Tag: images

Mythic

Mythic can do an 8-bit multiply and add in a single transistor

2020-10-17: AI Analog Compute

Mythic is the first and only company that have been able to implement a deep learning model like ResNet 50 in a non-digital architecture: > 50 layers, 1000 fps, 3W total, 9->2ms latency, 8 TOPS/W in 40nm silicon. 10x cost advantage over digital chips.

2023-04-11: Commercialization takes a long time

Mythic’s analog chip uses less power by storing neural weights not in SRAM but in flash memory, which doesn’t consume power to retain its state. And the flash memory is embedded in a processing chip, a configuration Mythic calls “compute-in-memory.” Instead of consuming a lot of power moving millions of bytes back and forth between memory and a CPU (as a digital computer does), some processing is done locally. Mythic’s success on that front has been variable: The company ran out of cash and raised $13 million in new funding and appointed a new CEO.
I asked him whether the state of analog computing today could be compared to that of quantum computing 25 years ago. Could it follow a similar path of development, from fringe consideration to common (and well-funded) acceptance?

It would take a fraction of the time. “We have our experimental results. It has proven itself. If there is a group that wants to make it user-friendly, within 1 year we could have it.” And at this point he is willing to provide analog computer boards to interested researchers, who can use them with Achour’s compiler.

Skull Logs

now i really want a fire pit!

Even though it’s spring, it’s still chilly outside at night and a human skull ‘log’ might be just what the witch doctor ordered to keep you warm. It’s definitely a conversation piece, I’ll give it that. The human skull ‘log’ is for gas or liquid propane pits only.

Settlement of the Americas

24 ka Americans?

About 24 ka ago, when much of North America was buried under the ice of the Last Glacial Maximum, a few hunters took shelter in a small cave above the Bluefish River in what is now northwestern Yukon. The hunters had killed a Yukon horse and were butchering it using super-sharp stone shards called microblades. As they sliced out the horse’s meaty tongue, the microblades left distinctive cuts in its jaw bone. Millennia later, archaeologist and doctoral candidate Lauriane Bourgeon spotted those marks through her microscope at the University of Montreal and added the fragment of ancient jaw bone to her small selection of samples for radiocarbon dating.

2017-04-26: 130 ka Americans? Those are fighting words, 6x-10x earlier than generally believed. Needs a LOT more evidence.

An unidentified Homo species used stone tools to crack apart mastodon bones, teeth and tusks approximately 130 ka ago at a site near what’s now San Diego.

2021-11-13: There’s plenty of evidence that North America was settled early, but not successfully:

The problem with the idea of an early, pre-Amerindian settlement of the Americas is that ( by hypothesis, and some evidence ) it succeeded, but ( from known evidence) it just barely succeeded, at best. Think like an epidemiologist – once humans managed to past the ice, they must have had a growth factor greater than 1.0 per generation – but it seems that it can’t have been a lot larger than that, because if they had averaged, say, 3 surviving kids per generation ( r = 1.5) , their population would have exploded, filling up all the habitable territories south of the glaciers in less than 2 ka. Maybe they didn’t have atlatls. The Amerindians certainly did. Maybe they arrived as fishermen and didn’t have many hunting skills. Those could have been developed, but not instantaneously. An analogy: early Amerindians visited some West Coast islands and must have had boats. But after they crossed the continent and reached the Gulf of Mexico, they had lost that technology and took several 1000 years to re-develop it and settle the Caribbean. Along this line, coastal fishing settlements back near the Glacial Maximum would all be under water today. Maybe they fought among themselves to an unusual degree. I don’t really believe in this, am just throwing out notions. Maybe their technology and skills set only worked in a limited set of situations, so that they could only successfully colonize certain niches. Neanderthals, for example, don’t seem to have flourished in plains, but instead in hilly country. On the other hand, we don’t tend to think of modern human having such limitations. One can imagine some kind of infectious disease that made large areas uninhabitable. With the low human population density, most likely a zoonosis, perhaps carried by some component of the megafauna – which would also explain why it disappeared.

2022-02-08: A more detailed look at the 24 ka hypothesis

I present this history of the last 36 ka of migration from the perspective of a scientist who places genetic evidence in the forefront of the investigation and then tests the models it produces with archaeological, linguistic, and environmental evidence. Around 36 ka BP, a small group of people living in East Asia began to break off from the larger ancestral populations in the region. 25 ka BP, the smaller group in East Asia itself split into 2. 1 gave rise to a group referred to by geneticists as the ancient Paleo-Siberians, who stayed in Northeast Asia. The other became ancestral to Indigenous peoples in the Americas.

24 ka BP, both groups independently began interacting with an entirely different group of people: the ancient Northern Siberians. Some archaeologists and geneticists argue that this meeting of the 2 grandparent populations of Native Americans—the group in East Asia and the ancient community in Northern Siberia—occurred because people moved north, not south, in response to the last glacial maximum (LGM), a period in which much of northern North America was covered by massive glaciers. Thus, many geneticists look north, to Beringia, for the location of the refugia that may have allowed the ancestors of Native Americans to survive the ice age.



2022-08-14: 37 ka evidence

About 37ka BP, a mother mammoth and her calf met their end at the hands of human beings.

Bones from the butchering site record how humans shaped pieces of their long bones into disposable blades to break down their carcasses, and rendered their fat over a fire. But a key detail sets this site apart from others from this era. It’s in New Mexico – a place where most archaeological evidence does not place humans until 10s of 1000s of years later. Based on genetic evidence from Indigenous populations in South and Central America and artifacts from other archaeological sites, some scientists have proposed that North America had at least 2 founding populations: the Clovis and a pre-Clovis society with a different genetic lineage.

Arquitectura de la Remesa

For her photo series, Aragón headed to those more remote regions, including San Mateo Ixtatán, to document the effects of emigration on the family members who stayed behind. She trained her lens on arquitectura de la remesa, structures built with money sent back from the US With the influx of capital, some families build grandiose, multi-story houses that suture together some elements of adobe homes with imagined characteristics of uniquely “American” ones. Still, some of the blueprints may seem puzzling: a staircase crawling down the outside of a house; 17 rooms with only 1 toilet; 6 light bulbs blazing in a single room. “The people who build them, they’re not architects or engineers. They don’t have any education about how to build a house; they have never even been in a house like you’d see in the city.”

Earth seen from Mars

The Earth and Moon, as seen from Mars. That image was taken by the phenomenal HiRISE camera on board the Mars Reconnaissance Orbiter, which was more than 200M kilometers from Earth at the time. It’s actually a composite of a few separate images, processed to show the relative size and position of our planet and its moon.

Studio D retrospective

This year we published a number of foundational research reports including Isa Pa EcoCash on mobile money adoption in Zimbabwe. It’s a good example of a project that required an experienced team and a nuanced appreciation of the political swirl that can occur. It also revealed the distortions between official quant data and what was happening on the ground.

Final Fantasy XV Food

clever, very clever.

Creating a recipe in Final Fantasy XV involved several development team members. According to Hasegawa, the process begins in the art department, where the dish’s ingredients and desired appearance are planned out. Another team takes it from there — takes it outside, specifically, to actually cook on a camp stove. “Our team members took out their gear and went camping to cook outdoors. You know how even the simplest foods can taste really delicious when you’re out camping? We wanted to focus on that same feeling while we created them.” Ignis serves up some very fancy-looking meals in the Coleman-branded camp dishware in the game, but it’s believable due to this detailed care in their creation. You can buy that the dedicated outdoor chef could make a beautiful croque madame at a campsite — because a team of dedicated outdoor chefs in Japan actually did the real-world work first. The completed dishes, “served” in the game’s camping and diner scenes, were then photographed from various angles. They were then scanned to create 3D data for the digital artists to work with, but artists weren’t just left to work off of static images. The digital art team also handled the physical dishes prepared by the food team and their ingredients — how are you really going to perfectly render a zucchini unless you’ve actually held a slice yourself? Recipes were then tasted by the teams creating the in-game models, and the 3D data tweaked as necessary.