"The more direct inspiration for me was an email from Stuart Russell in September 2020 to Alexander Koller and I about our ACL 2020 paper".

It's a good article, and worth reading. But somehow I also have enough trouble understanding how she would make this error that---in the inverse of the fluidity argument---I start to doubt her the rest of her logic based on one silly irrelevant grammar mistake.

Speaking on BBC Radio 4’s This Cultural Life (via The Mirror), Cocker addressed claims that the inspiration was Danae Stratou, a Greek woman who attended St Martin’s at the same time as Jarvis, but confirmed that “it wasn’t her because she had blonde hair and the girl had dark hair.”

https://www.nme.com/news/music/jarvis-cocker-is-on-a-quest-t...

Essentially, this means that all loops are already unrolled from the processors point of view, minus a tiny bit of overhead for the loop itself that can often be ignored. Since in binary search the main cost is grabbing data from memory (or from cache in the "warm cache" examples) this means that the real game is how to get the processor to issue the requests for the data you will eventually need as far in advance as possible so you don't have to wait as long for it to arrive.

The difference in algorithm for quad search (or anything higher than binary) is that instead of taking one side of each branch (and thus prefetching deeply in one direction) is that you prefetch all the possible cases but with less depth. This way you are guaranteed to have successfully issued the prefetch you will eventually need, and are spending slightly less of your bandwidth budget on data that will never be used in the actual execution path.

As others are pointing out, "number of comparisons" is almost useless metric when comparing search algorithms if your goal is predicting real world performance. The limiting factor is almost never the number of comparisons you can do. Instead, the potential for speedup depends on making maximal use of memory and cache bandwidth. So yes, you can view this as loop unrolling, but only if you consider how branching on modern processors works under the hood.

True, but is there some particular reason that you want to minimize the number of comparisons rather than have a faster run time? Daniel doesn't overly emphasize it, but as he mentions in the article: "The net result might generate a few more instructions but the number of instructions is likely not the limiting factor."

The main thing this article shows is that (at least sometimes on some processors) a quad search is faster than a binary search _despite_ the fact that that it performs theoretically unnecessary comparisons. While some computer scientists might scoff, I'd bet heavily that an optimized ternary search could also frequently outperform.

American military bases and other equipment in the Persian Gulf region suffered extensive damage from Iranian strikes that is far worse than publicly acknowledged and is expected to cost billions of dollars to repair, according to three U.S. officials, two congressional aides and another person familiar with the damage.

The Iranian regime swiftly retaliated after the Trump administration attacked on Feb. 28, hitting dozens of targets across U.S. military bases in seven Middle East countries. Those attacks struck warehouses, command headquarters, aircraft hangars, satellite communications infrastructure, runways, high-end radar systems and dozens of aircraft, according to the U.S. officials and an assessment by the American Enterprise Institute, a conservative think tank in Washington, D.C.

In the initial days of the war, an Iranian F-5 fighter jet bombed the U.S. base Camp Buehring in Kuwait, despite the base having air defenses, a rare breach that marked the first time an enemy fixed-wing aircraft has struck an American military base in years, according to two of the U.S. officials.

The U.S. bases that came under attack are home to thousands of American troops, and in some cases their families, though they were largely cleared out in the days and hours before the U.S. and Israeli went to war with Iran. The Pentagon has not detailed the extent of the damage to U.S. military bases publicly or, according to the U.S. officials, to members of Congress.

“We do not discuss battle damage assessments for operation security reasons,” a Pentagon official said in a statement. “Our forces remain fully operational, and we continue to execute our mission with the same readiness and combat effectiveness.”

U.S. Central Command declined to comment on battle damage assessments.

Last month, the administration asked private satellite companies, including Planet Labs, to withhold imagery of the bases from the public, making the extent of the destruction difficult to assess, according to the U.S. officials and experts, including a statement from Planet Labs to their customers.

The administration’s request remains in place, a Planet Labs spokesperson said. A White House spokesperson declined to comment.

Some Republican lawmakers privately have expressed frustration directly to senior Pentagon officials about their refusal to provide information about the extent of the damage or any cost estimate for repairs, according to two GOP congressional aides.

“No one knows anything. And it’s not for lack of asking,” one of the aides said. “We have been asking for weeks and not getting specifics, even as the Pentagon is asking for a record high budget.”

Asked for comment, White House spokesperson Olivia Wales said the U.S. had achieved the military objectives of Operation Epic Fury. “As the president has said, this was the last, best time to strike, and — thanks to our heroic warfighters — the operation was a tremendous success,” Wales said in a statement. “President Trump took decisive action to ensure that Iran can never have a nuclear weapon to threaten the United States or our assets and allies in the region, and Americans are already safer for it.”

The damage to and cost of repairing the bases could reignite a yearslong debate over the merits of maintaining U.S. bases in such close proximity to an adversary like Iran. Some national security officials, including some serving in the Trump administration, have for years pushed to move U.S. bases in the region further east and away from Tehran’s reaches. The issue also could embolden critics of America’s military presence overseas who have advocated for shrinking the U.S. presence in the Middle East, one U.S. official and one person familiar with the matter said. The three U.S. officials familiar with the damage to U.S. bases in the Middle East described it as extensive. The headquarters building for the U.S. Navy in Bahrain, the nerve center for the Navy’s operations in the region, sustained serious damage, the officials said. They said other parts of the base in Bahrain also suffered significant damage that is likely repairable.

Multiple hangars and warehouses at Ali Al Salem Air Base in Kuwait also were struck, according to the American Enterprise Institute’s unofficial assessment that was reviewed by NBC News. The assessment also shows a munitions storage facility at a military base in Erbil in northern Iraq was damaged and a runway at the sprawling Al-Udeid Air Base in Qatar was destroyed.

U.S. bases had been cleared of U.S. troops and other personnel, so many of the bases were left essentially empty and vulnerable to attack by Iranian missiles and drones. Many of the troops who were temporarily relocated are expected to return to the bases once tensions in the region subside.

Thirteen U.S. troops have been killed in the conflict and as many as 400 have been wounded, although more than 90% have returned to duty, according to the U.S. military. The Pentagon has refused to provide specifics, but during an April 8 briefing, Chairman of the Joint Chiefs of Staff Dan Caine said the U.S. and partners in the Gulf region intercepted 1,700 ballistic missiles and one-way attack drones during the war. A fourth U.S. official said only a fraction of the projectiles actually got through the U.S. and ally defenses.

Congress is considering legislation to support the cost of the war, including unspecified repairs and other costs in a so-called supplemental bill that could exceed $100 billion, according to two of the U.S. officials and two other people familiar with the matter.

According to the AEI’s assessment, Iran hit more than 100 targets across 11 bases in seven Gulf countries. Those attacks fell on U.S. and host-nation bases in Qatar, the United Arab Emirates, Bahrain, Jordan, Kuwait, Iraq and Saudi Arabia.

“As part of Epic Fury, the potential future costs to rebuild American military infrastructure overseas may include repair, reconstruction, outright replacement, or even abandonment/decommissioning of locales,” Mackenzie Eaglen, a senior fellow at AEI, said in a statement about the group’s assessment. “War damage also includes estimated costs for infrastructure that is unsalvageable.”

Eaglen’s cost estimate for repairing the infrastructure is more than $5 billion, but that amount does not account for some of the radar systems, weapons systems, aircraft and other equipment destroyed in the Iranian strikes, she said. Eaglen has worked on defense and budgetary and military readiness issues for years and is a former Pentagon official.

The Iranians damaged at least two air defense systems in the region, according to the U.S. officials.

Iran has also destroyed U.S. military aircraft. NBC News reported that at least one fighter jet, more than a dozen MQ-9 Reaper drones, two MC-130 tankers and four helicopters known as “little birds” were also destroyed.

Additional helicopters, tankers, an E-3 Sentry plane and two more helicopters known as Jolly Greens were also damaged, according to U.S. officials and information provided during a Pentagon briefing.

Radar systems in the UAE and a satellite communication system in Bahrain were also damaged the U.S. officials said, but it’s not clear whether Bahrain or the UAE would cover the cost of those systems.

https://old.reddit.com/r/neoliberal/comments/1svdezz/iran_ca...

Comments URL: https://news.ycombinator.com/item?id=47835635

Points: 706

# Comments: 483

I grow several varieties of sour cherries in my yard, and frequently use them whole and without further processing. Usually I use them in a recipe like this: https://en.wikipedia.org/wiki/Clafoutis. Sometimes I pit them first, sometimes I don't. Sometimes I'll even happily snack on them raw.

No, like most small fruit you aren't going eat them because you are desperate for calories. But they actually aren't any harder to prepare or use than lots of other tasty things that people traditionally grow.

But if it's merely [flagged] (with associated penalties keeping it off the front pages) there is no option for vouching to remove the penalty.

This one is currently [flagged] with no option for vouching.

A more accurate claim might be "More people voted for Trump in 2024 than any other candidate".

Sure, that would work significantly better than no predictor at all. But you'd agree that a better predictor would work better, right? The missing detail might be how expensive mispredicted branches are compared to other costs. If you can go from 50% accuracy to 90% accuracy, it wouldn't be surprising to more than double your performance.

> Didn’t realize it used the input value as well, which to me makes no sense — the whole point is to avoid having to inspect the value.

It doesn't, and can't for the reasons you hint at. The reason branch prediction is necessary is that the value often isn't available yet when the branch is taken. Was there something in the article that implied the opposite?

--

I wonder if Daniel's tricksy approach using a random number generator to simulate a complex pattern is misleading people here.

One of the main benefits of branch prediction is predicting the end of a loop, particularly, a loop within a loop. In assembly, a loop is just a comparison at the end and a branch back to the beginning. Assume you had a loop that always executes 8 times, or some other small fixed value. Also assume there is some reason you can't unroll that loop, and that loop is inside another loop that executes millions of times. It's a real boost to performance if you can consistently predict the end of the inner loop.

If you predicted just on the last time the loop closing branch was taken, you'd always miss the ending. But if you can remember a pattern that is longer than 8, you can always get it right. This is obviously valuable. The bigger question is how much more valuable it is to predict a loop (where "loop" might actually be a complex execution pattern across multiple branches) that is thousands long rather than just 8. But quantifying how long this pattern can be on different processors is part of the groundwork for analyzing this.

Referring to a dime as a dollar bothered me too. Going deeper, the absence of the olive branch is actually an intentional historical reference to the Revolutionary War, where peace was tragically lost. According to the artist who made it, the open claw is to symbolize the desire to regain it:

The image takes inspiration from the Great Seal of the United States, and represents the colonists before and during the American Revolution, Custer explained. While he included the arrows from the seal, he left out the olive branch to symbolize the fact that the colonies hadn’t yet reached peace — but left the claw open to demonstrate that they were waiting for it.

https://www.spotlightpa.org/news/2026/02/philadelphia-mint-c...

Custer’s design for the tails side of the coin — which features an eagle with one empty claw and one claw holding 13 arrows — won out in the selection process.

The image takes inspiration from the Great Seal of the United States, and represents the colonists before and during the American Revolution, Custer explained. While he included the arrows from the seal, he left out the olive branch to symbolize the fact that the colonies hadn’t yet reached peace — but left the claw open to demonstrate that they were waiting for it.

Note that this is probably a terrible strategy here, though. The most likely effect of adding SpaceX to the index in the manner described is that the price of SpaceX is likely to go artificially high as the index funds fight for the very few available shares.

If you are actually going to do this, you'd probably want to sell out of the ETF before it's added, wait for SpaceX to be fully added to all the index funds, and only then simultaneously buy into the ETF and short SpaceX.

Why would you think that? Because of Musk's reputation for magnificence and generosity? Because rich gigantic companies don't care about keeping operating costs low? Or because janitors at SpaceX require special skills that can only be obtained by paying many multiples of the market rates for standard janitors?

https://www-spiegel-de.translate.goog/kultur/philosoph-juerg...

They have CNN currently rated as "Lean Left", changed from "Left" a few years ago due to the changes you point out. You'll need to read the fine text to see their reasoning. While there were some votes that they are now biased to Right, this was a minority position.