Graphics processing unit

AMD’s Radeon GPUs are rare because they’re good at mining bitcoin copycats

Bitcoin and other cryptocurrencies may enable online threats like ransomware to spread, but they are also fueling demand for powerful and efficient new graphics cards.

It’s difficult to find an RX 500-series Radeon graphics card at retail or online right now in part because people are seeking them out to mine certain bitcoin-like cryptocurrencies, chip manufacturer AMD explained to CNBC. In April, AMD released its latest round of Polaris-powered GPUs which can render graphically intensive scenes without drawing excessive amounts of power. That combination of number-crunching capabilities and energy efficiency has attracted consumers who want to “mine” altcoin cryptocurrencies, which are alternatives to bitcoin.

“The gaming market…

AMD’s Ryzen Threadripper with 4 Radeon Vega GPUs runs Prey at 4K

Advanced Micro Devices announced its Ryzen Threadripper processor today, as well as its new Radeon Vega Frontier Edition graphics processing units (GPUs). And the Sunnyvale, Calif.-based company said that these chips are going to make gamers very happy.

In a demo at the Computex trade show in Taiwan, AMD executive Jim Anderson showed off the new flagship central processing unit (CPU), the Ryzen Threadripper. The processor is based on AMD’s Zen architecture, which can process 52 percent more instructions per clock cycle compared to AMD’s older CPUs.

“Ryzen Threadripper is targeted at the highest performance PC systems,” Anderson said.

The Threadripper is coming this summer, and it has 16 cores capable of running 32 threads. It supports 64 lanes of PCIe 3.0 devices, such as graphics cards. It has four channels of DDR4 memory. It’s about twice as fast as the 8-core Ryzen 7 desktop processor that…

How Nvidia’s Max-Q delivers 3 times the performance in a much thinner laptop

Last night, Nvidia CEO Jen-Hsun Huang announced Max-Q, a new approach to designing gaming laptops to deliver three times the performance in a third of the size of previous laptops. In doing so, he held up a 10-pound gaming laptop from a few years ago and compared it to a lighter and thinner laptop from Asus.

And today, Nvidia’s Mark Aevermann, director of notebook product management, further explained how Max-Q makes that possible. He noted that five years ago, efficient mainstream gaming laptops were just a dream, and the total market was about 200,000 systems a year. It was a niche market. In 2016, more than 10 million gaming laptops shipped, or more than the number of Xbox One game consoles sold.

“Gaming notebooks are a force today, and they’re the fastest-growing platform out there,” Aevermann said.

But gaming notebooks could still be lighter and more affordable. So, in partnership with original equipment manufacturers (OEMs), Nvidia set about to redesign notebooks to be more efficient, without changing the graphics chips that ship with them.

Above: Nvidia’s Max-Q improvements.

Image Credit: Nvidia

In one of the biggest changes, Nvidia figured out the peak efficiency for its graphics processing units. It found that running chips at peak performance cost a huge amount in terms of power consumption. By throttling back just a little, the power efficiency became much more manageable.

“Peak performance is not peak efficiency,” Aevermann said. “The last little bits of performance cost a tremendous amount of power.”

Nvidia also looked at the optimal game settings, advanced thermal solutions and software drivers that make use of them, better acoustics, and efficient power regulation.

Max-Q is a term taken from NASA’s mission to launch man into space. It is defined as the point at which the aerodynamic stress on a rocket in atmospheric flight is maximized. Thus, the design of the rocket is precision-engineered around Max-Q.


Nvidia’s Max-Q tech can make gaming laptops as thin as a MacBook Air

Graphics chip maker Nvidia introduced Max-Q, a new design approach that will enable thinner, quieter, and faster gaming laptops.

Laptops using the new Max-Q designs will debut from all major computer makers starting June 27, according to an announcement by CEO Jen-Hsun Huang in a speech at the Computex trade show in Taiwan. Max-Q is a new approach to designing gaming laptops, and Nvidia is working with original equipment manufacturers (OEMs) and system builders to make high-end gaming laptops thinner, faster, and quieter. With Max-Q, everything in the design is precision-engineered — including the laptop, the graphics processing unit (GPU), the drivers, and the thermal and electrical components — to ensure peak efficiency.

Max-Q is a term taken from NASA’s mission to launch man into space. It is defined as the point at which the aerodynamic stress on a rocket in atmospheric flight is maximized. Thus, the design of the rocket is precision-engineered around Max-Q. Nvidia said it has applied a similar philosophy to designing gaming laptops, enabling original equipment manufacturers (OEMs) to build laptops that are three times thinner with up to three times more performance of previous generation products.

The results of Max-Q technology, as applied to existing chips, is a high-performance gaming platform…

ARM wants to boost AI performance by 50X over 5 years

ARM is unveiling its first Dynamiq processor designs today, and the company said that the family will boost artificial intelligence performance by more than 50 times over the next three to five years.

The new family aims to spread AI processing from the edge to the cloud. The processors include the ARM Cortex-A75, which delivers massive single-thread compute performance at the high end; the ARM Cortex-A55, a high-efficiency processor; and the ARM Mali-G72 graphics processor, which expands the possibilities for virtual reality, gaming, and machine learning on premium mobile devices, with 40 percent more graphics performance. ARM’s partners are expected to launch chips in 2018.

To better handle AI processing, ARM realized that it needs to make basic changes to the computing architecture, with faster, more efficient, and distributed intelligence between computing at the edge of the network (like in smartphones and laptops) and in the cloud-connected data centers, said Nandan Nayampally, vice president and general manager of the Compute Products Group at ARM, in a blog post.

That AI technology also needs to be secure, as recent survey data shows 85 percent of global consumers are concerned about securing AI…

Why the AI hype cycle won’t end anytime soon

By now, we’ve all heard the hype around AI. According to Google Trends, only 5 percent of the U.S. population searched for information about artificial intelligence in 2012. In 2017, that figure has jumped to an estimated 60 percent. Unlike some other fads that have swept through the tech industry, however, the hype around AI is justified for a number of reasons. AI development will continue to exponentially infiltrate our day-to-day lives.

Unlike in the early days of AI, there are now many useful and powerful frameworks — like TensorFlow and Caffe — that enable easy implementation of AI technologies and remove much of the need for engineers to build code from scratch. These frameworks eliminate considerable time and resources and will continue to make AI more widespread and available to companies of all sizes and across industries.

There are also many pre-trained neural networks that are available for common use in different domains, like content, image, and voice recognition. These networks are also greatly contributing to the growth of entrepreneurial endeavors surrounding AI by allowing the use of pre-prepared neural network models which can be trained on an exclusive dataset. Some of the most trailblazing ones include Yolo, fasText, and Deep Speech 2. These initiatives are allowing AI to flourish at an exponential rate.

Increasingly affordable AI maintenance and the increased speed of calculations thanks to GPU are significant factors in the unbridled growth of AI. Maintenance is becoming more and more inexpensive as cloud providers like Amazon continue to mark down the costs of their services. The efficiency…

Nvidia CEO: ‘AI is going to infuse all of software’

Jen-Hsun Huang is a big fan of artificial intelligence, as it helps his company Nvidia sell a lot more AI chips.

In an earnings call yesterday, the CEO responded to a question by saying, “AI is going to infuse all of software.” He’ll talk more about this topic today at the Nvidia GPU Tech conference in San Jose, Calif., where he is delivering a keynote speech. The event draws about 7,000 people, many for talks on AI. Nvidia also said yesterday it plans to train 100,000 developers this year on deep learning technology, which is one form of AI that is delivering rapid advances across a variety of industries.

Huang wasn’t the only one singing AI’s praises during the call.

“AI has quickly emerged as the single most powerful source of technology,” said Colette Kress, chief financial officer at Nvidia, during the call. “And at the center of AI are Nvidia GPUs.”

Above: Nvidia Metropolis will use video analytics to monitor public safety.

Image Credit: Nvidia

Here’s what Huang had to say during the conference call.

First of all, AI… teams up with Nvidia to take machine learning to the enterprise and Nvidia today announced that they have partnered to take machine learning and deep learning algorithms to the enterprise through deals with Nvidia’s graphics processing units (GPUs).

Mountain View, Calif.-based has created AI software that enables customers to train machine learning and deep learning models up to 75 times faster than conventional central processing unit (CPU) solutions. The company made the announcement at Nvidia’s GPU Tech event in San Jose, Calif. will offer its machine learning algorithms in a newly minted GPU-edition and its Deep Water product on Nvidia GPUs. In addition,’s platform will now be optimized for the Nvidia’s DGX-1 AI processor.

Enterprises can use this end-to-end solution to operate on large data sets, iterate faster, deploy quickly, and gain real-time…

Nvidia Metropolis video analytics paves the way for AI cities

In a city of the future, it would be nice to know quickly if there’s a fire burning out of control, a crime in progress at a certain location, or a traffic snarl at a particular corner.

Nvidia hopes to detect such problems in smart cities using Nvidia Metropolis, which the company said could pave the way for the creation of smart artificial intelligence cities. Nvidia announced the tech ahead of its GPU Technology conference this week in San Jose, California.

Metropolis is a video analytics platform that applies deep learning AI to video streams for applications such as public safety, traffic management, and resource optimization.

Nvidia said that Metropolis could make cities safer, and more than 50 partner companies are already providing products and applications for AI city uses based on graphics processing units (GPUs) made by Nvidia.

“Deep learning is enabling powerful intelligent video analytics that turn anonymized video into real-time valuable insights, enhancing safety and improving lives,” said Deepu…

What All of Your Computer’s Specs Really Mean

Computer specs can be a baffling mix of acronyms and numbers at the best of times, but it’s worth learning something about them: It’ll help you choose a new computer, troubleshoot your old computer, and generally understand more about the relationship between the specs on the page and the experience you’re getting.

Such is the complexity of the modern-day computer, we could’ve written an article twice this size on any one of the categories listed below (look at any graphics card forum for proof)—but the main aim here is to help you understand the specs you see listed with desktops and laptops, and give you an idea of the difference they make to performance.


The Central Processing Unit, or CPU, or processor, is the brains of the operation: it handles all those calculations that keep your computer actually working. The CPU inside your machine is the main (but not the only) contributor to its overall speed and performance.

CPUs have a certain number of cores, mini computing units that are effectively CPUs in their own right—they let your computer work on multiple tasks at the same time, so the more cores the better. On top of this, each core has a clock speed, a measurement of how fast it can do its number crunching, usually measured in gigahertz (GHz).

Comparing the performance of CPUs based on core number and clock speeds is notoriously difficult (sorry shoppers). That’s because multiple factors are involved, most related to the microarchitecture of the CPUs. The microarchitecture is basically the way that the cores and the other bits of a CPU are packed together.

The two big computer CPU makers, Intel and AMD, have their own microarchitecture designs. When you see references to Intel Skylake, Intel Kaby Lake, or AMD Zen (on Ryzen chips), this is what’s being referred to, and newer is always better as successive microarchitectures allow the CPU to work faster and more efficiently (and use less power).

Intel and AMD also apply their own labels—i3, i5, and i7 in Intel’s case—to indicate relative performance within a microarchitecture family. It’s a useful shorthand reference to the power you can expect, with i7 CPUs the best of the bunch from Intel. In AMD’s case, you’re talking about Ryzen 3, Ryzen 5 and the top-end Ryzen 7.

If you want the very best processors around, you should also look out for what Intel calls hyper-threading and what AMD calls simultaneous multi-threading. These technologies effectively double the number of cores (virtually, not physically) so you’ve got significantly improved performance for demanding applications like video editing or CAD software.

Unless you’re building your own PC from scratch, that’s probably all you need to know when looking at system listings, but CPUs have numerous other specs, including the amount of high-speed memory cache and the extra graphics processing capabilities that are on board. If your CPU has enough integrated graphics oomph, you don’t need a separate card or chipset, of which more below.


The other big factor in computer performance, particularly if you’re gaming or working with a lot of video and images, is graphics.

We only gave it a brief mention in the processor section, but many Intel CPUs now come with a decent amount of graphics processing power built in, enough for most users to get by with a bit of web browsing, Twittering, essay writing and even light image editing and gaming. You can also get integrated graphics chipsets built into the motherboard…