New USB controller firmware, apparently. The fact that it can have any effect on the CPU temperature is a little puzzling.
I work in a office where the air conditioner is set to 28°C
Wow. As someone who is used to a room temperature of 20C, 28C is above the point of idle sweating and becoming lethargic. While that's still within the range that most consumer electronics are rated for, I believe a room temperature of 20C is the normal design target.
I'm from Australia and in summer the cooling can't cope and does indeed result in indoor temperatures exceeding 28°C. However, I'm drenched in sweat and certainly can't focus at that point. I will happily head outside and enjoy the crazy heat, but be productive? Nope!
I think heating and A/C target around 21°C in Melbourne.
However, I traveled to Vienna for a week long contract some years back. Sure it was cold outside (snowing), but I found all the offices and buildings excruciatingly hot and couldn't focus one bit. Made for a very miserable working week.
Of course, the locals all managed fine, so despite it seeming like a personal preference thing, I can only assume it's something you can adapt to.
I think average office people don't know how to use A/C and heating well.
Every time I work in an office with a large and diverse enough population I have to endure stuffy warm air in winter and being blasted by cold wind in summer.
I find it ridiculous, but, while wearing shorts, a t-shirt and sandals outside I have to keep a hoodie at my desk just to survive the office summer.
I relate to this so much. We are meant to adapt to our climate. Why do we resist so much? I ended up quoting osha regulation to ensure coworkers don’t turn the heat down below 68 degrees Fahrenheit. One coworker set it at 60 degrees at one point early this spring when it was a pleasant temperature outside. I biked in in shorts and a T-shirt and was shivering at my desk.
Haha I have the opposite issue and I also was quoting OSHA:
> OSHA recommends temperature control in the range of 68-76° F and humidity control in the range of 20%-60%.
When my desk thermometer started reading 76F+
Here is the thing, you can put on another layer of clothing or using a blanket, I cannot take off a layer of skin and my office does not allow for shorts/tank tops.
I hate that some people insist on cranking up the heat during the winter. Some of us wear the kinds of pants that will actually keep our legs warm out in freezing temperatures, and we shouldn't have to change into shorts to not be sweating at work.
I remember being seconded to Singapore for a bit back in the early 2000s. Second day there, I was scouring the shops for a warm sweater. At one point I was considering fingerless gloves.
Vienna in the summer is like a baking oven, the Austrians call it Hundstage (there is also a film by Ulrich Seidl called like that who has this heat as a topic).
A lot of the buildings are able to cool passively e.g. by keeping open a window to hallway and to the street which in old buildings results in a steady airflow.
Air condition only increases the cities temperature further so passive solutions are prefered.
It comes from when Sirius (the "Dog Star") rises at the same time as the Sun - I think it is more about the astrological culture than the language, although that will be a vector.
I am afraid you are mistaken, the ambient temperature increase is meaningful[0] in this case up to a predicted 2 degree increase for Paris if they double the numbers of their ACs.
ACs are more than just fans – they are basically fridges for your room, and like every thermodynamic machine they produce more heat than they can cool. Only not in your room but outside.
AC means cooling your room while heating your city.
Yeah, you'd think the places with very cold climates would have appropriately lower working temperature at the offices. Nope. Most sweltering offices I have ever been to were both in Canada and in Sweden in the middle of winter - easily -20C outside, and the office felt like it was set to 30C, it was insanely uncomfortable for me.
This seems to work the other way too. Lot's of hot countries seem to set the AC to 18c. I did trip to south america, and I needed warm clothing specifically for air conditioned busses and dormitories.
In the UK (where I live) I believe the recommended guideline is 21-24c
Yes, I also noticed it in SE Asia. I guess, due to the fact that AC was (and in some places is) too expensive for the majority of population, cold air feels like a luxury attribute.
American here, from the South. I've noticed the same sort of phenomenon travelling to the Canadian plains in winter, or among Northern transplants to the South in midsummer. Canadian sitting rooms felt like greenhouses to me (80+ Fahrenheit) despite it being 20 below outside, and my boss has commented before on visiting Southern households where the A/C is set to 68 in July.
My theory is that it's more about not-to-exceed ceilings [or the converse, floor] temperatures rather than average temperatures. When you have the huge temperature differential between the conditioned space and the outside, opening a door or something can affect the conditioned space's mean temperature by a huge amount for a short time. By over-conditioning or over-heating a space, you're more likely to be able to ensure the interior temperature never rises over ~75F, than if you'd set it at a comfortable 72F or so. Or in winter, it won't drop below a minimum comfortable temperature indoors when someone has to go outdoors.
Of course, once you've adapted to that habit, you may not always make seasonable adjustments to those settings.
The coldest parts usually have double doors at the entrances right? So temperature doesn't really swing so much from people transit. Beside, office buildings are usually extremely large compared to volume of exchanged air at the entrance.
Where I live in the tropics we have ~32 degrees every day all year. I put my AC on 27.5 and think that's more than cold enough. I guess the stable temperature makes the slight difference in temperature (and high difference in air humidity) more noticeable.
Presently located in Launceston, typically have my heat pumps at home set to 19 while I’m awake, 17 when I sleep.
I work in a workshop, so it’s both colder and warmer for longer depending on the season.
The -4 degree mornings don’t effect my cognitive not physical abilities, but the 30+ degree days certainly do. I have to constantly remind myself to slow down.
It has been 41 here lately and I leave the indoors at 22, and the shock is marvelous. That 41 is at greater than 50% relative humidity, so the big impact is that my sweating starts to be effective again.
Mist fans are great for this. I just hope my laptop will cope with the humidity. But after AC can't help (32+C), mist fan is still effective. North of Melb.
The level of comfort also heavily depends on the level of humidity, doesn't it? Seems pointless to compare temperatures only.
Would love to get all the complainers here to spend one month in winter where I live. We frequently have -35 to -45°C outside, and +30-35°C inside. How about a 70°C temperature difference?
That sounds like the city is a power outage away from severe hypothermia or death, depending on the insulation of buildings. How are system failures handled?
By prayer. It happened a couple times, although not yet on the city scale.
I don't personally know anyone who was affected. What they did, according to the local press, was stuffing every crack and cranny with old rags and sleeping fully clothed (as in fur coats and the like) and under as many blankets as they could get their hands on.
It's not only for heating to fail. If your car breaks down in the middle of nowhere you could be in for a bad time if you cannot stay warm.
It's standard advice to keep blankets or the like in your car for this specific scenario. Besides, for the majority of people in the Nordic contries, their normal winter comforter is warm enough to keep them warm should the heating fail.
I had to go to one of these cold places once for a couple of weeks. The car I was given was a modified Toyota Land Cruiser with two 100 liter tanks, additional insulation, additional battery inside, Webasto heater for both inside and engine compartment. It was also stocked with some useful stuff in case you do get stuck out in the middle of nowhere.
Those people were ready, when you live your whole life in such environments you quickly get "infused" with a lot of experience that was accumulated over generations.
So I'd be more worried about people in a regular country getting severe low temp weather than a failing heating system in one region used to extreme cold.
We have city-wide central heating system. If you don't pump (literally) boiling water into the system, it gets pretty cold in apartments furthest from the heating station. I live somewhere in between, so it gets even hotter the closer you live to the station. What do people do to combat this? They open the windows, and it's -40°C outside. This further lowers the efficiency of the whole system, so they get the water temperature even higher. Rinse and repeat.
Of course, instead of opening the windows, you could always install a gate valve (or whatever it is called in English) and close it to make the water go through your apartment's heating elements without actually (mostly) heating them, and that's what they recommend, but why bother? The level of self-entitlement of many people here, you wouldn't believe. I could tell stories all day long.
Like a caricature of a stereotypical American tourist.
Copenhagen has the same system (it's actually steam that is piped around the city; gases are much easier to move than liquids). It's called district heating.
In the basement of my building there's a heat exchange, which uses the steam to heat water. This water is piped around the building.
My fairly new apartment has a meter fitted on that hot water pipe. It measures how much heat I extract from the water, and I'm charged accordingly.
In some older buildings, they have some sort of temperature logging device on each radiator. I assume this is to charge the residents according to their usage.
Minor thing, but Copenhagen is odd with the steam. They're in the process of moving to hot water.
The trend is towards heating the water less, houses are getting better insulation, less energy is needed, and the hotter water is more expensive when heated by a heat pump and also in losses.
That sounds ridiculously ineffeicient. My post-Soviet city has district heating, and similar temperatures, but usually bottoms out at -35c.
In each building there is a system which takes the district heat and uses it to heat the water of the building. Both hot tap water and central heating. But the water is separate from district heat, so in each building the temperatures can be adjusted.
In older buildings you still have the issue you describe on a macro level - apartments on the bottom floor are too hot, and on the top are too cold - but it doesn't matter how close or far your building is from the district heating station.
Yeah, they've been talking about implementing something like that for as long as I can remember. The current system wastes so much heat that it's always 10-15°C warmer in the city than outside its borders.
I mean, how much coal do you have to burn to warm the air outside by 10-15°C?
By whom? They are given a job to provide heat for the city - and they are doing so. The people telling them to do that job are probably also the same people unwilling to schedule any budget to upgrade the system so make it more efficient.
Thermostats have been around for a hundred years or so. Could solve the problem...
What I've heard is that some portion of pension is given as free heating. So it makes no sense for the individual to conserve it. Giving it as money and then billing for the heating could create some other problems too.
I am not sure about pensions, but we pay quite a lot for heating. We do it every month and all year long, actually, because money management is a problem for many, and if you shift all the payments to one season (which is 4-8 months, depending on the weather), each monthly payment would be a problem for many. So they stretch it out across the year.
In which nordic country do you live? Im swedish and we have centralized water heating, and to me it sounds insane that people where you live does not have gate valves. The only time I heard of someone without that was when a friend changed the heating system in his building, and that was like a week of high temp. 30 degrees inside during winter sounds totally insane. Here we discuss that we should lower from 23 in the winter to save resources.
These buildings were developed in massive numbers back in post-WW2 years, across the whole Soviet Union. with the task to move people out of barracks in some sort of relatively comfortable housing. Back then there wasn't much money to spare (still isn't). They typically have pretty bad insulation and are built out of, as we say, "cardboard" walls.
When it's cold enough outside (-35°C and lower), you can see the energy losses with a naked eye, as hot air escapes through every crack in the building (and there are a lot of them). No need for thermal cameras.
People over here just don't give a fuck about nature. Now it's the middle of summer, it's +33°C outside. I am sitting in the office with two ACs running at full capacity and with all windows open, because "there isn't enough air". I tried arguing with this, but it's pretty difficult to go against all my coworkers at once. It's a cultural problem, and I have no idea how to fight this.
Thanks for the sanity check, though, at least I know it's not me going crazy.
> I am sitting in the office with two ACs running at full capacity and with all windows open, because "there isn't enough air". I tried arguing with this, but it's pretty difficult to go against all my coworkers at once.
What is your solution? To close windows and enjoy excess CO2?
When people complain about not having enough air it's almost always mean that ventilation system in building is insufficient and there in fact too much of CO2.
Sure. They don't seem to have any problems in the middle of winter, when windows stay fully closed for at least 3 months each year.
About a half also smokes, and then they "don't have enough air".
By the way, the office faces an arterial road, as they are called, with heavy traffic right outside. When you open a window, it fills very quickly with traffic fumes. This gives me persistent headaches almost every day.
> They don't seem to have any problems in the middle of winter, when windows stay fully closed for at least 3 months each year.
It's very much possible that during the winter ventilation performs better since there is hot air inside and cold outside. Or your office building just have similar hot air loss as your housing and it's helps to remove excessive CO2. Thermodynamics huh.
Also oxygen consumption and overall feel can be affected by temperature of environment.
> About a half also smokes, and then they "don't have enough air".
No surprise since heavy smoker lungs and hearts often perform worse and they going to be more sensitive to CO2. It's even worse if they're overweight or have other health issues.
> Is it really better than CO2?
I cannot say without some kind of measurement, but I guess yeah: even if there is massive road outside it's could still be better than staying with constantly high CO2. It's can be worse long-term for your health, but to be productive human brain need oxygen and many people more sensitive to CO2 level than others.
Personally I don't smoke and have healthy lifestyle, but I still extremely sensitive CO0 and without proper ventilation I just can't work efficiently if at all. So any room I stay in will always have open window no matter if there is AC working or not.
So it's great that you don't have same issue, but please keep in mind that people not just imagining things. Yeah they can quit smoking and watch their health, but CO2 is still extremely important. If you don't believe me rent some air quality meter and check it for yourself.
Thanks for the detailed response. I don't deny that it might be difficult for someone to work without ventilation. I simply find the air outside extremely irritating to my throat, nose, and upper airways. This "fresh air" frequently gives me headaches, which rarely happen if it is kept outside. I wasn't prone to headaches at all before I started working here. I rarely open windows at home and feel fine. It seems to be a no-win situation, and the better solution for us is to part our ways, which I am working on.
Sorry, I've spent enough time on the internet to know never to share any specific personal details.
It's cheap low-quality coal with high sulfur content. The level of air pollution it produces is insane. In winter time you can feel the taste of coal on your tongue. The level of visibility is like 150-200 meters, after that it's a solid grey wall. Nobody here cares though, and if you do, you pick up your things and move elsewhere. That's what I am currently preparing to do.
Pondering with rising wet bulb temperatures globally, whether Darwin will kick in and whether humans will evolve into a more heat resilient species. Natives in Nepal are adapted to lower oxygen, so why shouldn't that be possible?
It would require conditions where those who are better adapted to higher temperature produce more offspring than those who aren't, consistently over many generations. The Nepalese would have developed this change over thousands, if not tens of thousand of years. The climate crisis is likely to come to a burning head in the next hundred. In all likelihood, we'll use technology to circumnavigate the issue and it will never get to the point where humans are put into a position where adaptation occurs.
Evolution "cares" about one and only one thing - spreading of relevant genes. Technological advances means there's almost no difference in terms of procreation between more and less heat tolerant people.
In South Africa everyone is cold in the winter (if you are not in places like the KnP) due to bad designs and insulation, bad clothing perceptions and cultural factors.
A guy once told me that they install thermostats in corporate buildings that are actually disconnected from the system because people complain less if they can see thermostats in the room.
Basically, in the winter, we need to wear jackets while sitting at our desks in a place like JHB simply because the ventilation for some reason is pretty cold. I've heard that in European businesses you take off all your warm clothes indoors and just wear a T-shirt or other shirt.
There is also what we call it the "aircon" war. Usually black people (and myself) want the aircon off, but white people want it on. This is a joke though so don't take it too seriously (or racially). But this is mostly about when the aircon is cooling the building. When it is supposed to heat the building in a colder region most people will agree that the rooms are too cold, but for some reason we can't seem to learn this lesson.
By the way, for the sake of interest, Jan Smuts was well known for having a house with very cold winters and very hot summers. It is said that his guests would often comment on this when they stayed over. [1]
[1] Jan Smuts: Unafraid of Greatness. Richard Steyn
We have "aircon" wars over here in Romania, too. I'm now working from home but in the previous ~15 years of me working from an office one of the major stresses while moving in with new office colleagues was "will we be able to not kill each other in the summer over how the AC should work?" (I'm exaggerating, but only by a bit).
We also have what we call "curent" which is translated to "draft". There are many online articles about how foreigners view our (Romanians') obsession with "curent", this [1] is just one of them.
I think every office/home seems to have an air-con war between certain demographics. Our engineering company has started hiring more women, which I'm entirely for except for the fact the thermostat has seemingly crept up about 0.5C a day since the new hires have started coming in. I have cold showers every day, so anything above a cold breeze makes me sweat. I have the argument daily that you can wear jumpers scarves and hats if you're cold, but can't strip down any lower than a shirt when you're hot. Each day it still seems to climb...
I love a good Smuts anecdote, he was a fascinating person. I'm near Cape Town and I can confirm that it's also freezing cold here during winter.
You see more people investing in double glazed windows like they do in Europe. I've done this to my house and it works really well against both heat, cold and sound. Combined with an efficient fireplace, you can enjoy a toasty winter in the cold winter evenings.
I did it on an existing house. It's very simple and not too expensive. I went for uPVC double glazed windows, doors etc. For a typical 1x1.5m window you'd probably pay around R3500 including installation.
I have yet to experience an aircon war in Germany, but there is also a pattern where men want to open a window to keep the air fresh, while women find it too chilly and want to turn on a radiator instead (of course there are exceptions).
I can't get much done in radiator air, so that's one more reason to work from home, or in well-matched office rooms.
> men want to open a window to keep the air fresh, while women find it too chilly
This tends to happen a lot in environments where business attire is a must. It's harder to find a good compromise when some people wear business suit and tie and some wear skirts or open shoulders. And that's even before you factor in possible physiological factors (people have different comfort zones).
Force the women to wear suits too, or let the men wear shorts and a shirt with no tie.
If you've got menopausal women that's a problem too, and I think perhaps women and men just have different temperature preference distributions.
My kids teachers (all female) refuse to let the boys wear shorts except in Summer term because the teachers consider it not to be warm enough .. but they wear skirts and dresses all year .. which is weird.
Established norms on dress and room heating/cooling are pretty arbitrary I feel.
> perhaps women and men just have different temperature preference distributions
I bundled all that under "physiological factors". There are also sociocultural factors at play but that's a minefield of a discussion. It's how gender stereotypes are born and educated into kids from a young age.
Thankfully I've never worked in an office where suits were a thing. But this was a recurring issue even in a typical IT environment (jeans and T-shirts), and at school before that.
In Spain there is a regulation to only acclimate public buildings into the 21°-26°C range and no further (that is, don’t heat above 21° or cool below 26°), which I find rather sensible.
I think the range stems from a European regulation, although I’m not particularly sure.
For me, 20C is the point where I have to reach for a jacket, otherwise I start shivering from the cold and can't concentrate.
All air conditioner remote controls I've seen default to 24C whenever the batteries are replaced, so I believe 24C is the "standard" room temperature that AC manufacturers aim for.
The official guideline for sedentary office work in Denmark is between 20-22C.
If you begin shivering at 20C, put on a woolen sweater or the like. It's more comfortable anyway, because your body keeps the temperature that's optimal for it.
> The official guideline for sedentary office work in Denmark is between 20-22C.
That got me curious, so I went looking: the official guideline for Brazil is in the range 23C - 26C in the summer, and 20C - 22C in the winter (http://portal.anvisa.gov.br/documents/10181/2718376/RE_09_20...). But I question that division between "summer" and "winter": it's in the middle of winter right now, and the predicted high temperature for today is 33C (it's been feeling like summer for several weeks already).
> If you begin shivering at 20C, put on a woolen sweater or the like.
It seems wasteful to cool the air from over 30C to 20C, just to have nearly everyone wear heavier clothing, instead of cooling it to something like 22C where everyone can use normal street clothes.
> It seems wasteful to cool the air from over 30C to 20C, just to have nearly everyone wear heavier clothing, instead of cooling it to something like 22C where everyone can use normal street clothes.
Oh I agree, I was talking (not clearly), from my own context where we don't have to cool the office, but heat it, in that case it's the efficient choice.
I try to keep my home office no higher than 24c/75F when I'm working. Last year I had to work in an office where it was routinely 27c/80F and above, and my productively and mood was definitely affected. As you say, above that I get sweaty and lethargic. Of course there was someone in the office complaining that even that temperature was too cold.
The weird thing is I have no issues going outside in the heat - I used to live in the middle east and would walk outside at lunch, when it was 45c/110F+ - I just can't comfortably sit at a desk and work in the heat.
A few years ago I was living and working in an apartment where in the winter it was as low as 16c/60F. I much prefer too cold than too hot. I can combat cold just by wearing more layers, especially thick socks, and drinking more tea. Does that not work for some people?
You can always keep yourself warm by adding layers, but it's hard to look professional beyond a certain point. It helps somewhat if the clothing matches the weather, but I find it's often colder in offices during summer due to AC.
> New USB controller firmware, apparently. The fact that it can have any effect on the CPU temperature is a little puzzling.
At a guess it's something to do with power management and link idling. Version 0 of the firmware might leave the link up all the time with constant traffic; a later version gets it to power down, and along the way someone has to test and fix all the cases where it fails to power down or back up.
I know this is now totally unrelated to raspi but where do you live? Setting your AC to 20C seems freezing to me and I almost never set my AC below 26C (Tokyo)
I think it's an American thing. My workplace (Seattle) keeps the AC set around 72F (22C). In the summer, I walk to work in a t-shirt around 7AM, then put on a thin hoodie at work...
Feels weird to me, but college and people's homes are usually set to mid-70F too (or ambient temperature).
I hate this — yet it is very typical in southern European nations too.
The way I see it you always long for what you don’t have. That is why in torching hot places they tend to cool things really down and wear long pants and a second layer, while in northern nations people will run around in shorts and thin dresses literally the first time it seems to be a little bit summerish.
I think that big of a temperature difference is not only a tremendous waste of energy (that heats the rest of the outside at the same time), but it is also quite impractical: you always have to carry around a second layer if you are planing to stay somwhere for more than an hour. There you are, carrying around that hoodie, just because the rooms are cooled down to much.
In Tokyo, I'm not really trying to escape the heat. I'm trying to escape the humidity. So 26C sounds about right, and that's the typical setting I use there.
In California, I'm actually trying to escape the heat, which sometimes gets as high as 46C here. The setting on my home AC is 22C.
I guess my point is that your preferred AC setting can differ greatly depending upon the outside conditions.
>I'm trying to escape the humidity. So 26C sounds about right
>, I'm actually trying to escape the heat, which sometimes gets as high as 46C here. The setting on my home AC is 22C.
wait what? how does the weather outside have any bearing on what your comfort temperature inside should be? i mean, if you think 26C is comfortable, why would you need to turn it down to 22C in CA, where it's probably even drier?
In the US, 20-24C is a pretty typical AC setting, with commercial buildings usually closer to the cool end. I don't put on a jacket (other than for fashion) until 15C or so. For me, 26C would negatively impact my productivity.
Yeah, that's definitely true in my family at least. Radiating space heaters are one good solution to this problem since they allow you to warm up without requiring all the air in the room be heated up too.
The (northern) US, and previously the UK --- both places with mild and cool climates. In fact, in the UK (where the Raspbery Pi Foundation is, so this discussion isn't totally unrelated), 18C is considered a comfortable minimum to heat a house to, and it usually doesn't get much hotter than mid-20s in the summer. I suspect 20C (68F) is chosen in the US because it's a "round number" and happens to have an exact Celsius-Farenheit conversion, while also being within the typical comfortable temperature range.
Thus, it wouldn't surprise me if the RPis perform better at 18C than they do at 28C.
Heating to 18c is different than cooling to 20c. When it's hot you're typically wearing lighter cooler clothing and are more comfortable at hotter temperatures. When heating it's the opposite, you're wearing warmer clothes and better suited to cold.
Humidity also plays a big factor. In the southern US, most people have their AC set to between 23C and 25C during the summer and somewhere around 20C in the winter. In more arid regions the temperature settings creep up because you tend to feel cooler when it is dryer.
I think this is both regional and cultural. I’m in Ireland and tend to heat the house to about 16-18 degrees in the winter, which is on the low end of normal here (I think official guidance for workplaces is 18-22). But in Germany in similar weather, heating to about 25 degrees seems normal (I tend to be over there for a conference every December, and always find the indoor temperature pretty unpleasant).
Exactly, the comfort of temperatures is highly dependent on the ambient humidity. The normal human body temperature is a lot higher than a comfortable room temperature, so to be comfortable requires constantly dissipating a significant amount of heat. When the ambient humdidity is low, sweating works efficiently and you don't notice a moderately high temperature, but high temperature and humidity means sweating is inefficient and you end up overheating.
This law is great. I live in France and it seems to me that people are more and more wishing to have 18C when it's 35C outside and 25C inside when is 8C outside, which is completely crazy (and from what I understand from my US colleages, quite typical in the US - but I assume it depends on where in the US)
That kind of extreme heating/AC can sound like a luxury in an abstract sense, but it's not even practical IMO. When I go to countries that do that, I tend to get colds or sore throat from the sudden changes in temperature from outside to inside or vice versa, and I need to be putting on and taking off clothes all the time, which is hardly my idea of comfort.
I’m much more familiar with x86 hardware, but missing PCIe ASPM is a big deal. Some chipsets can’t properly idle themselves if any PCIe link is non-idle.
On my previous team (South Africa) I was located in an underground room with cheap air-conditioners set on 30 degrees celsius.
My eyes couldn't handle the dry air. It was unbearable. One of my colleagues was prescribed subscription eyedrops and another developed an eye condition.
After unsuccesfully trying to negotiate some sort of change to a more sustainable environment with the team of managers I evently started sitting in the above-ground building until I was repremanded and told I had no choice but to sit in that room.
Part of the reason I left that team was to protect my eyes and my health. Other people in the room thought 30 celcius was generous and that it should have been hotter if the aircon could do so. Seems utterly absurd to me.
There have been studies on office productivity based on temperature, as well as how it seems to differ between genders. Planet Money covered it fairly quickly[1], but they generally include sources so if you're interested you can dig deeper.
I have never understood this! My office has no AC and a window to the south. Temperature is 31C right now and probably will be 35C in the afternoon and I love it! :) 28C is barely OK and below 25C I will go and fetch a sweater. 20C is freezing cold.
I hope this does not mean that my hardware is suffering too much, though! I do underclock everything (Laptops) to around 700MHz, but still... looks like I'll stick to the Raspi 3 for now.
I WFH and I set my thermostat to 78F during the day. My office, because of the position relative to the sun as well as the equipment I have running gets about 83F during the summer. You get used to it. I find that it helps me tolerate the heat of the day better when I'm off work and out and about.
A jacket will not prevent the cold air from hurting a respiratory system. After 1 hour with such cold AC, I will get a sore throat, and come back home with a heavy headache.
It's pretty hard to code in a winter jacket, I've tried when being in an office set to 18°C. I absolutely can't work below 22°C and will refuse to work at those temperatures. (But then again, I'm a woman)
Luckily the french have a law saying that you shouldn't heat above 21°C and shouldn't cool below 26°C, and I'm a huge fan of this law, everyone should adopt it.
I took it to mean that the USB controller was doing double duty holding the now reflashable firmware for the main processor. Still a little weird, but makes a bit more sense.
> As someone who is used to a room temperature of 20C, 28C is above the point of idle sweating and becoming lethargic.
AC should not to be used for more than delta 8C. Maybe 10C. For Americans: if it's 104F outside it shouldn't be set to less than 86F. Setting to the customary 68F is very bad for your health.
Easy to say that when your summer is 2 months long with max temp of 30C something, try a 9 months summer ( most of south america, africa, south east asia and oceania ) with a max of 45C +.
We should turn off winter heating too then as that is probably as bad as Airco
BTW I am from a tropical country and I live in mid-northern Europe now, this argument not really fair, the amount of energy (electricity and gas) used here to keep people alive during winter is 10x larger than in my home country during summer.
This is one of the stupidest things I've read today. A lot of Europeans seem to claim that it never gets hot enough during summer to warrant A/C. Then they turn around and complain about "unusual heat wave" and can't work.
A/C is necessary in most parts of the world. I'm European as well and I can't work past 20C, which is why I have A/C and table fans to keep me cool.
And I'm European and can't concentrate below 22-24°C, because it's too cold. 24-26°C is my ideal temperature. (But then again, I'm a woman)
Additionally, having a too large difference between inside and outside isn't just annoying but actually dangerous and even painful when going from 18°C indoors to 34°C outdoors.
There's no contradiction there. Paying for an A/C system that only needs to work for one week per year during the peak of the heat wave is terribly inefficient. It doesn't mean you can't complain about heat especially since it affects far more than just your home (like being outside in any capacity - pedestrian, cyclist, driver, passenger).
Expect it is not a week. It is like 2.5 months of complaining. It has already started. People are talking about freak heat waves and breaking heat records, just like last year and year before that and the year before that and...
"Air conditioning was a most important invention for us, perhaps one of the signal inventions of history. It changed the nature of civilization by making development possible in the tropics. Without air conditioning you can work only in the cool early-morning hours or at dusk. The first thing I did upon becoming prime minister was to install air conditioners in buildings where the civil service worked. This was key to public efficiency."
What about heating system? It takes more energy to heat an apartment from -1° to 21° than to cool the same apartment from 38° to 25°.
For me it feels very strange to focus on AC when in Europe heating account for 33% of all CO2 emission.
The only thing that really matter if you want to act is building insulation, the only things that can change something regarding energy consumption (for both heating and cooling).
The climate change concern is certainly valid, but I don’t really understand your description. Surely it’s considered normal in Europe to have refrigerators and fans, both of which also match that description. It seems like overuse would be the issue with air conditioning, not just its purpose.
Hmm...I live in the USA in a place where nobody has air conditioning. Meanwhile when I visit Paris or Rome I look for an AirBnB with A/C otherwise I will get no sleep..
The Raspberry Pi 4 is really impressive hardware for the price. In these CPU benchmarks with heatsink attached, it slightly edges out the Jetson Nano developer board, which retails for $99.
Since one of the popular uses of the Raspberry Pi is to be an emulation box, this will mean it will now be capable of running most of the N64, Dreamcast and PSP library comfortably. The Raspberry Pi 3B+ could barely handle even select easier-to-emulate titles of these systems. It might even start running a handful of GameCube and Wii titles now using Dolphin. It also helps it to run tricks to vastly reduce input lag that troubles most emulation setups [1]. And the nice thing is that all of that should even work on the cheapest 1GB model.
It's even conceivable this can be used as a desktop in many ways. It is faster than some Chromebooks. However this will probably only truly enjoyable to use when using PXE for its main OS drive to have I/O performed on a server that can leverage SSD storage.
Mind you that the software is in a very pre release state, so none of this is really doable at the moment of writing. Especially the VideoCore (GPU and hardware video codec) driver seems to be in quite a sorry state and there's not even a hint of Vulkan support. The beauty of having a Raspberry Pi over another single board computer is that software support will come eventually though.
In fairness to the Jetson Nano, aren't you mostly paying for the CUDA graphics card? The CPUs in there are good, but I'll bet the Nano beats it in the GPU department (at least once the firmwares and drivers are super-optimized).
Though that may admittedly be cognitive dissonance, since I've been trying to talk myself into purchasing a Jetson for more than a month now.
Most likely yes. There is very little known about the VideoCore VI's capabilities, but it doesn't look impressive. They state OpenGL ES 3.0 support, which to me implies that it doesn't even support compute shaders (introduced in OGL ES 3.1). It looks like the hardware lacks features many modern graphics architectures can be expected to have. As a point of comparison, Nvidia has supported compute shaders in their architectures for eight full yearrs now. If I understand the spec sheet correctly Vulkan will never come to the RPi 4 because of hardware limitations.
On top of that, in basic initial benchmarks using OpenArena the GPU only shows a 50% improvement over the RPi 3B+, which is still using the same GPU architecture the RPi launched with in 2013. That benchmark doesn't test shader performance so it's not the best indicator, but it doesn't make the chip look any more impressive either.
I think Broadcom is not spending the R&D resources required to keep their graphics architecture up to date. It's a shame because the GPU of the original Raspberry Pi was really good for its time, and maybe the most impressive specification it had.
If you want to do graphics/compute tasks on an SBC the Nano (or just the full uncompromised Nvidia Shield TV) will undoubtedly be superior.
I have spoken too soon. Khronos lists the VideoCore VI as Vulkan compatible - latest Vulkan 1.1 compatible even [1]. Apparently compute shader support is being worked on for the VideoCore driver [2], which explains why it's missing from current reviews. Functionality to support compute shaders is not even present in current Mesa's git HEAD, let alone on Debian Buster's older Mesa version.
So it not having a modern featureset is a result of more driver issues. It's clear that this hardware has been a surprise release. Oh well, even then using an Nvidia Jetson gives you way better software support for graphics / compute tasks in the forseeable future.
The support is pretty unique in my experience. Purchased a different board (tinkerboard) and was susprised how much less info there is on the Internet, how outdated it is etc.
Def made me reconsider buying anothrr non rasp board even if stronger
From what I've read it consumes 1W extra over the 3B+. But it's understandable given that everything that could be upgraded was upgraded. Far more powerful CPU, USB/Eth, and possibly 3GB of extra RAM.
After some googling, it looks like VLI is the manufacturer that produced the USB controller chip [1] for the Raspberry Pi, so I guess that VLI firmware is the firmware that they provided for their chip.
I just tried out the Pi 4 yesterday and after a couple of hours of headless idling with no peripherals connected (just wifi) it was so hot, I almost burned my finger.
So this is definitely welcomed. But heatsinks or even fans will likely be necessary to do anything with the pi 4
I don't have a pi 4 to play with, but is there any chance you could try one of the mini heatsinks that just adhere onto the cpu, without any active cooling?
I'm just curious to see if that helps much, they're pretty cheap and could be useful other places too (I use one on the VTX for my drone)
You can check the temperature though the pi itself. From my experience depending on the material/length of contact, things start to get too hot to comfortably hold after ~55C/60C for most metals, but brief contact might barely feel warm.
Those are mostly useless (they only slow down temperature changes a little).
Only thing that really helps is active cooling / moving the air. You also have to figure out the placement of the board. The air will move via convection, and if you get air movement from a hotter part of the board (PMIC) to the SoC, you'll not improve things much.
I don't think perf gains in Rpi4 are of much use, if it idles at 65°C (my ambient temperature is ~32°C at home) and has to throttle down to <1GHz under persistent load.
> they only slow down temperature changes a little
That already helps a lot if you don't need to maintain full power for long periods. It gives just enough thermal inertia to handle spikes without throttling. For extended periods you may need to move to bigger heatsink or active cooling.
I'm curious what benchmarks you've seen that actually show a 3x compute improvement? Most of the benchmarks I've seen show a ~30% improvement in CPU compute, while I/O has improved tremendously, as well as the memory.
But again I've only seen ~30% improvement in CPU workloads, a far cry from "3x the performance" that I keep seeing.
For my use case, the main attraction for me in the RPi4 is the additional RAM and better IPC/ISA, not necessarily the higher clock (though higher clock is certainly welcome).
So some data is now in(Obviously with the old firmware, not the one mentioned in the article):
Without fan, idle: 58°C peak
With fan idle: 39°C (roughly 26°C ambient temperature)
With fan 4x parallism of `yes > /dev/null`: 51°C peak
The fan is a really small one, just for the CPU. And I didn't quite place in the center of the chip :/
Quite surprised of how much I overestimated idle temperature. 58°C feels really hot when you touch the SoC.
This obviously depends on the heat conductivity of the surface, but from my experience with my 3D printer, ~55C/60C is about the cutoff for pain tolerance for me. Just below that temperature is holdable, after that temperature is when it's painful to touch.
The fact that the RPi has writable onboard state/storage of any kind whatsoever (save for the fuses which are WORM) comes as news to me.
The SoC firmware is loaded from the FAT partition on the SD card, so there is no “flashing” (in the traditional EEPROM sense) of the system firmware. Apparently the USB chip has its own EEPROM?
On older Pi's, the Raspberry Pi had it's boot code completely burned into the SoC itself, permanently installed and unable to be tweaked post-launch.
With the Pi 4, this was no longer possible. The Pi 4 uses a much more complicated way of booting up as well as some RAM tricks inside the SoC to allow for more than 1GB of RAM to mostly work without redoing all of the SoC (e.g. the VPU and Hardware Codec Blocks still only sees 1GB of RAM, but they will never need more than 1GB of RAM, as far as I know). This Pi no longer uses the `bootcode.bin` file older Pis used for this reason.
Because of this, the Boot ROM is now a 500K EEPROM, which can be updated to allow things like USB boot, PXE boot, and similar to be added after launch. The USB Controller also has it's own EEPROM, which handles different things (I think).
> Because of this, the Boot ROM is now a 500K EEPROM [...] The USB Controller also has it's own EEPROM
Can these two EEPROMs at least be flashed through test points, so they can be recovered without de-soldering if they're corrupted? (The RPi4 schematics on the site are next to useless.) That would also allow dumping both EEPROMs to compare them with a "known good" copy, in case there's any suspicion they've been maliciously modified.
This is all documentation on the SoC EEPROM available. (I don't believe there is any USB EEPROM information publicly available at the moment.)
The good news: You don't need de-soldering to fix them. If you should corrupt the EEPROM, you can plug in an SD card with a special 'recovery.bin' file on it and it will re-flash the EEPROM itself. (This does mean that the Pi _technically_ has an extremely tiny boot ROM permanently installed in it still, but all it does is repair the boot EEPROM if it is damaged and start that boot EEPROM if it isn't damaged).
You can also write-protect the EEPROMs with a special soldiered-on resistor on the board.
It is very common for a controller chip (USB, ethernet, bluetooth, wifi, SATA, etc) to contain a firmware EEPROM.
Whether you can flash this firmware without an hardware programmer depends on the implementation. For embedded systems like the rPi, it is common for the engineers to implement peripheral programming capabilities from the main processor.
Most laptops have that capability too, on both my Apple and Lenovo laptops i've seen firmware updates happening. On generic PC motherboards, this is less common.
It comes down to the board designer connecting the programming pins of the peripheral device to the main CPU I/O.
To make things worse, AFAIK the USB chip is connected through PCIe, so unless the RPi has an IOMMU, it could in theory DMA to anywhere in the host memory.
"The ARM cores are capable of running at up to 1.5 GHz, making the Pi 4 about 50% faster than the Raspberry Pi 3B+. The new VideoCore VI 3D unit now runs at up to 500 MHz. The ARM cores are 64-bit, and while the VideoCore is 32-bit, there is a new Memory Management Unit, which means it can access more memory than previous versions."
An IOMMU serves a slightly different purpose than a normal MMU. The graphic on Wikipedia [0] sums it up pretty well. The lack of an IOMMU would mean that if the firmware on the USB controller was malicious in some way, it could access memory it would otherwise not have access to.
I don't know the details, but the memory controller is twice as fast for the SD card than the old Pi's are, and SOMETHING (I'm not sure for certain what, don't quote me) did change with the memory subsystem. This change is what allowed a workaround which allowed for 4GB of RAM for the first time (whereas before, the 1GB of RAM was a limitation throughout the SoC). So, it _may_ have a lite-IOMMU, but I'd open a forum posting for more info.
Edit: This could also be why the SoC uses an EEPROM. The RPF said that booting the Pi4 with this new memory system is _much_ more complicated than booting a Pi3, and that permanently burning the boot code into the SoC (like on the Pi3 and earlier) was not an option for the Pi4. Extremely high risk of backfiring.
As far as I know, the SoC supported higher SD card speeds all the way back to the first Pi and the original drivers even tried to use them. The trouble is, the older Pis couldn't switch to the lower IO voltages required to actually operate at those speeds. Early adopters actually had to track down older, slower SD cards in order to use their Pis because otherwise SD card access would fail due to the driver erroneously switching speeds.
You're talking about the early problems using Class 10 cards which has been solved for a long time now.
The Pi 4 upgrades the SD controller system (almost entirely) and is SD50 compliant, which means the SD card bus is roughly twice as fast as older models.
Basically all USB 3 host controllers run firmware of some description. Some can be updated in the field, some can't. You pretty much can't get away from proprietary code if you want USB 3, but in most cases it's hidden from you.
Yep, the same goes for any BT, WiFi or storage controller chip. You can download the specs for most protocols, you might even be able to the get datasheet for the controllers, but writing a new (FOSS) firmware for the chip would be a huge undertaking.
The spec isn't proprietary, you can download all 482 pages of it right here[1]. It's just that a USB controller needs firmware and none of the vendors making USB3 controller chips have open sourced their firmware. Most chip vendors are not exactly huge fans of open source and likely consider portions of the firmware to be trade secrets.
No, it's just complicated - easier to implement it on a reasonably general purpose core with firmware than to bet on getting it right first time in dedicated hardware.
The RPi has several proprietary binary blobs required for it to function, and always has. This has been one of my long-standing gripes about the platform’s security.
Unfortunately, there are no boards on the market (that I know of) that don't have any closed-source blobs. Also (unfortunately) it is probably going to stay this way because of legal requirements.
1) Broadcom has quite a bit of stuff they want hidden and don't allow any of their many customers (including the Pi) to open-source.
2) The Pi 4's SoC has a hardware H.264 block and hardware H.265 block. Older Pi's had a hardware MPEG-2, hardware VC-1, and hardware H.264 block. These blocks require royalties and are _heavily_ protected by MPEG LA. If open-source software was released that could control these blocks, you could possibly expect a huge lawsuit against the Raspberry Pi Trading arm.
3) To the Pi's benefit, the VideoCore is not open-source but is the most openly-documented GPU available in a mobile device. Most Pi competitors use Mali, which if you look at Hackaday's Pine64 Un-Review, works with almost nothing and has wretched documentation. There is some open-source work, but quite a bit remains closed or completely inoperable on Linux.
4) If you need platform security, look elsewhere. The Pi doesn't support secure boot, licensed code signing requirements, and the new Pi 4 allows almost anyone to update the USB and SoC EEPROMs just by putting in a new SD card and running a few commands.
>These blocks require royalties and are _heavily_ protected by MPEG LA. If open-source software was released that could control these blocks, you could possibly expect a huge lawsuit against the Raspberry Pi Trading arm.
I think you're confusing copyright and patents. Just because some technology is patented, it doesn't mean it can't be open sourced.
True, it could be open-sourced, but the RPT would no longer exist after the lawsuits are finished.
Edit: Even if the RPT wrote their own firmware, not from Broadcom, and released it, they would still risk being sued to pieces for patent violations galore. After all, if you license H.265 alone, there are over 1,100+ pages in the Table of Contents alone just listing all of the patents worldwide on it.
Edit 2: And they could be possibly sued by Broadcom for violating the IP sharing agreement used for developing the chip by developing an open-source blob software.
If I buy your patented thing from you, that exhausts your patent rights for that copy of the thing. If I modify it, and do something you don’t like with it, you can’t come after me under patent law.
Note that I didn’t say anything about contract or copyright law—they may have signed something saying they wouldn’t reverse engineer the chip, or the sale could come with an end user Eula to that effect.
You can find open-source firmware attempts by the community, but they haven't gotten much farther than booting the Pi.
I'm just saying that if RPT tried to write their own firmware, they could be in big legal trouble, so I wouldn't expect any open-source firmware from RPT anytime soon.
Also, be advised, that a one-line sed -i crack (open source software) is widely available for the bootcode of pre-4 Pis that universally enables the hardware codec blocks without a license.
It is “heavily protected” only in a legal sense, not a practical one.
Broadcom isn't really against anyone making their own firmware (and people have tried, but haven't gotten much further than booting). As for the parts that can never be open-source:
1. The decoder blocks, H.264 and H.265, Broadcom/RPT can never release the source code to this without violating potentially thousands of patents worldwide.
2. The GPU of the Pi runs a ThreadX RTOS, which is licensed, runs on "billions" of devices worldwide, and can never be open-source without being sued for, well, huge amounts of damages. (The GPU blobs will never be open for this reason, but perhaps a mostly-complete GPU firmware NOT based on ThreadX will be released someday by the community, but I wouldn't hold my breath).
Also, you know the Pi's new OpenGL driver? All it does is convert the OpenGL commands and stuff into commands for this proprietary ThreadX GPU software.
3. If you think this isn't great, it's still better than the Pi's competitor boards, which have missing gaps in functionality, proprietary blobs which don't work well or even at all, and open-source implementations that currently just suck. (Read Hackaday's Pine64 Un-Review to see what I mean. It's improved a bit since then, but this is why "Pi-killers" haven't caught on.)
Interesting. Thank you so much for the reply! Would it be feasible to engineer an open source firmware that doesn’t support any of the things you listed?
I use Pis as a stateless platform for an application that doesn’t use any graphics, GPU, or codecs. I imagine several other people do too.
> Also, you know the Pi's new OpenGL driver? All it does is convert the OpenGL commands and stuff into commands for this proprietary ThreadX GPU software.
You mean the old driver? The new one is based on Mesa.
I'm optimistic about the OLinuXino boards from Olimex. They're certified open source hardware by OSHWA, with all the bootloader code, schematics, and even CAD files on Github. I think the Mali firmware is the last holdout, but with the new Lima and Panfrost drivers landing in Linux we may soon have replacements for that too.
At least some of the Orange Pi boards don't require any blobs (if you ignore the boot ROM - which isn't user replacable in any case, and that's probably a good thing since it improves brick resistance). The open source drivers for the hardware video decode and graphics are a little bleeding-edge and require a very recent kernel and libraries though.
The Orange Pi uses Mali Graphics, which _may_ (I'm not certain) have a proprietary RTOS on it.
I do know that The Raspberry Pi has a RTOS for the VideoCore GPU, and it is based on ThreadX RTOS, which is proprietary and used on "billions" of devices. The RPF will never be able to make that OSS.
For (4) do you know of where else to look? Is there a board with a Pi-like price that does support trusted boot, or some sort of TPM-like remote attestation?
I work in a office where the air conditioner is set to 28°C
Wow. As someone who is used to a room temperature of 20C, 28C is above the point of idle sweating and becoming lethargic. While that's still within the range that most consumer electronics are rated for, I believe a room temperature of 20C is the normal design target.