BenchmarkXPRT Blog banner

Author Archives: Bill Catchings

Learning about machine learning

Everywhere we look, machine learning is in the news. It’s driving cars and beating the world’s best Go players. Whether we are aware of it or not, it’s in our lives–understanding our voices and identifying our pictures.

Our goal of being able to measure the performance of hardware and software that does machine learning seems more relevant than ever. Our challenge is to scan the vast landscape that is machine learning, and identify which elements to measure first.

There is a natural temptation to see machine learning as being all about neural networks such as AlexNet and GoogLeNet. However, new innovations appear all the time and lots of important work with more classic machine learning techniques is also underway. (Classic machine learning being anything more than a few years old!) Recursive neural networks used for language translation, reinforcement learning used in robotics, and support vector machine (SVM) learning used in text recognition are just a few examples among the wide array of algorithms to consider.

Creating a benchmark or set of benchmarks to cover all those areas, however, is unlikely to be possible. Certainly, creating such an ambitious tool would take so long that it would be of limited usefulness.

Our current thinking is to begin with a small set of representative algorithms. The challenge, of course, is identifying them. That’s where you come in. What would you like to start with?

We anticipate that the benchmark will focus on the types of inference learning and light training that are likely to occur on edge devices. Extensive training with large datasets takes place in data centers or on systems with extraordinary computing capabilities. We’re interested in use cases that will stress the local processing power of everyday devices.

We are, of course, reaching out to folks in the machine learning field—including those in academia, those who create the underlying hardware and software, and those who make the products that rely on that hardware and software.

What do you think?

Bill

TouchXPRT’s future

If you’ve been following the blog, you know that we’ve been reviewing each part of the XPRT portfolio. If you missed our discussions of HDXPRT, BatteryXPRT, WebXPRT, and CrXPRT, we encourage you to check them out and send us any thoughts you may have. This week, we continue that series by discussing the state of TouchXPRT and what we see down the road for it in 2017.

We released TouchXPRT 2016, an app for evaluating the performance of Windows 10 and Windows 10 Mobile devices, last February. We built the app by porting TouchXPRT 2014 performance workloads to the new Universal Windows App format, which allows a single app package to run on PCs, phones, tablets, and even consoles.

TouchXPRT 2016 installation is quick and easy, and the test completes in under 15 minutes on most devices. The app runs tests based on five everyday tasks (Beautify Photos, Blend Photos, Convert Videos for Sharing, Create Music Podcast, and Create Slideshow from Photos). It measures how long your device takes to complete each task, produces results for each scenario, and gives you an overall score.

As we think about the path forward for TouchXPRT, we’re aware that many expect 2017 to be a year of significant change in the Windows world, with two updates scheduled for release. Microsoft is slated to release the Windows 10 Creators Update (Build 1704) in April, and a subsequent version of Windows codenamed Redstone 3 may arrive this fall. Many tech observers believe that the Creators Update will introduce new creativity and gaming features, along with a UI upgrade named Project NEON. Major foundational shifts in the OS’s structure are more likely to appear with Redstone 3. At this point, quite a lot is still up in the air, but we’ll be following developments closely.

As we learn more about upcoming changes, we’ll have the opportunity to reevaluate TouchXPRT workloads and determine the best way to incorporate new technologies. Virtual reality, 3D, and 4K are especially exciting, but it’s too soon to know how we might incorporate them in a future version of TouchXPRT.

Because TouchXPRT 2016 continues to run well on a wide range of Windows 10 devices, we think it’s best to keep supporting the current version until we get a better idea of what’s in store for Windows.

If you have any thoughts on the future of Windows performance testing, please let us know!

Bill

CES 2017

I’ve attended many tech shows over the years, but this year’s CES has more energy than any I’ve attended in a long time. Part of the energy is the breadth of products. There are amazingly slim TVs that make my TV at home, which I thought was slim, look fat. And, there are beautiful 8K TVs that make my new 4K one feel old.

I’m seeing all manner of smartphones. I’m seeing mobile remote presence devices such as the one from Beam, and after seeing the latest Fenix 5 from Garmin, I think I’ve found my next smartwatch.

Photo Jan 05, 2 05 53 PM

There are many differing devices and approaches to VR and AR. There are drones everywhere. And, lots of massage chairs.

Photo Jan 05, 1 13 17 PM

There are also plenty of products, let’s call them longshots, that contribute to the Wild West feel of the show. Maybe you’d like the Hydreon FakeTV, a small device to make it seem like a TV is on in your house to keep away burglars? Or Dr. Fuji’s Body Shaper, a vibrating platform to “accelerate your workout”?

Another part of the dynamic feeling is the breadth of vendors. Almost all the big tech vendors are here, except for Apple, of course. The excitement for me, however, is the small vendors displaying things that may well never see the light of day, but give glimpses of the future. For example, while I doubt most of the drone vendors at CES will be around in a few years, I think the trend toward small, inexpensive selfie drones will be.

The main reason for the energy at this year’s CES could be the convergence of multiple big industries. The most obvious example of this phenomenon is the large auto presence. Cars have been at CES before—the first time I drove my current car, a BMW i3, was at CES 2014—but this time around they seem to really want to make a statement. Faraday Future is making a splash by trying to be the next Tesla with its FF91.

Photo Jan 05, 3 08 30 PM

Multiple vendors, including VW, Nissan, Toyota, and Mercedes, have concept cars on display—most of them are electric and all of them are heavy on technology. The biggest tech they’re touting is autonomous driving. The auto companies are showing their products while companies like NVIDIA, Intel, and Magic Eye are displaying the tech they have as well.

Regardless of the source of the energy at this CES, I see many opportunities for the existing XPRTs to continue to be important resources. I also see how important emerging technologies like machine learning and VR/AR are going to be and how the XPRTs can be of help there as well.

Exciting times!

Bill

Principled Technologies and the BenchmarkXPRT Development Community announce new effort to evaluate machine learning performance

Durham, NC – Principled Technologies (PT) and the BenchmarkXPRT Development Community, which PT administers, are pleased to announce an initiative to develop a new tool for evaluating systems’ machine learning performance.

Machine learning is a disruptive technology that has the potential to influence a broad range of industries. While there are many available consumer and commercial applications that utilize machine learning for computer vision, natural language processing, and data analytics, there is currently no comprehensive machine learning or deep learning benchmark that includes home, automotive, industrial, and retail use cases.

“These are still the early days of the technology. A fragmented software and hardware landscape and lack of standardization make it complex and challenging to evaluate performance in machine learning,” said Bill Catchings, Co-Founder and Co-Owner of PT. “We’ve decided to take on that challenge, and we invite all interested parties to participate in the creation, evaluation, and testing of this new tool.”

The new tool will join the rest of the industry-standard benchmarks from the BenchmarkXPRT Development Community: WebXPRT, MobileXPRT, TouchXPRT, CrXPRT, BatteryXPRT, and HDXPRT.

To learn more about and join the BenchmarkXPRT Development Community, go to www.BenchmarkXPRT.com.

About Principled Technologies, Inc.

Principled Technologies, Inc. is a leading provider of technology marketing and learning & development services. It administers the BenchmarkXPRT Development Community.

Principled Technologies, Inc. is located in Durham, North Carolina, USA. For more information, please visit www.PrincipledTechnologies.com.

Company Contact

Bill Catchings
Principled Technologies, Inc.
1007 Slater Road, Suite #300
Durham, NC 27703
bcatchings@principledtechnologies.com

Creating a machine-learning benchmark

Recently, we wrote about one of the most exciting emerging technology areas, machine learning, and the question of what role the XPRTs could play in the field.

Experts expect machine learning to be the analytics backbone of the IoT data explosion. It is a disruptive technology with potential to influence a broad range of industries. Consumer and industrial applications that take advantage of machine-learning advancements in computer vision, natural language processing, and data analytics are already available and many more are on the way.

Currently, there is no comprehensive machine-learning or deep-learning benchmark that includes home, automotive, industrial, and retail use cases. The challenge with developing a benchmark for machine learning is that these are still the early days of the technology. A fragmented software and hardware landscape and lack of standardized implementations makes benchmarking machine learning complex and challenging.

Based on the conversations we’ve had over the last few weeks, we’ve decided to take on that challenge. With the community’s help, of course!

As we outlined in a blog entry last month, we will work with interested folks in the community, key vendors, and academia to pull together what we are internally calling MLXPRT.

While the result may differ substantially from the existing XPRTs, we think the need for something is great. Whether that will turn out to be a packaged tool or just sample code and workloads remains to be seen.

What we need most your help. We need both general input about what you would like to see as well as any expertise you may have. Let us know any questions you may have or ways you can help.

On a related note, I’ll be at CES 2017 in Las Vegas during the first week of January. I’d love to meet and talk more about machine learning, benchmarking, or the XPRTs. If you’re planning to be there and would like to connect, let us know.

We will not have a blog entry next week over the holidays, so we wish all of you a wonderful time with your families and a great start to the new year.

Bill

WebXPRT in 2017

Over the last few weeks, we’ve discussed the future of HDXPRT and BatteryXPRT. This week, we’re discussing what’s in store for WebXPRT in 2017.

WebXPRT is our most popular tool. Manufacturers, developers, consumers, and media outlets in more than 350 cities and 57 countries have run WebXPRT over 113,000 times to date. The benchmark runs quickly and simply in most browsers and produces easy-to-understand results that are useful for comparing web browsing performance across a wide variety of devices and browsers. People love the fact that WebXPRT runs on almost any platform that has a web browser, from PCs to phones to game consoles.

More people are using WebXPRT in more places and in more ways than ever before. It’s an unquestioned success, but we think this is a good time to make it even better by beginning work on WebXPRT 2017. Any time change comes to a popular product, there’s a risk that faithful fans will lose the features and functionality they’ve grown to love. Relevant workloads, ease of use, and extensive compatibility have always been the core components of WebXPRT’s success, so we want to reassure users that we’re committed to maintaining all of those in future versions.

Some steps in the WebXPRT 2017 process are straightforward, such as the need to reassess the existing workload lineup and update content to reflect advances in commonly used technologies. Other steps, such as introducing new workloads to test emerging browser technologies, may be tricky to implement, but could offer tremendous value in the months and years ahead.

Are there test scenarios or browser technologies you would like to see in WebXPRT 2017, or tests you think we should get rid of? Please let us know. We want to hear from you and make sure that we’re crafting a performance tool that continues to meet your needs.

Bill

Check out the other XPRTs:

Forgot your password?