At the 2024 Taipei International Computer Show, Kneron, an AI chip startup, unveiled its next-generation edge AI products: the KNEO 330 server and PCs equipped with their new KL830 NPU chip.

The company was founded in 2015, and it focuses on Edge Computing AI

Founded in 2015, Kneron focuses on providing edge computing AI (edge AI) technology. They boast big-name investors like Horizons Ventures, Qualcomm, and Alibaba. Kneron aims to offer an alternative to cloud-based AI solutions from tech giants. Their technology allows companies to build internal servers for specific AI applications, addressing data privacy and security concerns. This could make AI solutions more accessible to businesses hesitant to rely on external cloud services.

Two products were showcased:

• KNEO 330 server: This server offers 48TOPS of AI computing power, supports up to 8 users simultaneously, and works with applications like LLM and Stable Diffusion. Kneron claims it can achieve cloud-like performance for RAG (Generative Adversarial Networks) at a lower cost, potentially reducing AI expenses for small businesses by 30-40%.
• KL830 NPU-equipped PC: Specific details weren’t revealed, but a demonstration HP device showcased its capabilities. Kneron is partnering with major PC manufacturers to make them the first customers for this chip.

The KL830 is a neural processing unit (NPU) designed to work alongside CPUs and GPUs for AI tasks. It offers high computing power (10eTOPS@8bit) with low power consumption (2W) and handles various large language models efficiently.

Kneron also mentioned a USB AI accelerator stick powered by the KL830. This could transform any device, like routers, IoT cameras, or even traditional computers, into an edge AI device.

(Via)

The post Qualcomm & Foxconn Backed AI Chip Startup Kneron Unveiled a New NPU Chip and Server, Competing with Nvidia appeared first on Gizmochina.

Samsung is gearing up to launch its latest flagship level Exynos SoC next year. This chip has been highly anticipated since AMD is working on its GPU. Although, a new leak has revealed that the performance gains might be disappointing.

According to @AhmedQwaider888, the South Korean tech giant’s cooperating with Team Red might not bring a big breakthrough. The tipster revealed that the Exynos 2200 might only offer a 5 percent gain over the Exynos 2100, while its GPU performance is higher by around 17 percent. By comparison, the Exynos 2100 offered a sizeable 30 percent gain in CPU and a 40 percent bump in GPU performance over the Exynos 990 SoC. In other words, the Exynos 2200 is offering quite lackluster growth. But, there is still hope for the new chipset.

While the leak shares modest performance improvements, there is one area where the upcoming chip shines. The Exynos 2200 SoC feature an NPU (Neural Processing Unit) that offers a whopping 117 percent performance boost in terms of artificial intelligence (AI). Through this, we can expect improvements in various areas like the Bixby voice assistance that could be faster. Similarly, other AI powered tasks will also receive a boost, including photography.

Although the figure points at just a 17 percent raw GPU performance growth, AMD being a pioneer in this field. So, it might have the SoC offer other kinds of improvements and optimizations. This could be in the form of better support for 3D games to enable a smoother gaming experience in high graphical fidelity titles. Another notable inclusion could be ray tracing support as well. Albeit, this would likely be a cutdown version compared to its PC counterpart. Unfortunately, only real world tests would reveal how well the new chip fare against its competitors and its predecessors. So stay tuned for more updates.

RELATED:

• Samsung Galaxy Tab A8 teased to launch soon in India
• Samsung Galaxy S22 Ultra features cinematography features and has 58% better video stability, tipster reveals
• Samsung Galaxy M33 5G appears on Geekbench with Exynos 1200 SoC

The post Samsung Exynos 2200 may offer lackluster performance gains, but its not all bad news appeared first on Gizmochina.

Earlier today (14th December 2021), the OPPO INNO Day 2021 event had officially kicked off. During this conference, the company unveiled a number of new technologies, which included its first self develop NPU chipset, the MariSilicon X.

The Chinese smartphone maker’s Founder and CEO, Chen Mingyong, showcased the new first self developed chipset. As per the official announcement, the new MariSilicon X NPU is an imaging based chipset that was independently designed and developed by the company. The senior executive stated that the chipset was named after the Mariana Trench, which is the deepest part of the world. This particular name was set to signify the great depth of the company’s “imagination” and innovation.

During the announcement, the Founder also added that tech giants need to develop core technologies to be prepared for any challenged that they might face in the future. In other words, this chip is also its step towards more self sufficiency. The new MariSilicon X is manufactured using the 6nm process from TSMC and seeks to improve imaging performance of upcoming Oppo smartphones. The company further added that the new chipset will be available in handsets starting from next year.

The new chipset will offer real time AI computing to improve energy efficiency while also offering “industry leading” ultra HDR capabilities. This chipset would also enable lossless real time RAW captures and maximize the RGBW sensor’s performance as well as offer extreme dynamic range via 20bit HDR. Notably, OPPO also officially announced that this new chipset will be featured in 2022 Find X series flagship phones, which would mark the arrival of the dual chip technology.

RELATED:

• OPPO Find N uses UTG foldable screen from Samsung Display
• OPPO Find N foldable phone live images leaked, company also drops new teaser video
• OPPO Reno7 Pro League of Legends Mobile Game Limited Edition goes on sale in China

The post OPPO unveils its first self developed MariSilicon X NPU imaging chip appeared first on Gizmochina.

During the upcoming INNO Day 2021 event, Oppo is gearing up to unveil its latest technologies and innovations. This will includes its first self developed NPU (Neural Processing Unit), smart glasses, and more.

We've developed a cutting-edge neural processing unit (NPU) called the MariSilicon X. Can you guess what it does?

Tune in to #OPPOINNODAY2021 to find out! pic.twitter.com/BbECacK70s

— OPPO (@oppo) December 8, 2021

Leading up to this event, the Chinese smartphone maker has already teased the aforementioned technologies and even offered us a glimpse at its retractable camera technology as well via Twitter. The brand has even officially announced that it will be showcasing its first foldable smartphone during the event, which is set between 14th December and 15th December in Shenzhen, China. Notably, the Founder and CEO of the company, Tony Chen, will also be sharing a keynote speech in the event, talking about the company’s updates corporate strategies and new research and development strategies.

Oppo added that “other senior executives will introduce the new NPU and smart glasses,” during this event. So, we can expect to hear more about the new Retractable Camera technology, as well as Digital Human and other technological breakthroughs in the fields of imaging, AI, AR, and even 5G. A highly anticipated announcement would also be for the company’s first NPU, which is being called the MariSilicon X.

As per our previous report, the company had also revealed on Twitter that it is gearing up to announce the Oppo Find N foldable phone during this event as well. It will be the company’s first foldable handset and has been in development for four years and has gone through six generation of prototypes. So stay tuned for more updates, as we will be covering this major company conference.

RELATED:

• OPPO Reno5 Pro 5G ColorOS 12 Open Beta program begins in China, India, Pakistan
• OPPO Find N foldable phone to launch on December 15, design teased
• Oppo, OnePlus phones tipped to use in house chips based on TSMC’s 6nm node

The post Oppo to showcase MariSilicon X NPU and new Smart Glasses at INNO Day 2021 appeared first on Gizmochina.

Earlier today (23rd October 2020), Huawei had officially launched the Mate 40 series. The lineup is the latest flagship from the company and feature top notch specs like the HiSilicon Kirin 9000 chip, which has even topped the list in an AI Benchmark.

The Chinese tech giant has also officially unveiled its new Kirin 9000 chips earlier this week. These were built on the new 5nm process from TSMC and boasts improvements in performance and energy efficiency over its predecessor. Furthermore, while the company had stated that the NPU has also been greatly improved with the new chips, no official figures were revealed. But now, the AI Benchmark list of ETH Zurich has the Huawei Mate 40 Pro take the top spot with the Kirin 9000 SoC.

Editor’s Pick: Huawei Mate 40, Mate 40 Pro, and Mate 40 Pro+ arrive with new designs, 5nm chipsets, and top class cameras

In the list, the HiSilicon Kirin 9000 is in the 1st position with a total of 216.6 points, beating both the MediaTek Dimensity 1000+ chip and even the Qualcomm Snapdragon 865+ SoC. For those unaware, the Kirin 9000 from Huawei is an 8 core CPU with a 24 core Mali-G78 GPU, alongside a 2 Big and 1 Tiny core NPU. It also houses Huawei’s most advanced ISP technology to date.

According to Huawei, the Kirin 9000 is by far one of the most powerful mobile processors in the market. It consists of 30 percent more transistors that Apple’s A14 chip and its CPU performs 10 percent faster than the Snapdragon 865+ as well. Furthermore, it beats its Qualcomm rival by a large margin of 240 percent in terms of NPU performance and offers double the network download speeds than the Snapdragon X55 5G Modem as well.

UP NEXT: PUBG Mobile might return to India as PUBG Corporation as hinted by a LinkedIn job listing

The post Huawei Mate 40’s HiSilicon Kirin 9000 chip ranks 1st in AI Benchmark appeared first on Gizmochina.

U.S fabless chipmaker Qualcomm first announced its new Snapdragon 700 series at MWC 2018 with the Snapdragon 710 being the first of the series to find its way into smartphones. The new SDM700 series lies in between the flagship Snapdragon 800 series and the premium mid-range Snapdragon 600 series. It draws a number of advanced features from the former and is very suitable for high-end models.

Apart from the Snapdragon 710, little is known of other members of the series but a new rumour has emerged which deals with the Snapdragon 720. While there is presently no information on the specification of the Snapdragon 720, it is only known to be an enhanced version of the Snapdragon 710. However, the SDM720 is said to have a Neural Processing Unit (NPU) similar to Huawei’s Kirin 970 which is dedicated to AI acceleration. If this turns out to be true, it will be the first time Qualcomm as the current Snapdragon platform doesn’t have a dedicated AI hardware unit but rather relies on the Artificial Intelligence Engine(AIE). This is software-based and the efficiency in the specific AI scene is still low.

In addition to Snapdragon 720, there had been a leaked poster that appeared a few months back containing details of the Snapdragon 730. The earlier leak tipped that Qualcomm’s SDM730 will be manufactured using a Samsung 8nm LPP process (10nm upgrade), integrating two Kryo 4xx 2.3GHz, six Kryo 4xx 1.8 GHz CPU core and specially equipped with NPU 120 unit!

Related: Mysterious Snapdragon 680 Appears on Geekbench with Impressive Scores

Finally, the new rumour also includes that at least three mobile phone manufacturers are currently developing Snapdragon 720-powered models, of which Samsung is expected to be the first to utilise the chipset on its mid-range Galaxy series.

(source)

The post Snapdragon 720 rumoured to feature a dedicated NPU for AI just like Kirin 970! appeared first on Gizmochina.

A few years back, Huawei was just a big networking and telecommunications giant from China. A regular phone user might have never even heard of this company. Fast forward to 2018, and Huawei’s smartphones are found in most of the top markets in the world. The company has slowly but steadily climbed up the smartphone rankings, now occupying the third place in the ever competitive smartphone market, just behind Samsung and Apple.

How did Huawei manage such a feat? Well, some of this credit has to go to its R&D department and it looks like this department will further fuel the company’s growth in the coming years.

Now, iFeng reports that Huawei has a formidable technology under development that will be launched very soon. The company has been making billions of dollars in investment in its R&D department, with the spending rising to nearly $14 billion last year. This makes Huawei one of the top companies to spend in R&D in the world. No wonder the company has managed to churn out innovative features and technologies like the triple camera setup found on the Huawei P20 Pro!

RELATED: World Intellectual Property Day: Huawei Reveals It Had 74,397 Patents As At Dec. 2017

Recent reports claim that the company’s R&D spending in the coming years would reach as much as $20 billion! With so much money being invested in this department, we are expecting some exciting features on upcoming Huawei smartphones. Unfortunately, iFeng wasn’t able to get more information on this new formidable technology from the company.

We don’t even know whether the company’s new breakthrough is in the hardware or the software department. If it’s in the hardware department, it could be something related to an independent NPU module alongside its processors for its phones for faster and efficient Machine learning, but at this moment, this is just a speculation.

RELATED: Huawei Said to Launch World’s First Foldable Smartphone in November

If its in the software department, it could be something related to its OS. Remember the Honor Magic smartphone that launched at the end of 2016? The phone had a lot of sensors that ensured privacy and special features such as WiseScreen, and FaceCode Intelligent Recognition. Then there was the Magic Live Assistant which added an extra layer of AI-based software for special tasks.

We are not sure of what exactly this breakthrough technology might be, but I’m sure it’s going to be something exciting. Maybe the company finally perfected foldable smartphones? We can’t be sure but stay tuned for more updates.

iFeng reports that the technology will launch alongside an Honor smartphone in the near future. An exact time frame wasn’t given, but it should be before the end of this year.

(Source)

The post Huawei To Release A Breakthrough Technology Later This Year, Could Be Inside A Honor Phone appeared first on Gizmochina.

The Huawei Kirin 970 is the first chip with a dedicated NPU (neural processing unit) for AI related tasks. Apple’s new iPhones are powered by the A11 Bionic which has a neural engine for machine learning tasks. Samsung is already behind as its Exynos 8895 doesn’t have any of the above mentioned features, so it is working on new chips that will fill this gap.

According to a source from Samsung’s partners who spoke to Korea Herald, “(Samsung) is in the middle of developing several types of chips that will be capable of processing massive data from AI applications on devices, eliminating the need to communicate with cloud servers”.

The advantage of having a dedicated processing unit for data processing is that data used by AI apps do not need to be stored in the cloud. They can be stored locally and accessed faster, with improvements in performance reaching up to 50%.

Samsung itself has officially mentioned that its chips will have Neural Processing Units. Speaking at a forum, the president of its semiconductor arm, Kim Ki-nam said “the existing central processing unit and graphic processing unit chips make it hard to achieve efficiency in AI computing. The NPU will help address the efficiency challenge. However, NPUs so far have storage capacities that are equivalent to a thousandth of a human brain.”

READ MORE: 270,000 Samsung Galaxy Note 8 Units Sold in Initial Two Days of Sales in South Korea

There is a race between tech giants as AI is predicted to create new marketing opportunities says market intelligence firm, TrendForce. Samsung has invested $300 million in Graphcore, a British AI chip startup that develops IPU (Intelligence Processing Unit) for machine intelligence to make up for areas where it is deficient.

With this news, we expect the Exynos chip that will power next year’s Samsung flagships to have dedicated processing units for AI tasks.

(Source, via)

The post Samsung Working On New Exynos Chips With Dedicated AI Processing Cores appeared first on Gizmochina.

The Kirin 970 is official and Huawei has set a record with the new flagship chip. The 10nm chip is the first mobile computing unit with a built-in AI computing Neural Processing Unit NPU.

Before we dive into the configuration of the Kirin 970, lets talk about its AI NPU. Instead of letting the CPU carry out AI computing tasks, the NPU is assigned the tasks and it does it better and faster. According to the figures provided, the Kirin 970’s NPU delivers performance that is 25 times more than the CPU’s,  and efficiency that is 50 times greater. That means lesser time and power required for AI tasks.

The TSMC-built Kirin 970 is an 8-core CPU with a clockspeed of uP to 2.4GHz i.e. 4 x Cortex A73 at 2.4GHz + 4 x Cortex 53 at 1.8GHz. It has a 12-core Mali G72MP12 GPU, making it the first commercially available chip to use ARM’s latest GPU. The GPU provides a 20% performance increase over the previous generation as well as a 50% increase in efficiency. Its Kirin NPU can perform 1.92 Teraflops when using 16-bit floating point numbers.

The Kirin 970 has 4.5G Cat.18 LTE and offers gigabit LTE speeds of 1.2Gbps. It also features an advanced dual ISP with face and motion detection, 4-hybrid focus, low-light and motion shooting. Since the Kirin 970 will be appearing in the Huawei Mate 10 series first, expect them to have an impressive camera performance.

The chip also has HDR10 support and 4K video encode and decode. It supports LPPDR 4X RAM (1883MHz) and UFS 2.1 storage. It also features an i7 sensor processor, iNSE and TEE security engines and support for HiFi audio.

READ MORE: Huawei Or Lenovo May Purchase Fujitsu’s Smartphone Business

Huawei says it has built an open AI ecosystem so apps can access NPU capabilities directly and via 3rd party AI frameworks. The Mate 10 and Mate 10 Pro are the first two devices that will be powered by the new chipset and they will be announced in Munich on October 16.

[embedyt] https://www.youtube.com/watch?v=5k566SwWDlc[/embedyt]

The post Huawei Unveils The Kirin 970, The World’s First Processor With A Dedicated NPU appeared first on Gizmochina.