> If demand for training and inference wanes, these hardware investments will be almost entirely wasted
Nvidia also need to invent smth then, as pumping mining (or giving good to gamers) again is not sexy.
What's next? Will we finally compute for drug development and achieve just as great results as with chatbots?
> Nvidia also need to invent smth then, as pumping mining (or giving good to gamers) again is not sexy.
They do! Their research page is well worth checking out, they wrote a lot of the fundamental papers that people cite for machine learning today: https://research.nvidia.com/publications
> Will we finally compute for drug development and achieve just as great results as with chatbots?
Maybe - but they're not really analogous problem spaces. Fooling humans with text is easy - Markov chains have been doing it for decades. Automating the discovery of drugs and research of proteins is not quite so easy, rudimentary attempts like Folding@Home went on for years without any breakthrough discoveries. It's going to take a lot more research before we get to ChatGPT levels of success. But tools like CUDA certainly help with this by providing flexible compute that's easy to scale.
There was nothing rudimentary about Folding@Home (either in the MD engine or the MSM clustering method), and my paper on GPCRs that used Folding@Home regularly gets cites from pharma (we helped establish the idea that treating proteins as being a single structure at the global energy minimum was too simplistic to design drugs). But F@H was never really a serious attempt at drug discovery- it was intended to probe the underlying physics of protein folding, which is tangentially related.
In drug discovery, we'd love to be able to show that virtual screening really worked- if you could do docking against a protein to find good leads affordably, and also ensure that the resulting leads were likely to pass FDA review (IE, effective and non-toxic), that could potentially greatly increase the rate of discovery.
From 0 day, I was shocked by Ukrainian elite twitter/facebook forces. They even continue fighting from Poland/Moldova, they never were affected by broken supply chains or blockades. I don't see any viable Kremlin alternative for this, so maybe with time they will take over and we will finally see democracy in Russia.
Most Ukrainians are forced to support Russia because the way how Zelensky "defends" Ukraine by freeing prisoners and giving advice to fight with Molotov cocktails.
AFAICS, it's all the same to him: If he were dreaming only of reviving the Tsarist (why "pre-"?) Russian Empire and not the Soviet Union, he'd hardly have gone on the record saying "the fall of the Soviet Union was the greatest geopolitical tragedy of thetwentieth century".
IMHO war isn't a scale of bodies. Each side has a strategic objective(s), and winning/losing should be measured against that.
Case in point: Vietnam. uSA and allies lost between 1/3 and 1/2 lifes than the North. And the USA loses were small (60.000 in a total of 1.5 million deaths). But it was a total defeat of the USA because none of the strategic objectives were achieved.
Ukraine: strategic objectives are being an independent country with a fairly working democracy. Russia: conquer and create a puppet state in Ukraine and keep pressing west againt NATO countries.
In the end we should measure against that, not by body count. I don't like it either.
The most important objective was achieved — preventing the expansion of the communism block to other countries. Vietnam today is a market economy, the country is communist in name only.
Also, please try to Vietnam. The US is very favorably perceived, more than China or Russia, although probably below European countries.
In the sense, the US lost the war, but won in the end, at a very steep price, of course
Sure! USA won a lot of wars just by not fighting: spreading wealth and culture via commerce is a far more effective way to _conquer_ countries than obliterating them.
And USA defeated the whole communist block using the same weapons. Also, don't dismiss tourism as a _weapon_: it's would be more effective to send millions of frienly tourists to Cuba than to embargoe them for decades.
But that's disgressing. Vietnam was only used as a sample of strategic millitary defeat albeit losing less people. I don't think anyone in the Whitehouse has planned the vietnamese frienship ahead.
It wasn't the US's intervention that succeeded in making it a "non communist" state, so the US failed in its objectives. Just because eventually Vietnam itself decided what path it wants to take doesn't mean, it was because of US.
US is perceived more favourably than China because their war with China was more recent and they have active disputes with PRC threatening island territories in South China sea.
Moreover, US has massive influence in terms of their media and their brands, something which China still doesn't match. Same reason even South Korea has massive appeal in Vietnam.
Perfectly valid standpoint, but USA and Vietnam is not Russia and Ukraine. Geography, culture, history - all of this. We can hardly guess what the hell is in Putin's head and what those plans really were aimed at. We are guessing thoroughly, though.
There is a great risk of massive starvation in besieged Ukranian cities, where it's too dangerous for sheltering civilians to venture out and procure or supply food.
Anecdotally, I live in Ireland and a number of locals have mobilised to gather food and medicine to send it to Ukraine. Then I came across this article and now think this is happening across the country: https://www.rte.ie/news/ireland/2022/0308/1285137-ireland-ai...
I wouldn’t be surprise if many other communities in Europe are doing the same. The problem might be getting the aid to the right place.
anecdotally - yes. We bought and brought some stuff following that https://www.facebook.com/lnyuz/posts/5490539337642083 (drop-off in San Jose, some people order stuff from Amazon with delivery right to there), and while i don't see food among the list of requested item, boxes packed with donated food were pretty prominent there (and even though food wasn't mentioned we actually brought food too - long term packaged high calories low weight, we thought that probably that may be needed too and it happened so, seems like many people do that).
During the first several days. When the outside world pulled Russian news channels off air and, possibly, blocked them on social media (I don't remember if that's what Facebook did); and started DDoSing its servers.
We are using our aviation too accurate for conventional war. As result, we've lost some (2..4) of our decent aircrafts. What supply chains do we need to use more (not ~15) bombers on higher altitudes, e.g. sacrificing precision for being protected from almost-dead Ukrainian ADS ("Buk"s or Stingers, maybe some S-300 near Kyiv)?
AFAIK, they are resupplied with Stinger ADS only, if you know more, pls tell me.
Yesterday we lost 1 aircraft in Kharkiv, which was bombing on terribly low altitude to preserve precision.
Two days ago we lost one AC in Chernigyv, same situation here.
Oryx reports 11 lost aircrafts during war, there are some listed whose nobody can prove (IMO about a half). Im not telling you that its a fake, but that resource definitely sometimes agrees with official Kyiv information.
The Russians confirmed the death of air-force general Andrei Sukhovetsky. The Ukrainians claim to also have killed army major general Vitaly Gerasimov.
>> The intelligence arm of the Ukrainian defence ministry said Maj Gen Vitaly Gerasimov, chief of staff of the 41st Army, had been killed outside the eastern Ukrainian city of Kharkiv, along with other senior officers.
>> Andrei Sukhovetsky was the commanding general of the Russian 7th Airborne Division and a deputy commander of the 41st Combined Arms Army, and by far the most senior Russian figure to have died in the conflict so far.
>> Mr Putin confirmed that a general had been killed in a speech updating the Russian people on the progress of the conflict, eight days into the deadly invasion.
Nvidia also need to invent smth then, as pumping mining (or giving good to gamers) again is not sexy. What's next? Will we finally compute for drug development and achieve just as great results as with chatbots?