Gigabyte Debuts X299X Motherboards for Intel Cascade Lake-X CPUs

Gigabyte today lifted the curtains on the its new X299X Designare 10G, X299X Aaorus Xtreme WaterForce and X299X Aorus Master motherboards fit for Intel’s Cascade Lake-X HEDT CPUs announced Tuesday. Those new chips come with half the pricing of previous-gen Skylake-X models, meaning these new motherboards should get more attention than usual for an HEDT lineup.

Credit: GigabyteCredit: Gigabyte

The trio of high-end motherboards sport the Intel X299X moniker; however, it should be noted that the motherboards are using the same X299 chipset that debuted two years ago. The “X” suffix merely serves to specify that the boards are built specifically for the new Cascade Lake-X processors. Still, Gigabyte’s new X299X offerings feature brand new power designs, thermal solutions and other features, such as Thunderbolt 3 support and 5 Gbps or 10 Gbps network connectivity.

The X299X Designare 10G, X299X Aorus Xtreme WaterForce and X299X Aorus Master are tailored to suit professional users and content creators. The motherboards boast robust 16-phase power delivery subsystems, server-grade components and eight DDR4 memory slots to house up to a whopping 256GB of RAM. They are also built to support multi-GPU configurations and are validated for Nvidia’s Quadro graphics cards.

The X299X Designare 10G and X299X Aorus Xtreme WaterForce have built-in Intel Thunderbolt 3 support, thanks to a pair of USB Type-C ports that deliver a bandwidth of 40 Gbps. Gigabyte X299X motherboards are equipped with built-in PCIe x4 M.2 slots, where users can choose to grab the lanes from the chipset itself or the processor for installing an Intel Optane module or creating a VROC disk array. The X299X Designare 10G and X299X Aorus Xtreme WaterForce even come with a Gen 4 AIC adapter offering up to four extra PCIe x4 M.2 slots for increased storage.

When it comes to network connectivity, Gigabyte X299X motherboards deliver Intel 802.11AX Wi-Fi 6 connectivity. Meanwhile, the X299X Designare 10G employs Intel’s X550-AT2 Ethernet controller and offers dual 10 Gigabit Ethernet ports.

Gigabyte didn’t cheap out on audio. Its X299X motherboards utilize the ALC 1220-VB IC and ESS 9218 Sabre DAC, complemented with the brand’s AMP-UP Audio technology.

As usual, you can expect to see the new Gigabyte X299X motherboards with competent heat dissipation systems, including the company’s Fins-Array heatsink, large diameter heatpipes, thermal pads and thermal backplates. The X299X Aorus Xtreme WaterForce flaunts an AIO cooling design that cools the processor, power delivery subsystem, chipset and M.2 SSDs. All three motherboards come bundled with Gigabyte’s Smart Fan 5 technology, so you have complete control over all the fans and temperature sensors.

The Gigabyte X299X Designare 10G, X299X Aorus Xtreme WaterForce and X299X Aorus Master motherboards should be available to shoppers soon. Gigabyte didn’t specify pricing, so we’ll have to wait until the motherboards hit shelves to find out.

Arduino Science Kit Taps Google to Help Middle Schoolers Learn

Arduino, the open-source hardware and software platform, has partnered with Google to create the Arduino Science Kit Physics Lab. Released today, it’s designed especially for middle schoolers ages 11-14 and features nine physics experiments that explore the topics of thermodynamics, kinetics, electromagnetism and kinematics. 

Two students are meant to share a single kit, which offers a series of plug-and-play projects based on classic carnival rides, such as Electric Fortune Teller and Haunted House Theremin. Data from the whimsically named experiments can be logged onto the Google Science Journal Android app (not Apple iOS yet). It also works with Google Classroom, which lets students share their work with others.

Of course, each kit comes with an Arduino board, in this case the Arduino MKR WiFi 1010. There’s also a carrier board, various cables, a magnet and Velcro accessories. There are also 2 PCB sticks and a PCB encoder, plus M8 screws and bolts, rubber bands a miniature slinky spring and silicone gaskets in two different sizes. Each kit also comes equipped with the necessary sensors, light and temperature, to complete the projects.  

A power source is not included, so users will have to supply their own LiPo battery or power bank to run the board.

The Arduino Science Kit Physics Lab is available now for $199 each, which comes out to $1,990 for a class of 20 students. 

Arduino offers boards and kits for DIY STEM enthusiasts and their children.

Photo Credits: Alberto Morici/Arduino

Bluetooth Flaw Enables Tracking on Windows PCs

Credit: ShutterstockCredit: Shutterstock

Boston University researchers have discovered flaws in the Bluetooth Low Energy (BLE) implementations of Windows 10, macOS and iOS that could allow devices running those operating systems to be tracked. ZDNet reported that the researchers shared details about the flaw at the Privacy Enhancing Technologies Symposium today in an effort to raise awareness for the flaw in a nigh-ubiquitous protocol.

BLE’s name suggests it was developed to make wireless connectivity less of a battery killer. That’s partly true, but it was also made to solve a privacy issue resulting from how Bluetooth connections were formed. Previous versions of the protocol had devices constantly broadcast their media access control (MAC) address; this made it easy to identify and track specific devices that were actively using Bluetooth. BLE changed things up to allow devices to broadcast a “periodically changing, randomized address,” as the researchers described it.

The problem stems from how Microsoft and Apple generate these ostensibly random addresses. The researchers explained in their paper that someone continuously monitoring the “advertisements” BLE uses to seek connections might be able to identify and track specific devices. The advertisements in question don’t have anything to do with promoting a product. Instead, they are the messages BLE devices constantly emit “to announce their presence to other devices,” as the researchers put it. (Should there be a better name for those messages in a time when people fear their privacy is constantly being compromised for someone’s profit? Probably.) Those advertisements are the problem.

The researchers said they developed an “address-carryover algorithm” that “exploits the fact that identifying tokens and the random address do not change in sync” to “continuously track a device, despite implementing anonymization measures.” This exploit doesn’t even require any real hacking–it simply uses publicly broadcast information in ways that Microsoft and Apple didn’t account for in their designs.

This exploit is thought to affect all Windows 10, macOS, and iOS devices. Researchers explained why Android isn’t affected:

“We observed Android advertising addresses to change in intervals of about 15-45 minutes. However, the observed Android smartphones use a completely different advertising approach than Windows or iOS/macOS, making them immune to the address-carryover algorithm. The tested Android phones never send out manufacturer-specific data or other potentially device-identifying data in regular intervals. Instead the OS scans for advertisements of other devices when the Bluetooth settings are opened by the user. Due to the lack of active, continuous advertising, identifying tokens cannot be assembled, making the observed Android devices immune to the carry-over algorithm.”

More information is available in the full paper, “Tracking Anonymized Bluetooth Devices. The researchers said they disclosed these issues to Microsoft and Apple in November 2018; it’s not clear if either company has responded. Neither has publicly acknowledged the issue. We’ve contacted both companies and will update this article if they respond.

ASRock Mutant X570 Board Uses Intel Cooler Mounting Instead of AMD's AM4

Credit: ASRockCredit: ASRock

Despite X570 being on the market for about a month now, it’s just now come to our attention that one of ASRock’s X570 ITX motherboards actually doesn’t support AMD stock coolers. That’s because it uses Intel’s current LGA 1151 mounting mechanism instead of the AM4 mounting mechanism that, to our knowledge, every single other AM4 motherboard uses instead. The X570 Phantom Gaming motherboard advertises its lack of AM4 support on its “special design that uses Intel LGA 115X mounting holes.”

You’re probably asking yourself why ASRock would make this choice because this is a weird thing to implement on an AMD board where you would expect people to use AM4-compatible coolers like AMD’s stock Wraith coolers. ASRock’s list of compatible CPU coolers is mostly made up of coolers that have AM4 brackets, which makes this decision seem nonsensical. However, on that very same list, there are CPU coolers that don’t support AM4, and most of these coolers are low profile, designed specifically for mini-ITX.

Although it’s been two years since Ryzen came out and AMD brought competition back, AMD still doesn’t quite have the presence Intel has on the motherboard side (though this is clearly changing), and this is especially the case on niche platforms like mini-ITX. Many of the mini-ITX-oriented coolers on the market only support Intel’s current mounting mechanism, so if you want a really good low-profile cooler for your Ryzen 3000 CPU, your options are limited on AM4 sockets. By switching to LGA 1151 (which is also compatible with older coolers for older sockets), ASRock has improved the cooler availability for these CPUs.

Although many companies have issued revisions to support AM4, the best cooler compatibility, old and new, is found on Intel’s mounting mechanism. While there are plenty of good coolers for AM4, even those designed for ITX usage, ASRock’s decision to use Intel mounting makes sense since there are simply more coolers that do support Intel but not AMD.

GlobalFoundries Files Patent Claims Against TSMC, Seeks to Ban Imports of Nvidia, Apple Chips

GlobalFoundries today announced that it’s filed lawsuits against Taiwan Semiconductor Manufacturing Company (TSMC) in the U.S. and Germany over the latter company’s alleged infringement of 16 patents. The company said that it’s looking to halt the import of processors made with the technologies as well as seeking “significant damages from TSMC based on TSMC’s unlawful use of GF’s proprietary technology in its tens of billions of dollars of sales.”

Note that GlobalFoundries said it wants to stop the import of processors made with the technologies it believes are covered by its patents. The company recognized that TSMC doesn’t usually import those processors into the U.S. or Germany; its customers do. That means the lawsuits could affect much of the tech industry: TSMC said that in 2018 it was “manufacturing 10,436 different products using 261 distinct technologies for 481 different customers.”

The list of companies supplied by TSMC includes AMD, Nvidia, Apple, Mediatek and many others, which means that GlobalFoundries could bring the tech industry to a halt if it’s allowed to stop imports to the U.S. and Germany. 

The lawsuits were filed with the U.S. International Trade Commission, the U.S. Federal District Courts in the Districts of Delaware and the Western District of Texas and the Regional Courts of Dusseldorf and Mannheim in Germany. GlobalFoundries leaned hard on TSMC being headquartered in Taiwan in its press release, effectively portraying the dispute as an Eastern company profiting off the innovations of its Western competitor.

That positioning was made clear in the following statement that was provided with the release:

“While semiconductor manufacturing has continued to shift to Asia, GF has bucked the trend by investing heavily in the American and European semiconductor industries, spending more than $15 billion dollars in the last decade in the U.S. and more than $6 billion in Europe’s largest semiconductor manufacturing fabrication facility. These lawsuits are aimed at protecting those investments and the US and European-based innovation that powers them,” said Gregg Bartlett, senior vice president, engineering and technology at GF. “For years, while we have been devoting billions of dollars to domestic research and development, TSMC has been unlawfully reaping the benefits of our investments. This action is critical to halt Taiwan Semiconductor’s unlawful use of our vital assets and to safeguard the American and European manufacturing base.” 

TSMC hasn’t publicly commented on GlobalFoundries’ lawsuits at time of writing. We’ve reached out to the company for a statement and will update this post if it responds. The company recently published its first blog post–in which it proclaimed that Moore’s Law isn’t dead–so responding to multiple lawsuits from GlobalFoundries would make a heck of a second entry. But we suspect the company will restrain itself to press releases and court filings.

GlobalFoundries announced in August 2018 that it would cease development of its 7nm process to focus on being a “specialty foundry.” The company recently showed off 3D chips it developed alongside Arm to enable a “new-level of system performance and power efficiency for computing applications, such as AI/ML [artificial intelligence and machine learning] and high-end consumer mobile and wireless solutions.” We said at the time that larger foundries–including TSMC–would likely follow. It seems that GlobalFoundries believes TSMC has been following its new technologies for a while.

India's First Three (Out of 60) Supercomputers Are Here

India’s Department of Science and Technology (DST) Secretary Ashutosh Sharma announced that two more supercomputers, which are part of the Indian government’s National Supercomputing Mission roadmap, will become functional in a few days. As reported by Indian newspaper Livemint, India’s first supercomputer from the project is already functional at the Indian Intitute of Technology (IIT) Banaras Hindu University.

Param Shavak, India's first supercomputer Credit: CDACParam Shavak, India’s first supercomputer Credit: CDAC

The National Supercomputing Mission aims to build 60 supercomputers, mostly with local technology, by 2022. The project is split into three main phases.

Phase one of the project includes integration and assembly at the motherboard level for the built supercomputers, according to Sharma. There are six supercomputers currently planned for phase one. 

The second phase will begin next year and see the building of 10 more supercomputers, with most of the assembly and integration done in India.

In phase three, the rest of the supercomputers should be fully built in India with the exception of the processors, which will continue to be outsourced.

Sharma told reporters that the National Supercomputing Mission will cost Rs 4,500 crore ($633 million). The project is meant to boost Indian research, as well as national security, according to Sharma.

Milind Kulkarni, a senior scientist at DST, gave a few more details about the goal of the new supercomputers, which will create a National Knowledge Network between various research-focused universities:

“They would help improve weather services, disaster management, ensure faster processing of data, support computational biology, flood control and aid research in various disciplines,” Kulkarni said. 

He added that as many as 3,000 people were already trained to use the supercomputers, including from chemistry, physics, biology and computer programming backgrounds. 

“So wherever they have been installed, they would be connected to all other institutions through the National Knowledge Network. So, if your university is part of the network, then anyone sitting on a desktop can make use of it, by seeking time to do the related high computing work,” Kulkarni said.

Even though the current National Supercomputing Mission roadmap doesn’t include using processors made in India yet, the Indian government launched its very first RISC-V processor, Shakti, recently. 

Going from building an experimental low-powered processor to building a supercomputer-grade CPU core should take some time. However, it’s clear that India has its eyes set on a super technological future. 

AMD Ryzen 9 3950X Hits 4.3 GHz On All 16 Cores With Water Cooling

AMD’s 16-core 32-thread Ryzen 9 3950X may have hit a small snag on its way to market, but the enthusiast world is eagerly awaiting the arrival of a new record-setting core count for the mainstream desktop. And according to Gigabyte, overclockers may have even more reason to be excited than “just” a generous helping of cores and threads. 

Gigabyte has released a Ryzen 9 3950X overclocking guide where the motherboard manufacturer was able to push its sample to an impressive 4.3 GHz on all 16 cores while using an AIO liquid cooler, and at a mere ~1.4V. The company accomplished this feat with a beefy watercooling setup.

Credit: Tom's HardwareCredit: Tom’s Hardware

*Ryzen 9 3950X benchmarks from Gigabyte, other results from Tom’s Hardware labs

Gigabyte paired its Ryzen 9 3950X with the brand’s own X570 Aorus Master motherboard, Aorus 16GB (2x8GB) DDR4-3200 memory kit and EKWB’s EK-KIT P360 liquid cooling kit. The motherboard manufacturer used Cinebench R15 to evaluate the 16-core chip’s stock and overclocked performance, which allows us to compare those results to ours.

At stock, the Ryzen 9 3950X scored 3,932 points in Cinebench R15, which is 92.4% faster than a stock Core i9-9900K and 81% faster than the Core i9-9900K overclocked to 5 GHz on all cores (which would be similar to a stock Core i9-9900KS). For now, the Core i9 series from Intel stands as its most powerful competing chip on a mainstream platform, and that isn’t likely to change soon.

At stock settings, AMD’s upcoming flagship also delivers up to 25.5% more performance than the existing Ryzen 9 3900X. As you can see above, the results are even more impressive once Gigabyte pushed the chip to 4.3 GHz.

The fact that the Ryzen 9 3950X could hit 4.3 GHz on all its cores is a great achievement, especially when Ryzen 3000-series processors are famous for not having much manual overclocking headroom. For comparison, our Ryzen 9 3900X sample, which has four fewer cores, maxes out at 4.1 GHz. That implies that AMD is setting aside the absolute best 7nm dies for its 16-core, 32-thread chip, especially given that Gigabyte hit a Prime95-stable (one hour run) 4.3 GHz with only 1.4 vCore. Gigabyte’s Ryzen 9 3950X sample could even hit a devastating 4.4 GHz, but the company insinuated that it was only stable enough to pass a Cinebench R15 run.

Credit: Tom's HardwareCredit: Tom’s Hardware

There are also a few other interesting takeways from Gigabyte’s Ryzen 9 3950X overclocking guide. For starters, Gigabyte states that 1.45V is the maximum safe voltage. The value might be too high for everyone’s taste, and keeping it around 1.4V sounds more reasonable. In terms of thermals, Gigabyte noted that the Ryzen 9 3950X’s operating temperatures are right in the same ballpark as last year’s Ryzen 7 2700X.

That’s pretty remarkable, so we’ll say it again: According to Gigabyte, the 16-core Ryzen 9 3950X is as easy to cool as an 8-core part.

SEP (USD)
Cores / Threads
TDP (Watts)
Base Frequency (GHz)
Boost Frequency (GHz)
Total Cache (MB)
PCIe 4.0 Lanes (Processor / Chipset)
Launch Date
Ryzen 9 3950X
$749
16 / 32
105W
3.5
4.7
72
24 / 16
September
Ryzen 9 3900X $499 12 / 24 105W 3.8 4.6 70 24 / 16 July 7, 2019
Ryzen 7 3800X $399 8 / 16 105W 3.9 4.5 36 24 / 16 July 7, 2019
Ryzen 7 3700X $329 8 / 16 65W 3.6 4.4 36 24 / 16 July 7, 2019
Ryzen 5 3600X $249 6 / 12 95W 3.8 4.4 35 24 / 16 July 7, 2019
Ryzen 5 3600 $199 6 / 12 65W 3.6 4.2 35 24 / 16 July 7, 2019

The $749 Ryzen 9 3950X, which originally was scheduled to launch last month, has been pushed to November. Let’s keep our fingers crossed that AMD doesn’t have any more setbacks.