AMD Scores First Top 10 Zen Supercomputer… at NVIDIA

https://www.anandtech.com/show/15874/amd-scores-first-top-10-zen-supercomputer-at-nvidia

One of the key metrics we’ve been waiting for since AMD launched its Zen architecture was when it would re-enter the top 10 supercomputer list. The previous best AMD system, built on Opteron CPUs, was Titan, which held the #1 spot in 2012 but slowly dropped out of the top 10 by June 2019. Now, in June 2020, AMD scores a big win for its Zen 2 microarchitecture by getting to #7. But there’s a twist in this tale.

via AnandTech https://ift.tt/phao0v

June 22, 2020 at 05:37PM

Twenty One Pilots is streaming a non-stop music video using fan uploads

https://www.engadget.com/twenty-one-pilots-never-ending-music-videos-145550118.html

Have you ever enjoyed a song that you could seemingly listen to it non-stop? Twenty One Pilots is betting that you will, and it’s using tech to make you part of the song. The band has debuted a never-ending music video for its pandemic-themed single “Level of Concern” that relies on fan material for many of its visuals. You just have to upload a clip between three and 30 seconds long (orientation doesn’t matter) and it’ll be added in “real-time.”

The integration is slick — the stream blends fan footage into overlays and transitions that accompany the band’s own performance. You may only get a fleeting moment of glory, but it’s hard to deny the allure of making yourself part of a music video that thousands will see at any given moment.

There is a question as to how the band moderates content. It’s easy to see someone using this to flash viewers or otherwise give them grief. We’ve asked for comment and expect to have an answer soon. However those crowdsourced films are handled, this is a clever concept that makes sense at a time when shooting a conventional music video would be impractical.

Update 6/22 12:05PM ET: Jason Nickel from Imposium, who worked on the project, told Engadget that there’s a moderation team reviewing footage and removing anything that “isn’t shareable material.” You won’t see trolls in the feed, then. Nickel added that there’s “already been a backlog” of videos due to demand, so don’t be surprised if your snippet takes a while to appear. You can read the full statement below.

“Almost all of the uploaded content we’re seeing from fans has been focused on being fun and creative. We do have a moderation team reviewing the videos as they come into the system and removing anything that isn’t shareable material. In general though, it’s been overwhelmingly positive. As videos are approved for use, they’re queued up to go into the never ending music video. With so many submissions so quickly, there’s already been a backlog of videos that are in line to go into the video stream. Each music video has about 70 different people in it so it may take some time to get caught up.”

via Engadget http://www.engadget.com

June 22, 2020 at 10:03AM

Scientists explored a buried Roman city without digging up an ounce of soil

https://www.popsci.com/story/science/falerii-novi/

The researchers used radar to reveal structures still lurking beneath the surface.

The researchers used radar to reveal structures still lurking beneath the surface. (Google Earth; Verdonck et al., Antiquity, 2020/)

Traditionally, checking out the towns and cities of our ancient ancestors has required a lot of digging. While we’ve been exploring ancient sites for centuries, scientists have been working to find newer, less destructive ways to learn about major sites like Pompeii, as well as the smaller communities that made up ancient Rome. Scientists at the University of Cambridge and Ghent University have used ground-penetrating radar to create a highly-detailed map of Falerii Novi, a Roman town that prospered from around 241 BC until 700 BCE and is now buried deep below the earth. The researchers published their findings last week in Antiquity.

Around six feet beneath the city 40 miles north of Rome, the researchers found evidence of a bath complex, large homes, a marketplace, public buildings and monuments, temples, and water pipes used by the city’s locals, which may have numbered in the thousands, says author Martin Millett, a professor of classical archaeology at Cambridge.

“What’s really interesting is not just what you can see, but how clearly you can see it,” says Millett.

Radar works by sending signals out and analyzing how they bounce back, providing evidence for objects we can’t see with the naked eye. But penetrating the solid soil of central Italy requires a great deal more power than it takes to search the sea or the skies. The antennae themselves might still be small, Millett adds, but higher energy and frequency allows the waves to penetrate the tough ground. The archeologists dragged these radar instruments across the entire city, just more than a tenth of a square mile, taking readings every five or so inches along the ride.

Materials like metal and marble have different frequencies than the earth surrounding them, says Helena Fracchia, a professor of the classics at the University of Alberta. That means archaeologists can get a pretty good picture of where things like buildings or pipes fit into to the city landscape. While a typical excavation can take years of fieldwork, using GPR only takes a couple of months.

“The results from Falerii Novi are absolutely fabulous as a preliminary step in studying the city,” she says.

While these results do give us a unique birds-eye view of a Roman city, Fracchia says, the use of ground-penetrating radar has its downsides. It shows you a whole town pretty quickly and cheaply, but there’s a lot more to archeology than just mapping out where things went.

Super-strong radars captured this image of a hidden ancient temple in the buried town of Falerii Novi.

Super-strong radars captured this image of a hidden ancient temple in the buried town of Falerii Novi. (L. Verdonck/)

When you wander around a museum, looking at the big picture diagrams of ancient communities is only part of the experience. The art, tools, and fossils that archaeologists can only find by digging provide much more insight into how people lived and died there. To take one example from the new study, the researchers managed to discover a large ancient monument by the northern gate of the city. But in order to understand what kind of practices and rituals went on inside the structure, Fracchia says, scientists will have to do some excavating to search for relevant art, tools, and remains.

Millet hopes to keep working on this site, finding as much detail as possible, but also intends to take the technology to other Roman cities. After all, with the help of these super-powered radars, you can find buried treasure without breaking out a spade.

via Popular Science – New Technology, Science News, The Future Now https://www.popsci.com

June 21, 2020 at 08:07AM

It Was 100 Degrees in Siberia Today. Yes, That Siberia

https://earther.gizmodo.com/it-was-100-degrees-in-siberia-today-yes-that-siberia-1844107792

A view of an energy plant’s dumpsite in Novosibirsk, a city in the Russian region of Siberia.
Photo: Rostislav Netisov (Getty Images)

A freak heatwave has been scorching most of the Arctic for weeks now, but it broke records Saturday when the temperature hit 100.4 degrees Fahrenheit in a town in Siberia, one of Russia’s northernmost regions.

It’s likely the hottest temperature ever recorded north of the Arctic Circle, CBS meteorologist Jeff Berardelli wrote on Twitter, though the recording is still pending verification.

If you’re having trouble parsing why this is such a big deal, here’s a dose of perspective that’ll make your eyes bulge out: Verkhoyansk, a town that’s even further north than Fairbanks, Alaska, has had the same number of 10o-degree days on record as Miami, Florida. (Sure, it’s only one, but tell me that’s not insane.)

Reports of the record-shattering heat quickly went viral online, shared by meteorologists worldwide and even Swedish climate change activist Greta Thunberg. While 2019 was Russia’s hottest on record, this year already seems poised to surpass that. Last month, Siberia reported temperatures almost 40 degrees above normal for this time of year. (Also, parts of the region caught on fire. And since you know what they say about bad news coming in threes, I’m waiting for some Godzilla-like kaiju to emerge from the melted ice).

Verkhoyansk, which is home to a little more than a 1,000 people, also holds the record for where you can find the greatest temperature ranges on Earth. While the weather there cooled down to around 80 by Saturday evening, temperatures regularly fall well below zero, with the lowest recorded at -90 degrees.

G/O Media may get a commission

This weekend’s historic heat is simply the latest horror in the Arctic as the region continues to get royally screwed by the impacts of climate change. Last month, sea ice extent reached the lowest level ever recorded in May. I know I’ve said before that 2020 seriously needs some chill, but this time, I mean it literally.

via Gizmodo https://gizmodo.com

June 20, 2020 at 08:33PM

NVIDIA Announces PCIe A100 Accelerator: 250 Watt Ampere In A Standard Form Factor

https://www.anandtech.com/show/15867/nvidia-announces-pcie-a100-accelerator-250w-ampere-in-a-standard-form-factor

With the launch of their Ampere architecture and new A100 accelerator barely a month behind them, NVIDIA this morning is announcing the PCIe version of their accelerator as part of the start of the now-virtual ISC Digital conference for high performance computing. The more straight-laced counterpart to NVIDIA’s flagship SXM4 version of the A100 accelerator, the PCie version of the A100 is designed to offer A100 in a more traditional form factor for customers who need something that they can plug into standardized servers. Overall the PCIe A100 offers the same peak performance as the SXM4 A100, however with a lower 250 Watt TDP, real-world performance won’t be quite as high.

The obligatory counterpart to NVIDIA’s SXM form factor accelerators, NVIDIA’s PCIe accelerators serve to flesh out the other side of NVIDIA’s accelerator lineup. While NVIDIA would gladly sell everyone SXM-based accelerators – which would include the pricey NVIDIA HGX carrier board – there are still numerous customers who need to be able to use GPU accelerators in standard, PCIe-based rackmount servers. Or for smaller workloads, customers don’t need the kind of 4-way and higher scalability offered by SXM-form factor accelerators. So with their PCIe cards, NVIDIA can serve the rest of the accelerator market that their SXM products can’t reach.

The PCIe A100, in turn, is a full-fledged A100, just in a different form factor and with a more appropriate TDP. In terms of peak performance, the PCIe A100 is just as fast as its SXM4 counterpart; NVIDIA this time isn’t shipping this as a cut-down configuration with lower clockspeeds or fewer functional blocks than the flagship SXM4 version. As a result the PCIe card brings everything A100 offers to the table, with the same heavy focus on tensor operations, including the new higher precision TF32 and FP64 formats, as well as even faster integer inference.

NVIDIA Accelerator Specification Comparison
  A100
(PCIe)
A100
(SXM4)
V100
(PCIe)
P100
(PCIe)
FP32 CUDA Cores 6912 6912 5120 3584
Boost Clock 1.41GHz 1.41GHz 1.38GHz 1.3GHz
Memory Clock 2.4Gbps HBM2 2.4Gbps HBM2 1.75Gbps HBM2 1.4Gbps HBM2
Memory Bus Width 5120-bit 5120-bit 4096-bit 4096-bit
Memory Bandwidth 1.6TB/sec 1.6TB/sec 900GB/sec 720GB/sec
VRAM 40GB 40GB 16GB/32GB 16GB
Single Precision 19.5 TFLOPs 19.5 TFLOPs 14.1 TFLOPs 9.3 TFLOPs
Double Precision 9.7 TFLOPs
(1/2 FP32 rate)
9.7 TFLOPs
(1/2 FP32 rate)
7 TFLOPs
(1/2 FP32 rate)
4.7 TFLOPs
(1/2 FP32 rate)
INT8 Tensor 624 TOPs 624 TOPs N/A N/A
FP16 Tensor 312 TFLOPs 312 TFLOPs 112 TFLOPs N/A
TF32 Tensor 156 TFLOPs 156 TFLOPs N/A N/A
Relative Performance (SXM Version) 90% 100% N/A N/A
Interconnect NVLink 3
6 Links? (300GB/sec?)
NVLink 3
12 Links (600GB/sec)
NVLink 2
4 Links (200GB/sec)
NVLink 1
4 Links (160GB/sec)
GPU GA100
(826mm2)
GA100
(826mm2)
GV100
(815mm2)
GP100
(610mm2)
Transistor Count 54.2B 54.2B 21.1B 15.3B
TDP 250W 400W 250W 300W
Manufacturing Process TSMC 7N TSMC 7N TSMC 12nm FFN TSMC 16nm FinFET
Interface PCIe 4.0 SXM4 PCIe 3.0 SXM
Architecture Ampere Ampere Volta Pascal

But because the dual-slot add-in card form factor is designed for lower TDP products, offering less room for cooling and typically less access to power as well, the PCIe version of the A100 does have to ratchet down its TDP from 400W to 250W. That’s a sizable 38% reduction in power consumption, and as a result the PCIe A100 isn’t going to be able to match the sustained performance figures of its SXM4 counterpart – that’s the advantage of going with a form factor with higher power and cooling budgets. All told, the PCIe version of the A100 should deliver about 90% of the performance of the SXM4 version on single-GPU workloads, which for such a big drop in TDP, is not a bad trade-off.

And on this note, I should give NVIDIA credit where credit is due: unlike the PCIe version of the V100 accelerator, NVIDIA is doing a much better job of documenting these performance differences. This time around NVIDIA is explicitly noting the 90% figure in their their specification sheets and related marketing materials. So there should be a lot less confusion about how the PCIe version of the accelerator compares to the SXM version.

Other than the form factor and TDP changes, the only other notable deviation for the PCIe A100 from the SXM version is the number of NVLink-connected GPUs supported. For their PCIe card NVIDIA is once again using NVLink bridges connected across the top of A100 cards, allowing for two (and only two) cards to be linked together. NVIDIA’s product sheet doesn’t list the total bandwidth available, but as the PCIe V100 supported up to 100GB/sec in each direction using two links, the PCIe A100 and its 3 NVLink connectors should be able to do 150GB/sec, if not more.

Otherwise the PCIe A100 comes with the usual trimmings of the form factor. The card is entirely passively cooled, designed to be used with servers with powerful chassis fans. And though not pictured in NVIDIA’s official shots, there are sockets for PCIe power connectors. Meanwhile, with the reduced usage of NVLink in this version of the card, A100’s native PCIe 4 support will undoubtedly be of increased importance here, underscoring the advantage that an AMD Epyc + NVIDIA A100 pairing has right now since AMD is the only x86 server vendor with PCIe 4 support.

Wrapping things up, while NVIDIA isn’t announcing specific pricing or availability information today, the new PCIe A100 cards should be shipping soon. The wider compatibility of the PCIe card has helped NVIDIA to line up over 50 server wins at this point, with 30 of those servers set to ship this summer.

via AnandTech https://ift.tt/phao0v

June 22, 2020 at 02:43AM