Live data from Hacker News

Intel Problems

stratechery.com

241–250 of 431 posts

Re: Intel Problems

#241

Earlier quoted context omitted.

It was more economics and political policy wonks, economists and politicians in general, who didn't just ask that question, but thrust forth the prescriptive rhetoric through a large grid of trade agreements, "Globalism is here to stay! Get over it! Comparative Advantage!" This is only the first in a long, expensive series of lessons these people will be taught by reality in the coming decades. I'm guessing this kind…

The US decided that it didn't like all the political disorder that came with managing a large proletariat. Instead, it decided to outsource the management of that proletariat to Asia and the 'global south.' Our proletariat instead was mostly liquidated and shifted itself into the service industry (not that amenable to labor organization) and the welfare/workfare rolls. There are so many things that the US would have…

An IP based economy just on its face seems like such a laughable house of cards. So your economy is based on government enforced imaginary rights to ideas? The proliferation of tax havens should be a sign that the system is bullshit - it exposes how little is actually keeping the profits of IP endeavors within a nation.

There is incredibly little respect for the society owning the means of production in a tangible real sense, instead we have economies that run on intangibles, where the intangibles allow 600lb gorillas like Oracle to engage in much rent seeking while simultaneously avoiding paying dues to the precise body that granted them their imaginary rights. The entire status quo feels like something some rich tycoons dreamed up to sell to the public the merits of systematically weakening their negotiating position on the promise that one day a Bernie Sanders type would descend from the heavens and deliver universal basic income fueled by the efficiency of private industry through nothing but incorruptability and force of personality.

China seems to be successful in part because they have no qualms with flexing dictatorial power to increase the leverage of the state itself. This may be less economically efficient but it means they actually get to harvest the fruits of any efficiency. Intellectual property law? They just ignore it and don't get punished, since punishing them would be anti-trade.

Re: Intel Problems

#242
post #151

I think one day we’re going to wake up and discover that AWS mostly runs on Graviton (ARM) and not x86. And on that day intel’s troubles will go from future to present. My standing theory is that the m1 will accelerate it. Obviously all the wholly managed AWS services (Dynamo, Kinesis, S3, etc.) can change over silently, but the issue is EC2. I have a MBP, as do all of my engineers. Within a few years all of these ma…

This is kind of what should happen right? I'm not an expert, but my understanding is that one of the takeaways from the M1 success has been the weaknesses of x86 and CISC in general. It seems as if there is a performance ceiling which exists for x86 due to things like memory ordering requirements, and complexity of legacy instructions, which just don't exist for other instruction sets. My impression is that we have b…

M1's success shows how efficient and advanced the TSMC 5 nm node is. Apple's ability to deliver it with decent software integration also deserves some credit. But I wouldn't interpret it as the death knell for x86.

Re: Intel Problems

#243

Earlier quoted context omitted.

> I think it'll be a good thing when people stop worrying about process node technology and start worrying about performance and power usage. I think it's more that people attribute too much significance to process node technology when trying to understand why performance & power are what they are. For single-core performance the gains from a node shrink are in the low teen percentage increases. Power improvements at…

While you are right that due to their design CPUs like Apple M1 can reach the same single-thread performance as Intel/AMD at a much lower clock frequency and such a clock frequency could be reached much earlier, e.g. already Intel Nehalem in 2009 reached 3.3 GHz as turbo, while Sandy Bridge in 2011 had 3.4 GHz as base clock frequency, it would have been impossible to make a CPU like Apple M1 in any earlier technology…

> To achieve its very high IPC, M1 multiplies a lot of internal resources and also uses very large caches. All those require a huge number of transistors.

Yes & no. Most of the M1 die isn't spent on CPU, it's spent things like GPU, neural net, and SLC cache. A "basic" dual-core CPU-only M1 would be very manufacturable back in 2011 or so. After all, Intel at some point decided to spend a whole lot of transistors adding a GPU to every single CPU regardless of worth, there were transistors to spare.

Re: Intel Problems

#244

A major incentive for the US government to get involved is touched on. Not only is Taiwan 'just off the coast' from China, China is coming for it and intends to assimilate Taiwan back into China just as Hong Kong and Macau have been. At that point, the only sustainable leverage the rest of the world would have in chip technology would be ASML.

It is thought provoking to think that the USA's interest in Taiwan is more about protecting TMSC than protecting a democratic state in East Asia. By this line of thinking, building capacity in Arizona, or anywhere outside Taiwan, is good for TMSC and for the USA but weakens Taiwan.

> It is thought provoking to think that the USA's interest in Taiwan is more about protecting TMSC than protecting a democratic state in East Asia.

Why it though provoking? It's always realpolitik. All wars are.

There's always a pretext but the subtext is what actually causes wars.

Re: Intel Problems

#245

Earlier quoted context omitted.

I think it'll be a good thing when people stop worrying about process node technology and start worrying about performance and power usage. Intel's 14nm chips are already competitive with AMD's (TSMC's, really) 7nm chips. The i7-11700 or whatever the newest one coming out soon is called, is going to be pretty much exactly on parity with AMD's Ryzen 5000 series. So if node shrinkage is such a dramatic increase in perf…

> I think it'll be a good thing when people stop worrying about process node technology and start worrying about performance and power usage. I think it's more that people attribute too much significance to process node technology when trying to understand why performance & power are what they are. For single-core performance the gains from a node shrink are in the low teen percentage increases. Power improvements at…

> We've been "stuck" at the 3-5ghz range for something like half a decade

It’s closer to two decades, actually. Pentium 4 (Northwood) reached 3.06 GHz in 2002, using 130 nm fabrication process.

Re: Intel Problems

#246
post #81

Earlier quoted context omitted.

"The vast majority of developers never profile their code." Protip: New on the job and want to establish a reputation quickly? Find the most common path and fire a profiler at it as early as you can. The odds that there's some trivial win that will accelerate the code by huge amounts is fairly decent. Another bit of evidence developers rarely profile their code is that I can tell my mental model of how expensive some…

> Find the most common path and fire a profiler at it as early as you can. The odds that there's some trivial win that will accelerate the code by huge amounts is fairly decent. I've found that a profiler isn't even needed to find significant wins in most codebases. Simple inspection of the code and removal of obviously slow or inefficient code paths can often lead to huge performance gains.

I mean I love finding those "obvious" improvements too but how do you know you've succeeded without profiling it? ;)

Re: Intel Problems

#247

Earlier quoted context omitted.

> Maybe I'm missing something, but don't the vast majority of applications don't care about what architecture they run on? There can be issues with moving to AArch64, for instance your Python code may depend on Python 'wheels' which in turn depend on C libraries that don't play nice with AArch64. I once encountered an issue like this, although I've now forgotten the details. If your software is pure Java I'd say the…

Sure, but you're talking about short term problems. RPi, Graviton, Apple Silicon, etc... are making AArch64 a required mainstream target.

That's true. AArch64 is already perfectly usable, and what issues there are will be ironed out in good time.

Re: Intel Problems

#248
post #140

I think one day we’re going to wake up and discover that AWS mostly runs on Graviton (ARM) and not x86. And on that day intel’s troubles will go from future to present. My standing theory is that the m1 will accelerate it. Obviously all the wholly managed AWS services (Dynamo, Kinesis, S3, etc.) can change over silently, but the issue is EC2. I have a MBP, as do all of my engineers. Within a few years all of these ma…

I commented [1] on something similar a few days ago, >Cloud (Intel) isn’t really challenged yet.... AWS are estimated to be ~50% of HyperScalers. HyperScalers are estimated to be 50% of Server and Cloud Business. HyperScalers are expanding at a rate faster than other market. HyperScaler expanding trend are not projected to be slowing down anytime soon. AWS intends to have all of their own workload and SaaS product ru…

>Which is why I was extremely worried about Intel. (Half of) The lucrative Server market is basically gone.

Right. I suspect in time we'll look back to this time, and realize that it was already too late for Intel to right the ship, despite ARM having a tiny share of PC and server sales.

Their PC business is in grave danger as well. Within a few years, we're going to see ARM-powered Windows PCs that are competitive with Intel's offerings in several metrics, but most critically, in power efficiency.

These ARM PCs will have tiny market share (Assuming these ARM PCs can emulate x86 with sufficient performance (as Apple does with Rosetta), consumers and OEMs will realize that ARM PCs work just as well as x86 Intel PCs. At that point, the x86 "moat" will have been broken, and we'll see ARM PCs grow in market share in lockstep with the improvements in ARM manufacturing capacity (TSMC, etc...).

Intel is in a downward spiral, and I've seen no indication that they know how to solve it. Their best "plan" appears to be to just hope that their manufacturing issues get sorted out quickly enough that they can right the ship. But given their track record, nobody would bet on that happening. Intel better pray that Windows x86 emulation is garbage.

Intel does not have the luxury of time to sort out their issues. They need more competitive products to fend off ARM, today. Within a year or two, ARM will have a tiny but critical foothold in the PC and server market that will crack open the x86 moat, and invite ever increasing competition from ARM.

Re: Intel Problems

#249

How does Moore's law figure into this? I suspect that TSMC runs into the wall that is quantum physics at around 1-2nm. Considering that TSMC has said that they will be in full production of 3nm in 2022, I can't see 1nm being much beyond 2026-2028. What happens then? Does a stall in die shrinks allow other fabs to catch up? It appears to me that Intel stalling at 14nm is what opened the door for TSMC and Samsung to ca…

Quantum effects haven't been relevant for a while now. The "nanometer" numbers are marketing around different transistor topologies like FinFET and GAA (Gate-all-around). There's a published roadmap out to "0.7 eq nm). Note how the "measurements" all have quotes around them:

https://www.extremetech.com/computing/309889-tsmc-starts-dev...

Re: Intel Problems

#250
post #173

I found the geopolitical portion to be the most important aspect here. China has shown a willingness to flex its muscles on enforcing its values beyond their borders. China is smart, and plays a long game. We don't want to wake up one day and find they've flexed their muscles on their regional neighbors similar to their rare earths strong-arming from 2010-2014 and not have fab capabilities to fall back on in the West…

> [...] and not have fab capabilities to fall back on in the West. I'm not too concerned: - There are still a number of foundries in western countries that produce chips which are good enough for "military equipment". - Companies like TSMC are reliant on imports of specialized chemicals and tools mostly from Japan/USA/Europe. - Any move from China against Taiwan would likely be followed by significant emigration/"bra…

Yeah, for some reason, I assumed that military equipment mostly used, like, low performance but reliable stuff. In-order processors, real time operating systems, EM-hardening. Probably made by some company like Texas Instruments, who will happily keep selling you the same chip for 30 years.
Post reply on HN