its materials arrive, rather than waiting for a central manager to issue commands. At a very basic level, this is the model of dataflow computing.

This model allows the hardware to devote far more of its silicon area to compute resources rather than control functions — a reversal of traditional CPU designs where the vast majority of transistors must be dedicated to instruction handling. In theory, a dataflow processor architecture should translate to significantly higher utilization of silicon resources for compute and much better power efficiency.

Next Silicon claims its Maverick-2 can achieve up to 10X the performance of top GPUs while consuming up to 60 percent less power, all while running unmodified C++, Python, Fortran, or other framework code. Developers and software engineers are accustomed to months-long porting efforts for each new platform, but this is not the case with Next Silicon’s solution. The company also underscores that code based on Nvidia’s CUDA programming language for GPU AI accelerators can also run on Maverick-2 efficiently, again completely unmodified.

Typically, 90+ percent of a workload is processed by only a few specific calculations and this is another area where Next Silicon’s technology reportedly shines. Next Silicon claims that Maverick-2’s software layer profiles existing code in real time, identifies computational hot spots, and dynamically reconfigures compute resources — effectively turning a static chip into a self-optimizing engine.

The optimizer constantly monitors what the application is doing, which parts of the code run most often and what data patterns emerge. Using that live telemetry, it automatically builds and compiles specialized hardware configurations known as “Mill Cores” and saves them as ready-to-use images in the chip’s memory. These optimizations happen seamlessly in the background, without slowing down the workload. When a performance-critical part of the application appears, the hardware instantly reconfigures itself in nanoseconds using those pre-built images.

Unlike traditional systems that make assumptions at compile time, Maverick-2 optimizes itself based on runtime behavior, reshaping itself for each workload. One moment it can be tuned for massive parallelism, the next for deep pipelining. As a result, you get near-ASIC efficiency with the flexibility to adapt dynamically as an application or system use case evolves. This kind of on-the-fly adaptability could be game-changing for workloads ranging from HPC simulations to data analytics, AI training and AI inference.

As an aside, Next Silicon details that Maverick-2 also offers either single or dual 100 Gigabit Ethernet connectivity (dual 100GbE for a dual-die Maverick-2 accelerator) for scalability.

Next Silicon’s Current Real-World Proof Points

Unlike most early-stage chip architectures, Maverick-2 isn’t just slideware. The chip is already deployed in systems such as Sandia National Laboratories’ Spectra supercomputer, where it’s undergoing production-scale testing now.

According to internal benchmarks shared by Next Silicon:

  • In GUPS workloads (Giga-Updates Per Second), Maverick-2 reportedly runs 22× faster than CPUs and nearly 6× faster than GPUs.
  • In the High Performance Conjugate Gradients benchmark, it delivers performance on par with top GPUs while using about half the power.
  • In PageRank, a graph analytics test for web page authority ranking, it completed large (25 GB+) graphs that leading GPUs failed to finish.

These figures were obtained in a company-controlled environment, which of course calls for independent, 3rd party validation. Sandia’s commitment to validating the Maverick-2 platform, however, suggests that Next Silicon’s solution is worthy of significant investment.

The Role of RISC-V Inside Next Silicon’s Maverick-2

Next Silicon also announced a new RISC-V core CPU core that was born from an integral part of the Maverick-2 architecture. Maverick-2 has a RISC-V control processor that handles the serial logic and orchestration to keep the chip’s massive dataflow grid running efficiently. Essentially, it serves as the air-traffic controllers of the system, coordinating the movement of data so its parallel compute fabric stays fully utilized. Next Silicon developed it for implementation in its Maverick-2 accelerator and it was so good they decided to market it separately as well.

Called Arbel, and scaled-up a bit to stand on its own for even host CPU applications, Next Silicon claims its RISC-V core tests chip is potentially the highest-performing RISC-V CPU core design on the market. The company points to features like a 10-wide issue pipeline, deep reorder buffer, and integrated vector units built on TSMC’s 5 nm node

If true, this would mean Maverick-2’s control plane isn’t merely keeping pace with its dataflow side, it’s also pushing the boundaries of RISC-V performance. Still, that claim awaits independent validation. For now, the key takeaway is that Next Silicon is implementing tight integration of high-speed scalar control with adaptive dataflow compute for utilization efficiency traditional accelerators lack.

Strengths, Challenges And Competitive Context

If Maverick-2 delivers its claimed performance-per-watt advantage, it could be one of the most important compute architecture breakthroughs of the decade. The energy savings alone would be huge for data centers that are constrained by cost, power and carbon footprint caps.

However, architecture is only half the story. The semiconductor market rewards ecosystem maturity, not just raw performance. Nvidia’s dominance and AMD’s emergence in AI aren’t just about silicon, they’re about complete platforms with programming libraries like CUDA, developer familiarity, full rack-scale solutions and years of optimization.

Next Silicon says Maverick-2 runs completely unmodified code, but its long-term success will hinge on real integration into existing HPC and AI frameworks. Profilers, debugging tools, and runtime schedulers will need to support the architecture seamlessly. That’s a non-trivial lift for a smaller company taking on giants with software ecosystems measured in millions of developers.

There’s also the question of manufacturing scale and supply chain. Maverick-2 is fabricated on one of TSMC’s advanced 5nm nodes, putting Next Silicon in line with the same foundry capacity crunch that affects every major semiconductor firm. How the company balances volume, cost, and delivery will matter as much as it performance metrics.

Next Silicon’s Market Implications And My Final Take

That said, it’s clear to me that if Next Silicon’s technology holds up under independent validation, it could carve out a niche in HPC, simulation, and AI-driven scientific research, where customers value throughput and efficiency above all. The company’s claim of “drop-in programmability”—if proven—might then also open doors in hyperscaler data center acceleration and big data analytics.

Longer term, Maverick-2’s success could push competitors to revisit traditional chip architecture assumptions. Nvidia is already pursuing tightly coupled CPU-GPU designs with its Grace-Blackwell architecture, for example. If Maverick-2 demonstrates that dataflow can coexist and run efficiently with standard code, it could force others to rethink how they balance parallelism and programmability.

Next Silicon’s Maverick-2 may represent one of the most credible attempts yet to make dataflow computing commercially viable. By fusing adaptive, software-defined hardware with embedded RISC-V control logic, it sidesteps the historical trade-offs between performance, flexibility, and cost that have constrained compute architectures for decades.

But real disruption in high performance processors doesn’t happen in benchmarks; it happens in ecosystems. The company’s next challenge isn’t proving its chip works, it’s proving developers and customers can adopt it easily and profitably. If it can clear that bar, Next Silicon won’t just have built a faster more efficient accelerator, it will have rekindled the industry’s approach to far more efficient computing in the AI, HPC and exascale era.



Source link

Related Topics:
Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Tech

X Issues November 10 ‘Account Will Be Locked’ Twitter Security Warning

Published

57 minutes ago

on

October 26, 2025

By


Given the continuing popularity of Elon Musk’s X social network, and the swathing staffing cuts made when the world’s richest man bought what used to be (and still is in my mind) Twitter, it doesn’t make the cybersecurity headlines as much as you might have thought. With PayPal users currently warned of ongoing attacks, ditto WordPress website owners, and even LastPass password manager customers, all being in the threat actor crosshairs, this is a good thing. However, X users have now been warned that unless they make a change to a legacy Twitter security setting, they will be locked out of their accounts from November 10. Here’s what you need to know.

ForbesAct Now — Microsoft Issues Emergency Windows Update As Attacks BeginBy Davey Winder

The X Safety Team Issues Clarification After Warning Of Twitter Account Lockouts

Whatever you call it, X or Twitter, the social network isn’t immune to security threats. This year alone, I have reported on outages caused by a claimed DDoS attack and a warning for 650 million X users not to change their passwords. Sometimes, though, the perceived security threat comes from inside the building. Such was the case after the X safety team tweeted on October 24: “After November 10, if you haven’t re-enrolled a security key, your account will be locked until you: re-enroll; choose a different 2FA method; or elect not to use 2FA.”

This, rather unsurprisingly if you ask me, created a wave of concern amongst both ordinary users and security experts on the social media platform. One asked whether not using 2FA meant their account would remain active; another asked whether there had been a security breach; and another asked whether this only impacted passkey users?

ForbesGmail Account Lockout Warning — Users Must Check This 1 Setting NowBy Davey Winder

The confusion sat with X warning that “all accounts that use a security key as their two-factor authentication method to re-enroll their key to continue accessing X,” and adding that users could “re-enroll your existing security key, or enroll a new one.” A typical example of someone who knows what they are talking about but not how to communicate that in such a way to people who do not. Translating tech-speak into ordinary language is an essential skill and one that the X safety team appears to have misplaced on this occasion.

What X should have said, and ended up being forced into actually saying a day later, was: “To clarify: this change is not related to any security concern, and only impacts Yubikeys and passkeys – not other 2FA methods (such as authenticator apps). Security keys enrolled as a 2FA method are currently tied to the twitter.com domain. Re-enrolling your security key will associate them with x.com, allowing us to retire the Twitter domain. If this relates to you, you’ll be prompted automatically to re-enroll.”

ForbesOrganizations Can’t Deploy Passwordless, Declare Victory And Walk AwayBy Davey Winder





Source link

Continue Reading

Tech

Google’s Pixel 10 Series Could Soon Receive A Significant Performance Boost

Published

1 hour ago

on

October 26, 2025

By


Key Takeaways

  • Some Pixel 10 users have reported underperforming graphics and battery life.
  • Google has confirmed it will release driver updates, potentially addressing these issues and unlocking the Pixel 10’s full potential.
  • A new GPU driver update (version 25.2) is available that Google could deploy for significantly improved performance.

October 26 Update Below: A driver update may be essential for security reasons. This article was originally published on October 24

The Pixel 10 could soon receive a significant boost in performance, and potentially battery life, thanks to an upcoming graphics driver upgrade.

Reports indicate that the Pixel 10 series currently underperforms in graphics performance, notably scoring lower than the Pixel 9 Pro in some benchmark tests. The Pixel 10 series uses a PowerVR DXT-48-1536 GPU from Imagination Technologies, rather than the ARM Mali component found in previous models, which has led to complaints of poor performance, especially when playing popular, graphically intensive games such as Genshin Impact, and reduced battery life during video playback in apps like Netflix.

A primary cause of these performance issues appears to be an outdated GPU driver. The Pixel 10 shipped with driver version 24.3, lacking key features and optimizations necessary for the new GPU to reach its full potential. While it’s not unusual for a smartphone to ship with slightly older drivers, the switch from Mail to PowerVR appears to have exacerbated the issue.

A New GPU Driver Is Already Available

Fortunately, Imagination Technologies has since released driver version 25.2, adding official support for Android 16, along with significant performance enhancements, including support for the latest Vulkan 1.4 specification. That means this particular update could greatly benefit Pixel 10 users, and early indications suggest Google will most likely implement it.

A Google representative recently confirmed to Android Authority that the company plans to continue releasing GPU driver improvements in its regular system updates.

“We are continuing to improve driver quality in our monthly and quarterly system updates. For example, the most recent September and October patch releases included driver improvements. In future releases we are planning further GPU driver updates.”

Google has a strong track record in this area, having delivered a significant GPU performance boost for Pixel 8 Pro users in its December 2023 update, and several generations of Pixel smartphones received performance gains thanks to an updated GPU driver in the March 2025 Feature Drop. However, Google’s statement stops short of any commitment to any specific driver versions or performance benefits for the Pixel 10 range.

Google Pixel 10: Performance Upgrades Are Likely

If this pattern continues, Pixel 10 users can expect significant performance gains with future driver updates. Given that the new driver is already available from Imagination Technologies, we can hope it arrives sooner rather than later. For those currently experiencing graphics performance issues, a fix appears to be on the horizon.

October 26 Update: Existing vulnerabilities make a driver update extremely likely.

Driver updates typically offer not only performance improvements but also essential security patches, increasing the urgency of releasing updates.

The Pixel 10’s GPU driver is vulnerable, putting additional pressure on Google to provide an update.

According to Imagination Technologies’ published GPU Driver Vulnerabilities list, versions of the PowerVR GPU driver, up to and including version 24.3, currently deployed in the Pixel 10 series, contain critical vulnerabilities that could result in system instability, reboots and non-privileged access to secure data.

Addressing these vulnerabilities will be a high priority for Google, making a driver update urgent. Imagination Technologies doesn’t appear to have released any driver updates between versions 24.3 and 25.1, which fixes all of the listed vulnerabilities and adds most of the improvements listed in this article.

This means Google’s next Pixel 10 GPU driver update will likely be to least version 25.1, if not the latest version 25.2 released to partners on Oct 8. Pixel 10 Users can therefore expect significant improvements in performance, reliability and security once Google releases this update.

Follow @paul_monckton on Instagram.

ForbesNano Banana AI Image Editing Comes To Billions Of Google UsersBy Paul MoncktonForbesGoogle Photos’ New Update Leaves International Users Wanting MoreBy Paul Monckton





Source link

Continue Reading

Tech

Edifier Unveils Upgraded S880DB MKII Active Speakers For Hi-Res Performance

Published

2 hours ago

on

October 26, 2025

By


Edifier is a speaker brand that’s rapidly building a reputation for producing premium mid-range speakers that sell for very affordable prices. Edifier has just announced the launch of the S880DB MKII, its fully upgraded successor to the brand’s acclaimed S880DB model. This latest version features enhanced acoustic performance, redesigned circuitry and an improved user experience.

The S880DB MKII speakers now feature a completely upgraded driver system that can produce a richer and more precise sound. The tweeters have been upgraded to a 1.25-inch titanium dome model with a 25mm voice coil and neodymium magnet. The output through the tweeters is 12W per channel. The previous model used a 1-inch tweeter, so the improvements deliver clearer and more transparent highs that offer more accuracy and detail.

Complementing the upgraded tweeters are a pair of redesigned 3.75-inch mid-bass drivers with an output of 32W per channel for a clearer sound across the entire mid-range. The woofers also provide an extended low-frequency response that goes as low as 50Hz for a fuller and richer bass response. The overall sound of the upgraded drivers is a more solid and balanced soundstage.

High-Quality Sound

At the heart of the new S880DB MKII speakers are high-performance chipsets sourced from Texas Instruments. Edifier has used the powerful XMOS XU216 onboard processor, which provides advanced digital audio processing and can support high-resolution audio playback up to 24-bit/96kHz, for clearer sound and ultra-low distortion.

The S880DB MKII speakers are designed to meet and exceed the specifications required to gain Hi-Res Audio certification. The speaker’s digital interfaces support resolutions up to 24-bit/192kHz, while the Texas Instruments analog front end operates at 24-bit/96kHz, for precise signal handling and clarity across the entire frequency range.

At the rear of the S880DB MKII’s primary speaker unit are all the inputs for connecting the speakers to a digital or analog source. Users can choose from two RCA line-in connectors, a digital coax and optical inputs, plus a USB-C port, which is ideal for connecting these speakers to a computer so they can be used on a desk setup. There’s no 3.5mm input, but a cable is provided that goes from 3.5mm to RCA.

Bluetooth Connectivity

Finally, these speakers also include a Bluetooth 5.3 receiver for wireless input. The S880DB MKII speakers support the LDAC audio codec with transmission rates up to 990kbps, which means high-resolution wireless audio streaming up to 24-bit/96kHz.

Using the LDAC codec ensures audiophile-grade sound quality over Bluetooth from compatible Android devices running Android 8.0 and higher. Anyone using the speakers with a Mac or iPhone could use a Bluetooth dongle that supports LDAC. In addition, the speakers also support the vanilla SBC codec, but not AAC. The sound of these speakers can be shaped using the three rotary controls at the rear of the primary unit. The three knobs control volume, treble and bass.

The S880DB MKII also includes a dedicated SUB OUT port for the connection of an external subwoofer — like Edifier’s recently announced T5S model — and an even deeper bass response. This feature provides a fuller and more immersive low-end that’s suitable for music, movies and gaming. The one thing missing is a front-mounted headphone jack, which a lot of people use for private listening when seated at their computer.

User-Friendly Controls

Controlling the speakers is easy using the included puck-shaped remote control, which is just as well because those rear-mounted controls aren’t particularly convenient to reach. The S880DB MKII speakers come with a newly designed and rechargeable 2.4GHz remote control for quick access to playback and volume levels.

On the front of the primary speaker is an integrated OLED display featuring auto-dimming and providing clear and real-time feedback without too much glare. The speakers are also fully compatible with Edifier’s ConneX smartphone app, which can be used to fine-tune settings and manage the source inputs directly from a smartphone’s screen.

Pricing and Availability:

The Edifier S880DB MKII speakers are available now for $399.99 from Amazon.com.

Tech Specs:

  • Total output power (RMS): Treble: 12W x 2 / Mid-low: 32W x 2.
  • Drivers: 3.75″ long-throw aluminum diaphragm / 1.25″ titanium diaphragm dome tweeters.
  • Frequency response: 50Hz – 40kHz.
  • Signal-to-Noise ratio: ≥ 85dB(A).
  • Audio inputs: USB-C, Optical, Coaxial, Line In 1, Line In 2, Bluetooth.
  • Input sensitivity: USB-C: 500 ± 50mFFS OPT: 500 ± 50mFFS, COAX: 500 ± 50mFFS, Bluetooth: 500 ± 50mFFS, Line In 1: 500 ± 50mV, Line In 2: 700 ± 50mV.
  • Bluetooth: V5.3.
  • Supported audio codecs: SBC, LDAC.
  • Dimensions (W x H x D): 145 x 237 x 207mm (Primary speaker) /145 x 237 x 192mm (Passive speaker).
  • Weight: 6.27kg.



Source link

Continue Reading

Trending