McAfee believes that the transition to chiplets design in the mobile field will inevitably have a negative impact on processor power consumption.

Chiplet, also known as chiplet or small chip, is a type of die (bare chip) that meets specific functions. Through die-to-die internal interconnection technology, multiple module chips and underlying basic chips are packaged together to form a system chip. , to achieve a new form of IP reuse.

Processor TDPs for ultrabooks and thin-and-light notebooks typically range between 15-30W. AMD currently relies entirely on monolithic die design, and that won’t change in the short term.

When asked about why AMD’s chiplet design has been successful in desktop processors but has not been extended to the mobile field, McAfee’s answer is as follows:

 When developing new products for desktop and mobile, we consider monolithic and chiplet architectures.

 However, there are bound to be huge challenges in energy efficiency when notebook computers transition to chiplet architecture. If mobile chips adopt the chiplet architecture, they must make compromises in terms of power consumption.

 All things considered, we are now seeing processors with monolithic architectures that are more energy efficient than chiplet architectures. If this changes in the future, we may consider using chiplets in the mobile space.

It should be added that McAfee’s answer is limited to AMD’s mobile processors (APUs). Some Dragon Range processors in AMD’s Ryzen 7000 series of mobile processors use chiplet architecture, which are essentially desktop Ryzen 7000 chips but in a mobile BGA package.

THIS IS A SPONSOR PROMOTION: >>>>>>>>>>>>>

Geekwills is an online shop that connects consumers with millions of products and brands around the world with the mission to empower them to live their best lives. Geekwills is committed to offering the most affordable quality products to enable consumers and sellers to fulfill their dreams in an inclusive environment.

The post AMD explains why mobile APUs do not use chiplets architecture appeared first on TechGoing.

“Avatar Pandora’s Frontier” will be officially released on December 7th, landing on PS5, Xbox Series X|S, PC and Amazon Luna. The game is now available for pre-order on Ubisoft’s official website, starting at RMB 348.

The game introduction attached is as follows:

After being abducted by the human militarization corporation RDA, you play as a Ne’vi child who is forced to train and assimilate to serve their purposes. Fifteen years later, you are free, only to find that you are a stranger in your birthplace. Retrieve your lost heritage, discover what it means to be a Na’vi, and unite with other tribes to resist the RDA and fight to protect Pandora.

Not long ago, AMD and Bethesda Game Company cooperated to announce a limited edition RX 7900 XTX graphics card with the theme of the “Starry Sky” game, which was also given out through the event.

The post AMD Announces Avatar Pandora’s Frontier theme RX 7900 XTX graphics card appeared first on TechGoing.

In addition to bug fixes for its various components, the new driver also introduces application support for “new program support and some new drivers”. In addition, the new package includes three new components – the AMD SFH 1.1 driver, the AMD PMF-7040 series driver, and an oddly named AMD interface driver.

SFH, or “Sensor Fusion Hub”, is a component related to mobile devices that can present various sensing devices (such as cameras, motion sensors, etc.) to operating systems, such as Windows Hello.

The AMD PMF (Platform Management Framework) here is a version of the driver specific to the Ryzen 7040-series “Phoenix” mobile processors, while the “AMD Interface Driver” is a bit of a mystery.

According to the TechPowerUP test, the new version of the driver has improved the performance of some old processors with 3D-V cache, but the improvement rate of different models is also different.

Older 3D-V cache processors such as the R7 5800X3D are said to be single-CCD-based models because the cache management mechanism is easier, whereas the AM5 platform’s 3D-V cache processor has two CCDs with a separate cache subsystem.

To ensure that the correct workloads are allocated to the different CCDs, AMD has been using special drivers to assist Windows in this scheduling task.

They believe that the “AMD interface driver” here may be a backend system used to enhance the 3D cache management system, and the application may achieve a better scheduling scheme by strengthening some kind of control over the 3D-V cache control driver.

The post AMD released Ryzen CPU chipset driver to optimize 3D cache processor performance appeared first on TechGoing.

According to the exposed shipment list, AMD’s new-generation Threadripper Threadripper is now available in three models – 7945WX / 7985WX / 7995WX – and its power consumption appears to have been bumped up from 280W to 350W.

Currently, AMD’s flagship Threadripper model for sale is the PRO 5995WX, which has specifications of 64 cores and 128 threads, an acceleration frequency of 4.5GHz, 256MB L3 cache, and a default TDP of 280W.

Summarized the Threadripper 7000 breakout below:

This series of processors will be divided into two platforms, workstation and HEDT, based on AMD Zen4 CPU architecture. Threadripper PRO 7000 series processes up to 96 cores, support 8-channel memory, with 128 PCIe 5.0 lanes does not support overclocking. Threadripper 7000 support 4-channel memory, with 64 The Threadripper 7000 supports 4-channel memory with 64 PCIe 5.0 lanes and supports overclocking.

AMD has not announced the release date of the Threadripper 7000, some sources say that it will be released in September this year.

The post AMD Threadripper 7945WX / 7985WX / 7995WX models revealed with 350W appeared first on TechGoing.

Biostar A620MT AM5 motherboard adopts AMD Socket AM5 A620 chip architecture, supports 7000 series processors including AMD Ryzen 5 7500F, supports up to 96GB DDR5 memory, PCIe 4.0, and supports 64 Gb/s bandwidth.

The motherboard is equipped with a USB 3.2 Gen 2 Type-C port on the front and supports HDMI 4K resolution, Realtek Gigabit LAN, and ALC897 7.1 channels.

Currently, the price information is unknown.

The post Biostar Introduces A620MT AM5 Motherboard Supporting AMD Ryzen 7000 Series Processors appeared first on TechGoing.

Pixabay

In the past, the company’s AI engine, Ryzen AI, was introduced at the event, saying that AI has been around for a long time, but it is only because of the emergence of ChatGPT that the market has begun to perceive AI. Lisa Su emphasized that there are great opportunities for AI, but the market is changing very fast. She believes that AI will be ubiquitous in the next five years, but in addition to hardware, the success of AI also requires software support, and AMD will invest in this area as well.

For the PC market, Lisa Su still expressed optimism, she pointed out that AMD has quite a good relationship with the computer supply chain and chip foundries. Lisa Su stressed that in 2017, AMD’s market share was only 1%, and then because AMD and TSMC and other chip foundry partners have a close working relationship, the establishment of a cooperative ecosystem, has led to today’s achievements.

Lisa Su this trip to Taiwan will meet with Xiangshuo, Sun Micron Group, and visit TSMC President Wei Zhejiang and other senior management, to the PC, server supply chain, customers, its latest product blueprint and outlook, and listen to the views of important partners for the industry market conditions, then will also discuss with TSMC 4/3 nm process and advanced packaging cooperation issues.

The post AMD’s Lisa Su: AI will be everywhere in the next five years, still optimistic about the PC market appeared first on TechGoing.

There is a guess that the code-named Phoenix2 chip uses “big and small cores”. The former contains 2 Zen4s and 4 Zen4cs, and the latter is 2 Zen4s and 2 Zen4cs. The Zen4c core is mainly the third-level cache halved (16MB to 8MB ), also reduces the frequency.

Now, the exposure blogger “Golden Pig Upgrade Package” has released the real photos of Phoenix1 and Phoenix2. It can be seen that the latter is obviously smaller, but the package is the same FP7.

He also put forward a new point of view. The Phoenix1 chip has a maximum of 8 CPU cores and 12 GPU cores (Radeon 780M/760M), while the Phoenix2 chip has been simplified to a maximum of 6 CPU cores and 4 GPU cores (Radeon 740M). The contribution of shrinking mainly comes from the streamlined core display.

However, it is unclear whether the Phoenix2 chip also streamlines the AI ​​engine unit.

Some people may ask: Is it worthwhile to tape out a chip for such two entry-level models?

The answer is obviously yes, after all, the complexity and difficulty will be much smaller if some cores are streamlined and then taped.

Ryzen 5 7540U and Ryzen 3 7440U have not yet been listed on the relevant equipment, but they should have been shipped to OEM manufacturers.

The post AMD will really save money: not only the size of the core, but also the core display appeared first on TechGoing.

ExpertCenter PN53 can be equipped with two PCIe 4.0 M.2 SSDs supporting DDR5 memory. IT House learned from the report that the mini PC is equipped with cleverly designed cooling functions and artificial intelligence-driven noise reduction technology, and can be connected to 4 external 4K displays at the same time, or an external 8K display.

Compared to the previous PN51 model, the ExpertCenter PN53 offers a 34% increase in single-thread performance and a 29% increase in multi-thread performance. In the PCMark 10 Essentials panel test, the ExpertCenter PN53 showed 10 percent faster web browsing, video conferencing, and application launching.

It also outperformed the ExpertCenter PN51 by 10 percent in the productivity test group, which measures system performance for everyday office applications, including spreadsheet and writing workloads.

The post With Ryzen 7000 series processors, ASUS launches ExpertCenter PN53 mini PC appeared first on TechGoing.

Aris / crmaris used the Powenetics v2 test tool and found that the peak power consumption of the AMD Ryzen 7950X3D processor can reach 130W when it is idle.

And the peak CPU load is only 3.53%, which is obviously unreasonable. Even with curve optimization and the -15 setting enabled, peak standby power was still close to 125W.

Aris found in the test that the Ryzen 7900X processor also has this problem. The peak power of the idle state is 109W, and the CPU load is 5.12%, which is obviously unreasonable.

The post AMD Ryzen 7000 series processors again exposed to high standby power consumption appeared first on TechGoing.

As shown in the figure above, the 3DMark TimeSpy graphics card with 780M core display of R9 7940HS and DDR5-5600 memory is divided into 2994 points. In contrast, the 680M core display with DDR5-4800 memory scored 2490 points, and the 780M improved by 20%. According to the UP, the performance of the 780M is basically equivalent to that of the NVIDIA GTX 1650 Max-Q.

It should be noted that the R9 7940HS also supports LPDDR5-7500 memory. With higher frequency memory, the performance of the core display will be even higher.

Notebooks and mini hosts equipped with R9 7940HS or R7 7840HS are expected to be available in large quantities in May.

The parameters of R9 7940HS attached are as follows:

The post AMD R9 7940HS core scores: up to GTX 1650 Max-Q with DDR5-5600 memory appeared first on TechGoing.