> That's not to say it's an easy problem to solve.
Incorrect. You simply decide that having less than 5 suppliers at any level is unacceptable and you bust companies up, repeatedly until you have those suppliers. That way when one goes bankrupt, you don't wind up with complete supply chain disintegration.
The solution is quite straightforward. However, it requires an electorate that has a couple of brain cells to rub together in order to understand the solution. And 30% of the US is willfully hostile to any real solutions while another 30% is happy to fiddle while everything burns.
> You simply decide that having less than 5 suppliers at any level is unacceptable and you bust companies up, repeatedly until you have those suppliers.
This is a very extreme solution, and eliminates many of the benefits from horizontal integration even when the benefits are passed onto customers. Consider:
- insurance companies
- banking
- utilities
It’s also hard to implement. What counts as a supplier? Is Google the sole supplier for search functionality? If 4 suppliers provide 1% of demand, and one supplies 96%, does that comply? If there’s only one company offering some new service (e.g. driverless cars), do they immediately get broken up?
Yes. Always. At all levels. I might provide a limit below which that doesn't happen (like $50 million in revenue), but as soon as you cross that limit, scrutiny should be automatic.
> This is a very extreme solution, and eliminates many of the benefits from horizontal integration even when the benefits are passed onto customers. Consider: - insurance companies - banking
There is no advantage to horizontal integration for consumers in those industries. If anything, the value is negative. The fact that people are quite a bit happier about credit unions than Chase says everything you need to know.
Sure, there are "efficiencies" to be gained by horizontal integration. What we have seen is that the horizontal integration is so strong that the industries are sclerotic in the face of crisis or change (see: toilet paper manufacturers in Covid who couldn't switch gears). It has become repeatedly clear that we need resilience and competition more than we need efficiency.
> utilities
Should be limited to natural monopolies and strongly controlled by the government. We have seen what happens when you create hybrid-type utilities that try to have some existence in the market (rather than being solidly government regulated) and the result is poor (see: PG&E).
> There is no advantage to horizontal integration for consumers in those industries. If anything, the value is negative. The fact that people are quite a bit happier about credit unions than Chase says everything you need to know.
IMO this claim is just too strong. I think you'd end up breaking up (or trying to) Lloyds of London, Spacex, Fedex, DHL, Boeing, Panasonic, ASML, Google, Apple, and many other very specialized companies. These businesses would be very expensive if they could only supply 1/5 of the market, to the point that many people would be totally priced out. The world can barely support 1 ASML, imagine if we had to pay for 5 of them. We'd be sent back to the 2000s, and that's _just_ computing.
> I think you'd end up breaking up (or trying to) Lloyds of London, Spacex, Fedex, DHL, Boeing, Panasonic, ASML, Google, Apple, and many other very specialized companies.
I see exemplars and no counterexamples.
Boeing turned to garbage when it took on McDonnel-Douglas--we were better off with the separate companies. YouTube not being bought by Google means that you don't have a single giant ad juggernaut and the copyright infringement that goes along with it. Apple being busted up means we have a division that actually focuses on computers in their own right rather than being a vestigial graft to the phone services division. Fedex was enough of a monopoly problem that Amazon bought carriers and, very painfully, set up its own delivery system.
> The world can barely support 1 ASML, imagine if we had to pay for 5 of them.
So, you prefer that we are two Chinese drone strikes from having a chip economy meltdown?
This is the kind of stuff that absolutely needs diversity. And part of the reason the ASML stuff is so expensive is because it doesn't have enough volume. So, for example, if the US had multiple fab lines that could consume the ASML machines, that would reduce the costs for ASML.
> Canned fruit, like what these farmers were producing, has been losing popularity for years. You can't force consumers to like it.
Has canned fruit actually lost popularity? Or did the grocery stores decide that the shelf space had a higher profit margin pushing something else?
The last couple of times I tried to get canned fruit for a recipe I had to actively hunt for the particular cans of fruit I needed (I needed to hit 3 different grocery stores).
I haven't tracked peaches recently, but I can tell you that canned apricots have been a bit thin on the ground for at least a couple of years.
> Has canned fruit actually lost popularity? Or did the grocery stores decide that the shelf space had a higher profit margin pushing something else?
> The last couple of times I tried to get canned fruit for a recipe I had to actively hunt for the particular cans of fruit I needed (I needed to hit 3 different grocery stores).
> I haven't tracked peaches recently, but I can tell you that canned apricots have been a bit thin on the ground for at least a couple of years.
Groceries stores with canned fruit being harder to find is entirely consistent with it being less popular. Pushing you to go to another store for something is bad, if you're a grocery store. That's a great way to drive off customers. There's a lot of shelf space at my local grocery stores still dedicated to fairly-redundant products or high amounts of extra copies of items, so I don't think they're being pushed out because something else is way more profitable. (My local stores have much larger selections of canned beans than canned peaches, for instance.)
I think it's just generational trends. Generally health-conscious consumers these days are more skeptical of canned vs fresh, and non-health-conscious have more junk food options than ever. It's also gotten easier to source fresh fruit across seasons than thirty or forty years ago, further squeezing canned options.
Yes. Global supply chains have improved, so it's easier to get fresh fruit year round (or closer to it) than it used to be. If they can, people will choose fresh over canned, for obvious reasons.
> people will choose fresh over canned, for obvious reasons
Not at all obvious. A lot of "fresh" produce in the US was refrigerated for more than a week before it arrived in the supermarket, from varieties that were designed to hold up to transport rather than flavor. Fruit that was canned at the height of the season is often much more flavorful than "fresh" off-season fruit.
The US has a problem with packing fruit in added sugar, which is sad but not inherent to canned fruit.
Where I live peaches are rare. It's all pears, oranges, and fruit cocktail. Not joking, there's five different variants of pears on the shelf at the grocery store, from sugar free to light syrup, and from three different brands. Canned plums? Nope. Apples? Nope. Strawberries? Nope. Cherries? Only around Thanksgiving and Christmas.
At the very least I can get all of those fresh and not canned, but honestly I'd prefer having canned versions as well because of all of the import uncertainty that ended up affecting things this past winter.
> Has canned fruit actually lost popularity? Or did the grocery stores decide that the shelf space had a higher profit margin pushing something else?
Do grocery stores make their own decisions about what goes on their shelves? I thought they mostly rented the shelf space to food vendors who were responsible for that.
For example, a while ago I complained on HN that a particular flavor of Triscuits was reliably out of stock whenever Safeway discounted Triscuits, and I was told that the way to address that, were I so minded, is to reach out to Nabisco on Twitter, because they - and not Safeway - make the stocking decisions.
Abandoning aether was NOT a "simple" answer. One you abandon aether, all manner of weirdness suddenly pops in.
Light being the same speed irrespective of observer is weird. Velocity dilating length and time is weird. Not having a preferential observation point is weird. Not needing a medium for transmission is weird. Not being able to agree on simultaneity is weird. etc.
Aether wasn't just something that a bunch of dullards clung to. You have to abandon some very long held common-sense understanding when you give it up.
> Occam's razor would point towards some general misunderstanding on which we have no evidence to reasonably speculate a cause.
Dark matter is the Occam's Razor theory. It explains almost all of the observations while assuming the least.
Why do you assume we have no evidence just because we don't have direct observations? Black holes were a similar phenomenon that we had no direct evidence for a fairly long time even though we had lots of other indirect phenomenon that really couldn't be explained any other way.
William of Ockham objected to his fellow theologians inventing things out of whole cloth (like dark matter question mark). That’s the point, not that a simpler explanation is more likely to be true.
The common understanding would have us believe that creationism, being simpler, outshines evolution, or that there is no such thing as a color revolution because the simplest explanation is that the mass protests are earnestly aggrieved locals.
At this point, online fraud control is getting absurd, and AI is just making it untenable. I simply won't use ebay for anything above $50 anymore.
Having physical locations that you have to come to pick up your Thneed protects both buyer and seller. Buyer can verify that what was described is delivered and seller can verify actual pickup with ID.
If they apply a bit of logistics for shipping between stores, Gamestop could crush it.
Fraud is forcing the pendulum to swing from everything-online back to everything-in-person.
Zig also differentiates between the wrapping and non-wrapping operators. The for loop example would toss a runtime error when the index underflowed in most compiler modes.
The if statement won't work since Zig would force a cast.
The tricky wrap sucks unless you use a power of 2. Then the Zig type can match (u4, u5, u7, etc.) and you would use wrapping arithmetic operators. And on smaller CPUs you NEED to use a power of 2 because division and mod are expensive.
AltaVista had a Java applet that would visualize the "clusters" that a search produced. You could then click on a "cluster" in order to exclude all the irrelevant ones and the search results would update.
For example: Searching on "python" would give you two obvious clusters one for "reptiles" and one for "programming languages". Clicking on the appropriate cluster would screen out all the irrelevant ones.
This is a feature still unmatched by any search engine today.
Um, yes? The big value that AMD had in the x86 market over competitors was their verification model. This has been known for decades.
> 3-seed nextpnr P&R on a Gowin GW2A-LV18 (Tang Nano 20K) — median Fmax × CoreMark iter/cycle = fitness
Every single "improvement" is basically about routing around how absolutely abysmally bad the Gowin FPGAs are. Kudos to that, I guess?
Gowin FPGAs have extraordinarily bad carry chain and block to block routing systems. They are literally so bad that a 32-bit ripple carry is almost as fast as the carry skip version even if you manually route it. Jump prediction is almost all about avoiding arithmetic computation at all (which most other FPGAs would have no problem with).
Memory accesses are super slow and locked to clock edges rather than level sensitive (why ID/RF and WB take entire cycles and nothing optimization could do could change it). The additions are all routing around that (Note the immutability of the ID and WB phases).
To top it off, the 5-stage pipeline is an annoying quirk of the RISC-V architecture having an immediate value offset on its load instruction. If the RISC-V load mandated 0 as the offset, the MEM read phase could overlap the RX phase since no ALU would be necessary (Store doesn't care because the result goes to memory rather than back to the register file so RF writeback isn't an issue). The absolutely horrific add performance of the Gowin FPGAs makes this acute.
Finally, try to put this on a board. I found that anything above about 175MHz out of Nextpnr failed to execute on actual hardware (please correct me if this isn't valid. It's been over a year or more since I tried Nextpnr on the SiPeed Tang Primer 20K). That's simply right around where a 32-bit add plus some routing sits on these FPGAs. There's something a bit off in the timing analysis code for Nextpnr and the AI is almost certainly optimizing into it.
That having been said: I would LOVE somebody to bounce AI off of reversing the architecture and bitstreams for the stupid-ass closed-source FPGAs. Now THAT would be a project worth throwing a couple of grad students and a bunch of subsidized AI tokens at.
Please apply this loop to nextpnr for any of the commodity Xilinx, Altera, or Lattice parts. For example, everything about Lattice has been stuck for almost a decade at this point.
Assuming that your claims about GoWin FPGA flaws are correct, isn’t the point of this experiment that it was able to exploit these flaws without manual guidance?
His claims are indeed correct;
Yes, you got my point tks!;
AND the loop produced architecture gains that are not exclusive to the GoWin FPGA (CoreMark/Mhz is higher than VexRiscV)
As a non-hardware guy, I read, “well, duh, for a 20yr practitioner dealing with the intricacies of specific FPGA series, all this makes tons of sense”.
It only makes sense to me because I tried to implement a RISC-V on these Gowin FPGAs and banged into the limitations and can distill them down. A junior engineer looks at this post-AI, shrugs, and says "I'm done."
The AI doesn't flag "Hey, my adder sucks. Move to a better FPGA architecture." A junior engineer pre-AI would have to bang on this a while, get frustrated at the critical paths, and eventually ask for help. At which point we would both look at this, identify that the adder was doing a 32-bit ripple carry, both have a "WTF?!" moment, and switch FPGA families.
In addition, the AI also doesn't flag how close to the margin you are. To my eye, almost all the Fmax gains look like PnR (place and route) noise. The DIV/REM obviously isn't and the replay predictor looks real. To top it off, the branch predictor wins look anomalously low to my eye.
This is what a bunch of us are yelling about with AI. AI gets you a thing. AI gets you no insight into that thing. And because the juniors will use the AI, they will never learn the insight.
Side note: The granularity of the CM/MHz numbers look a bit suspicious. Why are there identical entries?
The frontier is the verifier not in the sense of this project, but to every project. If we have a good verifier for a task, any task, this type of loop can be applied to it. Today LLMs are good enough to tackle FPGA projects, but what this type of loop will be applicable to many more things
Board should be arriving next week. I will let you know!
Incorrect. You simply decide that having less than 5 suppliers at any level is unacceptable and you bust companies up, repeatedly until you have those suppliers. That way when one goes bankrupt, you don't wind up with complete supply chain disintegration.
The solution is quite straightforward. However, it requires an electorate that has a couple of brain cells to rub together in order to understand the solution. And 30% of the US is willfully hostile to any real solutions while another 30% is happy to fiddle while everything burns.
reply