Tag Archives: CPU

Kunpeng 920: Huawei Unveils World’s Highest-Performance ARM-based CPU

Just in time for CES 2019, Chinese tech leader Huawei, strategically unveiled the Kunpeng 920, the industry’s highest-performing ARM-based processor. At Huawei’s headquarters in Shenzhen, China, Huawei released the impressive CPU, intended for handling applications like big data processing and distributed storage.

William Xu, Director of Huawei’s Board and Chief Strategy Marketing Officer, said during the press conference that “Huawei has continuously innovated in the computing domain in order to create customer value. We believe that, with the advent of an intelligent society, the computing market will see continuous growth in the future. Currently, the diversity of applications and data is driving heterogeneous computing requirements. Huawei has long partnered with Intel to make great achievements. Together we have contributed to the development of the ICT industry. Huawei and Intel will continue our long-term strategic partnerships and continue to innovate together.”

William Xu, Director of Huawei’s Board and Chief Strategy Marketing Officer (credit: Huawei)

Kunpeng 920: High-Performance ARM-based CPU

Huawei’s high standards of leading the industry are very clear with the release of the new Kunpeng 920.  The new CPU, similar to Huawei’s HiSilicon Kirin 980 (launched in October 2018), was  designed in-house and manufactured on a 7-nanometer processor. According to Huawei, the gain in performance between these two CPU’s emerges from a series of optimized branch prediction algorithms and an increased number of OP units, along with an improved memory subsystem architecture. The Kunpeng 920 provides both higher computing performance for data centers while decreasing overall power consumption. The Kunpeng 920 is packed with 64 cores, clocking an impressive 2.6GHZ, and paired with 8-channel DDR4 memory allows the Kunpeng 920 to score over 930 on the SPECint Benchmark, 25% higher than the industry benchmark.  For the sake of comparison, the previous industry leader was the Fujitsu 7-nanometer A64X. Beyond its speed is the CPU’s power efficiency, which performs 30% better than its competitors.

Kunpeng 920 performance (credit: Huawei)

In terms of system integration, the new CPU has increased with two 100G RoCE ports. Kunpeng 920 also supports PCIe Gen4 and CCIX interfaces, and provides 640 Gbps total bandwidth.  In addition, the single-slot speed is twice that of the incumbent offering, effectively improving the performance of storage and various accelerators.

TaiShan Series: Huawei’s ARM-based Server Built to Perform

To complement the Kunpeng 920, at today’s press conference Huawei also released its TaiShan series servers, powered by the company’s new CPU. Aimed at tackling three distinct issues, storage,  high-density, and balancing both requirements, the Taishan series designed three unique models. Powered by the Kunpeng 920, the TaiShan servers are built for big data, distributed storage, and ARM native application scenarios. These unique scenarios are ideal for the ARM architecture offering many advantages in many-core and performance per watt. The TaiShan series are ideal computing platforms for enterprises with high performance and low power consumption. Designed especially in big data scenarios, the TaiShan servers are tuned for optimal many-core high concurrency and resource scheduling to deliver a 20% computing performance boost.  

As Mr. Xu stated at the press conference, “The ARM industry is seeing a new development opportunity. The Kunpeng 920 CPU and TaiShan servers newly released by Huawei are primarily used in big data, distributed storage, and ARM native applications. We will work with global partners in the spirit of openness, collaboration, and shared success to drive the development of the ARM ecosystem and expand the computing space, and embrace a diversified computing era.”

TaiShan server (credit: Huawei

Open and Collaborative ARM Ecosystem is Geared for Mutual Success

For years, Huawei has continuously promoted industry cooperation and collaboration among its industry partners in terms of hardware, basic software, and applications. The company’s unique approach is predicated on the growing trend that an intelligent society, with all things connected, and sensing, is well underway. In light of these industry trends and application requirements, a new era of diversified computing is emerging. The ability to process multiple data types and scenarios are the driving force behind optimizing computing architecture.   

Summarizing the company’s vision of serving as an industry leader across all verticals, from telecommunications to servers to handsets, Mr. Xu noted that “With Kirin 980, Huawei has taken smartphones to a new level of intelligence. With products and services (e.g., Huawei Cloud) designed based on Ascend 310, Huawei enables inclusive AI for industries. Today, with Kunpeng 920, we are entering an era of diversified computing embodied by multiple cores and heterogeneity. Huawei has invested patiently and intensively in computing innovation to continuously make breakthroughs. We will work with our customers and partners to build a fully connected, intelligent world.”

The War of the Processors: Where AMD Went Wrong

 

The microprocessor has undoubtedly come a long way. The leap from four to sixty-four-bit chips has been a tremendous feat. And as expected, this remarkable advancement has brought to life fierce rivalry between industry players.

Although the recent influx of praiseworthy Chromebooks has popularised ARM chips, nearly all computers are powered by what are known as x86-based processors. Technicalities aside, these CPUs are primarily based on the Intel 8086 – a 16-bit microprocessor that was introduced in 1978 – and are a result of the gradual additions and extensions made to the 8086 over the years.

As leading manufacturers of x86-based CPUs, Intel and AMD have steadily achieved global domination in the world of computer processing. Each company has been kept on its toes by the other, and we’ve all benefited greatly from this rivalry.

In the recent years, however, AMD seems to have dialed down quite significantly. Simply going through the laptops in your college class, or the PCs around the office will give you a good picture of the dwindling competition. Intel processors are in far wider use than AMDs.

Some might argue that Intel has always been ahead. Albeit only by one year, it is the older of the two companies.  Moreover, it is only as a consequence of the conditions made by IBM for its first PC in 1981 that AMD became a second-source manufacturer of the already patented Intel 8086.

So, Intel may have made AMD into the company it is now, but that doesn’t mean the latter has always been in the former’s shadow. For instance, it was AMD that designed the highly successful Am2505 in 1971, the fastest multiplier of its time.

Today, however, no one can deny that Intel has cast a gloomy shadow onto AMD. But where did a company that freely benefited from the other’s technology go wrong?

Performance

With its FX-8350’s glass-shattering 8794.33MHz, AMD is the Guinness World Record holder of the highest frequency ever achieved by a computer processor. But we have all come to the realization that more clock speed doesn’t necessarily mean better performance.

AMD has historically taken pride in its emphasis on the number of cores and clock speeds, and on paper, this would make its chips faster than Intel’s. In reality, however, it is the issue of efficiency that negatively impacts AMD processors.

For illustration’s sake, let’s take the best of Intel’s Skylake architecture, the Core i7-6700K. The chip triumphs over the AMD FX 9590, despite having lower maximum clock speed and only half the number of cores.

What Intel does better is, instead of adding cores, it uses hyperthreading to keep the existing ones active at all times. When performing tasks that require the same processing power, therefore, Intel chips have always offered better performance over similarly classed AMD CPUs.

Pricing

A while back, PC enthusiasts marveled at AMD for offering reliable processors at fair prices. However, because competition has so far heated up, this is no longer the case.

Of course, you could buy the entry-level Athlon X4 860K – a 3.7GHz quad-core processor – for around $70, which is not bad, but the performance you’ll get will be way too low to justify the money saved if you chose it instead of something like the Intel Core i3-6100.

What makes the situation even worse is the high rate of CPU pricing fluctuation. Customers are opting for older Intel processors which still perform better than AMD chips, but having been in the market for a while are now almost as cheap.

You can, therefore, be duped into thinking you’re getting better value for the money, but instead end up with a processor that throttles with even the simple tasks. Many AMD chips are cheap, but it is because pricing them higher wouldn’t make any sense. This is why Intel largely dominates the exorbitant $200 and above price range.

Too much emphasis on dedicated graphics

There was a time when the graphics processing capabilities of AMD-based computers were far superior to anything offered by Intel.

Fast-forward to today, and Intel’s integrated graphics have all but closed the gap. The 6th generation Intel chips have largely been accepted in a market that is increasingly showing its flexibility to compromise on graphics power for more portable devices.

While not built for graphics-intensive tasks such as heavy gaming, the latest Intel HD, and Intel Iris Pro GPUs have proven capable of handling some demanding titles comfortably.

Admittedly, AMD has gone back to the drawing board, and the integrated graphics on its new A-series APU processors are currently giving Intel GPUs a run for their money. But the company has evidently taken its time to realize that not everyone shopping for a computer needs it to run The Witcher 3 or Mortal Kombat X at maximum settings.

Availability and support

AMD enthusiasts might have something to say regarding all the points we’ve made so far, but they will agree that their processors of choice are simply not compatible with as many components as Intel chips.

Take motherboards for example. You can find several options for your AMD chip, but the choices are not nearly as many as with Intel. Moreover, Intel boards make a little more sense, design-wise.

AMD has their metal connector pins on the chip, rather than on the board, as it is with Intel motherboards.

AMD’s design choice is great because you can rule out the board’s faulty pins as the cause of any processor issues. However, replacing a high-end chip is usually a lot more expensive than getting a new motherboard.

Wrap Up

Despite Intel’s unwavering command over the CPU market, AMD is finally gearing up a notch. The future, therefore, looks brighter for a company that has seemingly grown tired of being known only as second-best. 2017 is particularly an exciting year for the decades-old rivalry.

We are all eager to see how AMD’s highly-anticipated Ryzen processors will compete with the latest and most powerful 7th generation Intel chips.

 

Another iPhone 4G Prototype Pops Up In Vietnam, Gets Dissected and Filmed On Video

Another day, another Apple’s fourth-gen iPhone spotted in the wild, this time by the Vietnamese guys from Taoviet.com who’ve managed to purchase a prototype of Cupertino’s future device on the black market for a whopping $4,000. And while Gizmodo’s acquired phone had no capacity numbers behind its back, besides the “XXX” caption, this one, on the other hand, indicates a 16GB of internal storage, housing the usual “Bonfire” testing interface, instead of Apple’s coming OS 4 platform – what a pity. Other than that, it apparently packs the same line of hardware (minus the two external screws) we’ve already seen before with a micro SIM slot stuffed to the right side of the box and these dual silver volume buttons placed to the left, while rear and front-face cameras adorn this “magical” bad boy as well, providing a clear evidence to support future arguments about authentication. Hands-on video after the break.

Update1: The guys at Chipworks claim that the “339S0084” serial number etched on top of the big chip (the one with Apple’s logo) found in Cupertino’s next-gen iPhone during last teardown session, is actually THE Apple A4 chip, fabricated by Samsung as a result of Apple’s design acquisitions of PA Semi and Intrinsity. And if that allegation should be taken seriously, it means that Apple’s fourth generation iPhone will share the same micro-processing capabilities you can find inside the iPad – huzzah!

Update2: We’ve just added two more videos of Apple’s alleged iPhone 4G, posted by the same group of the Vietnamese guys. Check them out after the break.

[via Tinhte]

Continue reading Another iPhone 4G Prototype Pops Up In Vietnam, Gets Dissected and Filmed On Video

AMD To Announce a Break Up

AMD is expected to announce a break up tomorrow morning, due to growing demand for independent, leading-edge foundry production capabilities and increasing R&D costs. According to sources, the giant chipmaker will be splitting up into two companies, as the first one will be responsible for chips designing and the second, for the manufacturing. Additionally, the company has issued a PR release, saying:

“AMD and Advanced Technology Investment Company of Abu Dhabi to Create New Leading-Edge Semiconductor Manufacturing Company. Mubadala increases investment in financially-stronger AMD, which is simultaneously unlocking the value of its manufacturing assets”

The meaning is that Advanced Technology Investment Company will strengthen AMD’s financial position by purchasing 55.6 percent of The Foundry (manufacturing company), for $5.1 billion dollar, and both sides will cooperate together, in order to create a clear stand alone leader in global semiconductor manufacturing, that will be able to equally confront Intel. Stay tuned.

[AMD]