> AMD has been outsourcing verification to its customers for many years
You have no idea what you're talking about if you think for a second that a large CPU vendor like AMD would delegate verification to "its customers". It's like saying that Boeing just builds planes and tells airlines to make sure they fly correctly before allowing passengers to board them.
Seriously? So when I see someone on HN who I think is clearly wrong, I can't say that? I agree that my tone could be improved, but I only said it that way in response to the parent's tone.
Also, how was our argument a "flame war"? It only lasted for 3-4 replies, and was quite civil in my opinion.
I tend to always agree with your decisions, but this one is a bit too extreme.
It's not about "tone" but content and it's quite simple, though always harder to see in one's own case. The HN guidelines say this:
When disagreeing, please reply to the argument instead of calling names. E.g. "That is idiotic; 1 + 1 is 2, not 3" can be shortened to "1 + 1 is 2, not 3."
Similarly, "You have no idea what you're talking about if you think for a second that a foo like bar would baz" can be shortened to, "A foo like bar wouldn't baz."
HN's rules apply regardless of how other commenters are behaving. If they didn't, we might as well have no rules, because it always feels like others are behaving worse.
That's a bit too extreme, don't you think? Disliking AMD is one thing, but claiming that the entire initial run of an AMD processor might be faulty is just silly.
My guess is that you have no idea how much effort goes into verification and testing of something as complex as a microprocessor. A significant chunk of NRE costs goes into verification and test AFAIK.
Around 80-85% is the projected cost of Verification and Validation (pre/post - Silicon). This is what historical data shows for us (as a CPU design firm doing ARMV8)
Intel's FDIV bug was an outlier. Besides, nowadays most of the ISA is implemented in microcode[1]. There are two advantages to this approach: 1) it is much easier to verify the microcode unit (it's simpler/smaller), and 2) it allows CPU vendors to "fix" ISA implementation issues post-release by issuing microcode updates.
[1]: ISA => microcode is equivalent in some respects to C => LLVM IR
There was also an issue with transaction memory with both Haswel & Broadwell. The fix was to disable TSX support via micro code update. I wouldn't call disabling a fix personally. I doubt Intel even compensated the folks who bought it for the TSX support.
Yeah, that list is far from exhaustive. Years and years ago, Google's websearch cluster validation suite found a tricky bug that isn't on that list. The exchange with the CPU manufacturer was amusing.
Most of the ISA may be implemented in microcode, but the parts that matter are certainly not microcoded!
For example the TLB access and the fast path of the TLB miss are not microcoded. You only get to microcode if you have to set an accessed bit or a dirty bit, or if there is a page fault.
OP said "will [they] even work?". That's a very different thing from a TLB bug. Although I concede that depending on the nature of the TLB bug, it may lead to a significant performance hit.
Regarding your criticisms of AMD's past processors, are you coming at this from a consumer standpoint, or are you just disappointed with their architecture and/or implementation?
I ask because the Phenom II seems quite loved by its users[1], and the FX series was also quite good for its price. I had a FX-8320 in my last PC and it was quite a capable CPU in my opinion, and many others seem to agree[2].
For datacenter operators there is no difference between "performance is so crippled that the TCO doesn't pencil out" and "does not work." No difference at all.
My guess is you haven't been in this business long enough to comment. Rev. A AMD "Barcelona" parts did not work in any meaningful way. They all had to be thrown out.
The linked decision, if you read it, discusses this. Google has in the past responded to similar requests but after the Microsoft decision they responded differently.
Lunatic conspiracy theories about what Google does with email abound. There are several likely and reasonable explanations for why you can't find the mails you expect. Two of the most likely being 1) the messages were never acknowledged with a 220 response from gmail's smtp protocol translators to begin with; 2) the messages were accepted but are still in flight for some reason. Gmail does not accept and then silently drop messages. If they are accepted, they will be delivered.
You need to remember the fact that already Snowden's revelations have proven that the NSA and other government agencies all have specific budgets for astro turfing activities (manipulating the public opinion by massively participating in online discussions).
And a couple of days ago, there was a nice post on Reddit's front page summing up the situation on Reddit. Reddit is basically completely compromised by whoever has lots of money (government, big industries, etc). Any company can buy astro turfing services nowadays.
So no, you can't trust public online discussion anymore. Not on Reddit and not here. Unless for topics you are absolutely certain that no economic interest is part of the equation.
Yes, rights aren't absolute. If the governments wants your data on a self hosted server they need a warrant. In comparison, you have basically zero privacy protections when your data is in the hands of a third party.
You could "self-host" on a cloud server in, say, China, or Russia, or Iran (if they have any hosting services.)
I mean, the governments of those places will probably snoop your emails, but if their contents have nothing to do with them, they won't care. And they have no treaties with the US to force their hand to turn anything over.
Think of your server as Edward Snowden. What country should it hide in, so the US can't legally get to it?
You're forgetting the possibility of rubber hose cryptanalysis applied on you. In fact just by hosting in such places, you're probably inviting more attention.
>the governments of those places will probably snoop your emails
Uhm, how? Gmail supports Transport Layer Security (TLS), and >80% of their emails to and from other providers do as well (https://www.google.com/transparencyreport/saferemail/). Reject non-TSL emails, give the server a public key and tell it to throw away the email plaintext, and the only remaining threat vectors seem like "get rubber hosed into disclosing your private key" and "server gets compromised, causing future emails (but not past ones) to get exfiltrated".
Are we talking about bulk requests? The case we seem to be discussing here involves "data associated with three Google accounts held by an individual who resided in the United States."
>Acquire the free time and money to not have to work,
I do a lot of work while walking and hiking; as a programmer, most my time is spent thinking which I do not need a computer for. Well, I need a phone or tablet to find stuff on internet to think up a solution, but I can do that during walking. And I do. It won't give me a month of 8 hours / day, but it does give hours of walking per day. I did hikes of 4-6 hours / day in fulltime work days while sleeping / working in hotels on way points of the hike. 4-5 hours of walking/dictating/looking up stuff and then 2 hours of typing in/testing/delegating when arriving at a hotel.
I was going to do that but did not yet. I must say that I generally try to do not do any 'frontend' stuff. So sure I do prototypes of the frontends but someone else does a UX/UI and the coding of that. So my thinking is architectural, the different components, algorithms and low level implementations. All of these I can work out in my head with some Google + paper which means I can walk during that. If I would have to do HTML/CSS it would not work; I am not sure if that's because i'm just not that fast with it, but it's way too much iteration to not make me sit in front of a (large) screen for too long. For instance, when I have to do embedded work, thinking up the coding, let's say for a RISC chip, can be worked out to quite a lot of detail before typing 1 line. Worse; if I don't work it out in detail, I'll just end up frustrated behind a debugger and a lot of crashes. I have done that since I was young as assembly or low level C never worked well for me while iterating; only in recent years I realised that this applies to other projects as well.
Another thing that has to be said is that peripheral vision is important; when I walk writing or staring in my phone, I don't have issues tripping, walking into things etc. I would be very handicapped if I did not have that. And another thing with my eyes is that I read very very small print which is why I can work quite comfortably on small tablets / computers (I am looking forward to receiving the Pyra).
There are plenty of people hiking the AT who are decidedly not wealthy. Heck, I've met quite a few hikers who hike specifically to save money (living off social security).
No, that's irrelevant because of all parent's 3 points could easily be part of a job like land or mineral surveying, field biology, forestry, and so on. You don't have to quit working to go outside and walk around all day.
Not really because you choose foods that are calorie-dense (i.e. foods without water weight, and fatty foods).
3000-4000 calories per trail day was typical during my 1,900 PCT miles. I never had to restrict intake due to weight concerns. Also binged on burgers and ice cream at every opportunity in town.
This is what one of my larger resupplies looked like, 31,750 calories for a 7-day stretch:
When I did the JMT I was restricted in the last leg by the volume of my bear cannister. That's mostly what I was thinking of. I dropped from around 3500 calories per day to around 2800, and I sure felt it (also the last leg is by far the most difficult, southbound).
Yes, the southernmost stretch of JMT is also the burliest stretch of the PCT!
Many thru hikers bend the rules for bear canister use. We carried canisters because we had to, but I don't know any PCT nobo'er who was able to cram all their food inside for the long Sierra Nevada (JMT) stretches. We slept next to our food so that we could chase away the bear.
I think the complication comes in trying to balance exercise with everything else you want to do. Not everyone has the interest or the means to go on a 30 day hike.
Completely agree. But while I usually find overly-distilled things unhelpful, for the vast majority of people it really does come down to consuming fewer, better calories, and moving more (i.e, burning more).
But I agree that the overthinking comes when people are trying to fit 27 hours of daily activities into 24, and also when people are pursuing results without burning more and/or consuming fewer calories.
It is HUGELY wasteful. The statement "one of the most energy-efficient buildings in the world" is a prime example of obscuring the truth, as well. When considering full-system efficiency this is undoubtedly one of the most wasteful structures in the world. 12000 employees, 14000 parking spaces, and a 1-hour walk to the nearest point of mass tranportation (Lawrence Caltrain, which is poorly served; the more frequently served Sunnyvale Caltrain is a 75-minute walk from this building).
As far as I can tell from the schedule the minimum bus trip from this site to Sunnyvale Caltrain is 44 minutes, plus an average 10 minutes for headway. That assumes you did not start off on the east face of the UFO, in which case you should add a walk of up to a mile to get to the VTA stop on Wolfe.
That doesn't seem obviously true to me. Google for example doesn't run shuttles to Mountain View Caltrain.
Apple operates door-to-door shuttles to SF and other population centers but that isn't really the point. The point is they have obviously provisioned enough parking to give each and every one of their employees a parking space, with some left over. It's just irresponsible given the traffic situation in the area, and to call it energy efficient is greenwashing.
All states spend much more on roads than they take in on gas taxes, so the issue of whether gas taxes are "diverted" is moot. Every state also "diverts" a huge amount of general revenues to road building and maintenance. To stay with your example of Illinois, which recently passed a "lockbox" ballot measure for road funds, tolls and fuel taxes take in 1.6 billion a year but the state spends over 4 billion a year just on road maintenance alone, not even counting ancillary expenses like DMV, State Police, etc.
Yes Trumps loves some infrastructure. He loves it so much that the first act of his Dept of Transportation was cancelling a long-planned $700m project to electrify Caltrain. Sad!
