Connect with us

Gadgets

Stanford’s Doggo is a petite robotic quadruped you can (maybe) build yourself – TechCrunch

Published

on

Got a few thousand bucks and a good deal of engineering expertise? You’re in luck: Stanford students have created a quadrupedal robot platform called Doggo that you can build with off-the-shelf parts and a considerable amount of elbow grease. That’s better than the alternatives, which generally require a hundred grand and a government-sponsored lab.

Due to be presented (paper on arXiv here) at the IEEE International Conference on Robots and Automation, Doggo is the result of research by the Stanford Robotics Club, specifically the Extreme Mobility team. The idea was to make a modern quadrupedal platform that others could build and test on, but keep costs and custom parts to a minimum.

The result is a cute little bot with rigid-looking but surprisingly compliant polygonal legs that has a jaunty, bouncy little walk and can leap more than three feet in the air. There are no physical springs or shocks involved, but by sampling the forces on the legs 8,000 times per second and responding as quickly, the motors can act like virtual springs.

It’s limited in its autonomy, but that’s because it’s built to move, not to see and understand the world around it. That is, however, something you, dear reader, could work on. Because it’s relatively cheap and doesn’t involve some exotic motor or proprietary parts, it could be a good basis for research at other robotics departments. You can see the designs and parts necessary to build your own Doggo right here.

“We had seen these other quadruped robots used in research, but they weren’t something that you could bring into your own lab and use for your own projects,” said Doggo lead Nathan Kau in a Stanford news post. “We wanted Stanford Doggo to be this open source robot that you could build yourself on a relatively small budget.”

In the meantime, the Extreme Mobility team will be both improving on the capabilities of Doggo by collaborating with the university’s Robotic Exploration Lab, and also working on a similar robot but twice the size — Woofer.

Source link

Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published.

Gadgets

Intel’s 13th-gen “Raptor Lake” CPUs are official, launch October 20

Published

on

Enlarge / An overview of the improvements coming to Intel’s 13th-gen desktop chips.

Intel

If there’s one thing Intel has gotten good at in the last few years, it’s refining a CPU architecture. Between 2015 and 2020, manufacturing troubles pushed Intel to release not one, not two, but five processor generations based on iterations of the sixth-gen Skylake core, while still managing to increase clock speeds and core counts enough to stay competitive through most of that timespan.

It’s an approach Intel is returning to for its 13th-generation Core CPUs, the first of which are being officially announced today. Codenamed Raptor Lake, Intel says it has made some improvements to the CPU architecture and the Intel 7 manufacturing process, but the strategy for improving their performance is both time-tested and easy to understand: add more cores, and make them run at higher clock speeds.

Intel is announcing three new CPUs today, each with and without integrated graphics (per usual, the models with no GPUs have an “F” at the end): the Core i9-13900K, Core i7-13700K, and Core i5-13600K will launch on October 20 alongside new Z790 chipsets and motherboards. They will also work in all current-generation 600-series motherboards as long as your motherboard maker has provided a BIOS update, and will continue to support both DDR4 and DDR5 memory.

Raptor Lake uses the hybrid architecture that Intel introduced in its 12th-generation Alder Lake chips last year—a combination of large performance cores (P-cores) that keep games and other performance-sensitive applications running quickly, plus clusters of smaller efficiency cores (E-cores) that use less power—though in our testing across laptops and desktops, it’s clear that “efficiency” is more about the number of cores can be fit into a given area on a CPU die, and less about lower overall system power consumption.

There have been a handful of other additions as well. The amount of L2 cache per core has been nearly doubled, going from 1.25MB to 2MB per P-core and from 2MB to 4MB per E-core cluster (E-cores always come in clusters of four). The CPUs will officially support DDR5-5600 RAM, up from a current maximum of DDR5-4800, though that DDR5-4800 maximum can easily be surpassed with XMP memory kits in 12th-generation motherboards.

The maximum officially supported DDR4 RAM speed remains DDR4-3200, though the caveat about XMP applies there as well.

Extra cache, faster memory speeds, and boosted clock speed are responsible for the single-threaded performance gains for the i9-13900K. Additional E-cores make the multi-core improvements much more significant.
Enlarge / Extra cache, faster memory speeds, and boosted clock speed are responsible for the single-threaded performance gains for the i9-13900K. Additional E-cores make the multi-core improvements much more significant.

Intel

As far as core counts and frequencies go, the Core i5 and Core i7 CPUs each pick up one extra E-core cluster, going from four E-cores to eight. The Core i9 gets two new E-core clusters, boosting the core count from eight all the way up to 16. All E-cores have maximum boost clocks that are 400MHz higher than they were before. P-core count stays the same across the lineup, but the maximum boost clock has been increased by 600MHz, 400MHz, and 200MHz for the Core i9, i7, and i5, respectively. As K-series chips, these are all unlocked for overclocking when used with Z690 or Z790 motherboards.

Launch pricing is going up by $30 for the Core i5 models, but staying level for the other two. As usual, Intel doesn’t include any CPU coolers with K- or KF-series chips. Here’s how each CPU stacks up to its predecessor:

CPU Launch MSRP P/E-cores Clocks (Base/Boost) Total cache (L2+L3) Base/Max Power
Core i9-13900K $589
$564 (F)
8P/16E 3.0/5.8 GHz (P)
2.2/4.3 GHz (E)
68MB (32 + 36) 125/253 W
Core i9-12900K $589
$564 (F)
8P/8E 3.2/5.2 GHz (P)
2.4/3.9 GHz (E)
34MB (14 + 30) 125/241 W
Core i7-13700K $409
$384 (F)
8P/8E 3.4/5.4 GHz (P)
2.5/4.2 GHz (E)
54 MB (24 + 30) 125/253 W
Core i7-12700K $490
$384 (F)
8P/4E 3.6/5.0 GHz (P)
2.7/3.8 GHz (E)
37 MB (12 + 25) 125/190 W
Core i5-13600K $319
$294 (F)
6P/8E 3.5/5.1 GHz (P)
2.6/3.9 GHz (E)
44 MB (24 + 20) 125/181 W
Core i5-12600K $289
$264 (F)
6P/4E 3.7/4.9 GHz (P)
2.8/3.6 GHz (E)
29.5 MB (9.5 + 20) 125/150 W

According to Intel, all of the changes together will boost the i9-13900K’s single-threaded performance by around 15 percent, with most of the improvement attributable to P-core clock speed increases. That’s short of the 29 percent AMD accomplished across the lineup with its Zen 4 chips, and it will be lower for the i7 and the i5. But it’s reasonably respectable for a year-over-year increase. Multi-threaded performance is where you’ll see the biggest gains, with the added cache, boosted clock speeds, and increased E-core counts all coming together to improve the i9-13900K’s performance by 41 percent compared to the i9-12900K (though, again, that number may be less impressive for the i7 and i5).

Since the manufacturing process is, at best, only improving modestly, the price you’ll pay for the extra clock speed and core counts is higher power usage. Intel is keeping the base power of these 13th-gen CPUs unchanged at 125 W, but the Maximum Turbo Power numbers have gone up quite a bit—the Core i9-13900K’s 253 W maximum is the maximum amount of power officially supported by the LGA1700 socket, though it’s possible that some high-end motherboards could let it go even higher.

The i9-13900K can be quite power-efficient compared to the i9-12900K, though its default configuration allows for higher power use overall.
Enlarge / The i9-13900K can be quite power-efficient compared to the i9-12900K, though its default configuration allows for higher power use overall.

Intel

But that doesn’t mean Intel is throwing power efficiency entirely out the window, either. When restricted to a 65 W base power, Intel says that the improvements to Raptor Lake will allow the chips to run multi-threaded workloads just as quickly as a Core i9-12900K running at 241 W. As has become the norm for these kinds of high-end parts, they will default to fast performance with high power usage, but users can rein them in if they want.

As for the accompanying Z790 chipset, it has a few improvements over the previous-generation Z690, but is likely nothing worth upgrading for if you’re already using a 600-series motherboard you like. The chipset now sports a total of 20 PCIe 4.0 lanes for SSDs and other accessories, plus eight PCIe 8.0 lanes—Z690 has 12 PCIe 4.0 lanes and 16 PCIe 3.0 lanes, so clearly Intel is just shifting the balance in the direction of the faster interconnect. Z790 also supports one additional 20Gbps USB 3.2 Gen 2×2 port, for a total of five, and removes support for basic USB 2.0 ports entirely. The platform’s PCIe 5.0 lanes for GPUs and next-gen SSDs are still built into the processor, not the chipset itself.

The Z790 chipset has more PCIe 4.0 lanes and 20Gbps USB ports than Z690. But if you have a 600-series motherboard you already like, it will probably get a BIOS update allowing it to support 13th-gen CPUs, and none of the improvements to Z790 are worth an additional purchase.
Enlarge / The Z790 chipset has more PCIe 4.0 lanes and 20Gbps USB ports than Z690. But if you have a 600-series motherboard you already like, it will probably get a BIOS update allowing it to support 13th-gen CPUs, and none of the improvements to Z790 are worth an additional purchase.

Intel

Intel didn’t announce any other 13th-generation CPU models today, but it teased that the standard range of chips would be following in the coming months—lower-wattage, lower-cost desktop parts, as well as laptop CPUs designed for everything from thin-and-light ultrabooks to bulky LED-festooned gaming laptops. Intel says that we can expect other desktop CPUs in the lineup to get more E-cores, too, something that previous rumors had already suggested. We’d expect to learn more about these chips at CES in January.

Continue Reading

Gadgets

Alienware QD-OLED monitor picks open standards over G-Sync, is $200 cheaper

Published

on

Enlarge / Alienware’s latest QD-OLED monitor, the AW3423DWF.

Alienware

Alienware announced today a new QD-OLED monitor SKU that looks awfully similar to the Alienware AW3423DW released for $1,300 this spring. The AW3423DWF has many of the same specs but skips Nvidia G-Sync certification and hardware in favor of AMD’s and VESA’s open standards for fighting screen tears, while costing $200 less than its predecessor.

Like the AW3423DW, the AW3423DWF uses QD-OLED technology from Samsung. This is a form of OLED that uses a blue self-emitting layer as its light source, which goes through a layer of quantum dots. The primary goal is better color coverage, including more consistent colors across brightness levels, combined with the deep blacks and incredible contrast for which OLED displays are known.

The 34.18-inch AW3423DWF and AW3423DW’s specs sheets match closely, including 3440×1440 resolution, an 1800R curve, 99.3 percent DCI-P3 and 149 percent sRGB color coverage, up to a 165 Hz refresh rate via DisplayPort and 100 Hz via HDMI 2.0, and 0.1ms gray-to-gray (GtG) response time.

But while the AW3423DW uses G-Sync Ultimate, which confirms 1,000 nits’ brightness with HDR in addition to fighting screen tears when paired with an Nvidia GPU, the new AW3423DWF uses AMD’s FreeSync Premium Pro and VESA’s AdaptiveSync standards.

There's ventilation around the back for heat dissipation.
Enlarge / There’s ventilation around the back for heat dissipation.

Alienware

AMD’s FreeSync Premium Pro works with AMD GPUs, including those within the latest Xbox and PlayStation consoles. The Premium Pro qualifier confirms the feature works with HDR (the monitor is VESA DisplayHDR TrueBlack400-certified) and adds low frame rate compensation, which displays frames multiple times to make up for when frame rates fall under the monitor’s lowest supported refresh rate.

VESA announced its AdaptiveSync certification program in May. It includes testing for judder, dropped frames, and more strenuous GtG response time testing that looks at 20 GtG combinations.

VESA’s AdaptiveSync requires functionality with refresh rates from at least 60–144 Hz and a 5 ms GtG response time as per VESA’s testing. Support with Nvidia, AMD, and Intel graphics cards, meanwhile, simplifies the compatibility question.

Even before VESA announced its AdaptiveSync program, the lines between G-Sync and FreeSync were already diminishing somewhat, with many FreeSync monitors also being able to run G-Sync without Nvidia hardware and, thus, maintain lower prices.

By skipping Nvidia certification or using the G-Sync module, Alienware has a $200 cheaper QD-OLED monitor with largely the same performance expectations. When VESA announced the AdaptiveSync performance tier (there’s also a MediaSync one), the industry group acknowledged to Ars Technica that while this isn’t its goal, the tiers could lead to the end of GPU-specific flavors of variable refresh rates (VRR).

Alienware also threw in a couple of extra features with the new AW3423DWF, including purportedly improved cable management and a five-way joystick that can bring up different image modes, including a new Creator Mode that switches the color gamut to sRGB and lets you adjust gamma. However, at 149 percent sRGB color coverage, you’ll need to do significant calibration for color-critical work in this space.

Alienware also claims the new monitor will be easier to mount than the prior one, thanks to a slimmer profile. The AW3423DWF is 5 inches deep without the stand versus the AW3423DW’s 5.4 inches.

A peek at port selection, which includes one HDMI 2.0 port and two DisplayPort 1.4 ports.
Enlarge / A peek at port selection, which includes one HDMI 2.0 port and two DisplayPort 1.4 ports.

Alienware

With a 21:9 aspect ratio that lends to black bars when console gaming at 16:9 and HDMI 2.0 connectivity maxing out at 100 Hz at the monitors’ max resolutions, neither QD-OLED display is the perfect fit for modern console gaming. However, the display should support VRR at up to 120 Hz on the Xbox Series X/S and PlayStation 5 at 2560×1440 resolutions.

Alienware will start accepting orders for the AW3423DWF this fall.

Ars Technica may earn compensation for sales from links on this post through affiliate programs.

Editor’s note: This article originally stated that orders for the monitor would start on a specific date, but an Alienware representative reached out to Ars, saying that the company’s no longer providing a specific date for taking orders. 

Continue Reading

Gadgets

The Roomba j7+ learns to mop with a dramatic swing-arm setup

Published

on

iRobot—soon to be owned by Amazon—is announcing a flagship Roomba with a new feature: It can vacuum and mop simultaneously. Meet the Roomba Combo j7+, a $1,100 combo cleaning robot that ships on October 4.

iRobot is not doing a ground-up redesign of the j7+ series to add mop functionality. In fact, the update almost looks like a retrofit. The new j7 looks just like the old j7 with a camera in the front, a big dust bin in the back, and a bottom layout that is almost identical to the old bot. There’s a new dust bin and… is that a rear spoiler?

The mop functionality lives on the top (yes, the top) of the j7+, which has a big rear cutout now. The top of this cutout is plastic, and the bottom is the wet mop pad, which is connected to the robot by two side arms. When it’s time to do some mopping, a dramatic, Transformers-like transition occurs. Two flaps on the side of the Roomba open up, revealing that the top mop cutout is actually connected to the robot by a pair of swing arms. The cutout section on top of the robot is lifted up and swings down and under the robot in a big, 180-degree motion. Now you’re dragging a wet mop pad across the floor with minimal changes to the layout of the j7+.

It doesn’t sound like the j7+ will provide iRobot’s top-of-the-line mopping experience, which will still belong to the purpose-built Braava Jet M6. iRobot’s dedicated mop bot beats the j7+ because about 50 percent of the Braava’s body is a water tank, and it has a front water sprayer. The j7+ has a redesigned dust bin that is now a dust and water combo tank. The dust area is smaller to make way for a 210ml water tank. That means way less dust capacity, but that doesn’t really matter since the Roomba will just visit the base station to get cleaned out anytime the bin is full.

The mop raising in action. When it’s time to go over a carpet, the mop swings up to the top of the robot, so it will never get your carpet wet.

iRobot

The state of the art in combo cleaning robots is probably the $1,400 Roborock S7 MaxV Ultra or $1,550 Ecovacs Deebot X1 Omni, both of which have a dust bin and water tank in the robot, along with a giant base station that empties the robot of dust, fills it with fresh water, and then washes the mop and stores the wastewater in a separate tank. iRobot has a long way to go to catch up to that.

The swing-arm setup looks cool, but you have to wonder about its longevity. These robots take a lot of abuse by bumping into everything and constantly dealing with dust—it’s just the nature of the job. That large motion requires a ton of small moving parts and tiny pivot points. Something like a Roborock, by contrast, just has a mop permanently affixed to the bottom of the robot, and the mop setup is on a variable-height suspension, so it can be lifted up slightly to avoid touching carpet. That works for thinner carpets, but if you have a serious pile going on, the robot can sink down too far and get your carpet wet. iRobot says the swing arm setup is designed to avoid that.

iRobot’s press release throws some shade at the competition, saying, “Other 2-in-1 mops lift only a few millimeters—if at all—and, as a result, can leave wet messes on many carpets or rugs. The Roomba Combo j7+ is the only 2-in-1 with a mop that lifts itself to the top of the robot, completely away from carpet and rugs. Two fully retracting, durable, metal mop arms raise the mop pad when the robot senses carpet and rugs, preventing unwanted wet messes.”

The new combo bot is up for preorder now.

Ars Technica may earn compensation for sales from links on this post through affiliate programs.

Listing image by Roomba

Continue Reading

Trending