spatular

I have 9360 with i5-7200U. I've just checked, fan kicks in after 90 seconds of sha1sum /dev/zero and stays always on lowest RPMs. 25C ambient. With 50% core load at 3.1GHz it stays silent most of the time...

Idea is to treat (ip, port) pair as single 6-byte endpoint address. Each endpoint would be able to run just one service (well, two, one for tcp, another for udp). Type of service would be determined by looking up PTR record.

On the downside you won't be able to easily tell if two services run on a same machine or not. SSH and HTTP endpoints would be related only through DNS.

I think doing it in compatible way that doesn't require replacing routers would require NATs on both sides, and client that looks up PTR records before connecting.

One thing is that monks usually exist inside religious organizations. Religions are highly traditional, do not concern themselves with western science and are slow changing.

The other thing is the question of what do they think is worth doing. Imagine a person who's been enlightened. He/she may have been miserable, depressed and suicidal in the past, but now has reached state of no suffering and is completely happy. Very likely goal of such a person is to reduce unneeded suffering of others. The most straightforward way is to help others reach enlightment.

That's the point. It's the mind that separates objects out from the global mess, from the single process that unfolds. It's a good thing in a sense that it allows humans to function in the world and create more humans. Downside is that it leads to lots of unnesessary suffering when things change. And they always do, all the time.

I'll add that normal people act mostly out of desire to have something or to not have or experience something. After completion of 1st "path" Self is seen to be an illusion, after 2nd path craving and aversion are greatly reduced. Saying that it affects motivation is a bit of understatement, since it completely changes frame of reference.

About people getting unwanted outcome, dissolution of self is known to be unpleasant (aka "Dark Night"), but it's a temporarily unpleasant stage. I've not heard of anyone regretting completing 1-st path after it's done.

Hmm, on my XPS 13 (9360) fan kicks in only after about 5 minutes with one core fully loaded, and it's not that loud. It usually happens only during Hangouts videocalls or heavy dev work. Browsing / playing 1080p videos / text editing / gimp is not enough to trigger active cooling. But I use linux. Maybe he had windows install with antivirus or whatever hogging CPU...

XPS's Atheros QCA6174 works fine with recent FW. I use it in AP mode in 5GHz band in ac mode, though I had to adapt a kernel patch from OpenWRT that ignores frequency regdomains. I've also tried to use Intel 8260 for this, but with no luck.

I have XPS13 with Killer wifi (Atheros QCA6174). At first there were stability problems but they were completely solved with FW update from Atheros's github repo and a kernel update (to >= 4.9 I think).

Later it turned out that an Intel 8260 card I've bought to use in PC-based AP cannot be coerced to work in 5GHz band no matter what I did, so I've swapped them, now Atheros is in AP and Intel is in XPS, with everything working flawlessly. So it turned out that the Intel card is less functional...

I was making 25$ per day trading manually during a specific 20-min time slot for about a year. That wasn't simply by chance -- nearly monotonic increase in total earned sum with 2-week averaging during the year. Though I didn't hedge against black swan events, and I'm not sure it would be profitable if I've tried to. That said my understanding was that nobody else cared to take those money. I've eventually lost all intrest too since it was impossible to scale.

That was not algorithmic trading, but maybe-could-be-possible to automate.

Is there any info if linux compilation segfault (kill-ryzen) has been fixed?

Edit: and what is ECC status? Still not officially supported?

I don't have windows installed, but it's highly unlikely that it'll help. This feature is baked into panel firmware, and while Dell releases firmware updates for some models, there is no official fix for 9360 FHD. For models with the fix there are two versions of firmware: with CABC support and without. Switching requires firmware reflash with a special vendor tool. So it's unlikely that there is a dynamic setting to switch it off.

I also own 9360 with 1080p panel for half a year.

The only thing that bothered me was adaptive brightness (CABC) that was fixed in my case with an unofficial panel firmware at some risk to brick the device. Coil whine is perceptable only if I keep ear close to the keyboard. No other problems.

If there is someone from Dell here: please fix CABC! Unofficial patch proves that it can be done. It's especially important for linux devs who switch between dark terminal windows and browser: CABC increases backlight brightness if contents of screen are light, and decreases if it's dark. So, I switch from terminal to browser, brightness jump is already unpleasant as is, but then CABC kicks in, increases brightness and burns out eyes. Switch back to terminal -- too dark, difficult to see anything. How it's even supposed to work... Another unpleasant thing happens when CABC starts to adjust brighness back and forth while I edit code, it's highly annoing when it happens.

Edit: There were also a problems with Atheros WiFi, but new firmware from Atheros's github fixes them completely for me. Works stable since June 2017.

You can also lower switching threshold, in my experience it helps:

echo 50 >/sys/devices/system/cpu/cpufreq/ondemand/up_threshold

Yes, looks like this. And in addition in average human this pathway is limited to 300g amino-acids -> glucose per day if I remember correctly. Still what should happen if protein intake is tripled say from 50g/day to 150g/day? Excess amino acids have to go somewhere...

Well, fat digestion seems to be effective even comparing to hydrocarbons.

I think longer hydrocarbons including sugar have to be broken into fructose and glucose during digestion. Fructose cannot be immidiately consumed by cells, so it's converted into glucose in liver. On top of that, there are about 5g of glucose total in bloodstream at any given moment, so exess have to be accumulated by liver/muscle cells that have to store it in a long-chain hydrocarbon form (glycogen). Each conversion requires some energy, so 5-15% isn't that much...

This isn't so. Quoting [1], energy dissipated as side-effect heat during gestation and storage:

Carbohydrates: 5 to 15% of the energy consumed

Protein: 20 to 35%

Fats: at most 5 to 15%

[1] https://en.wikipedia.org/wiki/Specific_dynamic_action

To expand a bit on construction: it's possible to take a long tube, put a fresnel lens at front, pyroelectric sensor with rotating shutter at the other end. Scan by rotating it with reductors/steppers as a whole. It's also possible to cool tube with contents down to -70 with dry ice to cut on background heat radiation. Maybe it'll lessen noise in pyroelectric sensor, maybe not (and it's quite easy to buy dry ice in Moscow).

Have you tried zone plates instead of lenses? Feature sizes should be larger for IR than for visible light, so it's should be possible to fit enough zones to get some focusing.

It's quite possible that in SPEC benchmarks they used 2 Elbrus cores and 1 Pentium M core, so IPCs are about the same. And it's confirmed by 7zip benchmark.

In theory Elbrus can dispatch 23 non-SIMD ops per cycle, it has only 6 integer units and 4 among them can do FP, so it's not clear what remaining 17 operations can do. Modern x86 CPUs have about the same number of integer/FP ALUs.

Current IPC is quite good, but it's hard to tell if core clock can be increased without lowering IPC. If ALU pipelines are short, it may be difficult to increase frequency without adding more stages. If more stages are added then pipeline flushes due to unpredicted jumps will result in heavier penalty. And it looks like Elbrus doesn't have a branch predictor at all. Pipeline stalls caused by memory access will also increase with CPU frequency. In x86 CPUs they are somewhat masked by out-of-order execution and SMT, which Elbrus doesn't have, so it should depend on memory preload instructions inserted by compiler.

Sure. Just in this case variant 2 is very unlikely. I'd say getting sources after buying one or two Elbrus-based servers would be a success for a company not affilated with government.

Company I work for once tried to get linux kernel sources for an embedded system produced by another russian company. We got just honest "we have our proprietary module in our linux tree, so we won't be giving you any sources. Still, we are nice enough to recompile it for you with options you need." Ugh...

Translation is correct.

They also state that in realtime mode it's possible to set up different modes of external interrupt handling, "computation scheduling", "disk i/o" and "some other things", whatever that might mean. Kernel is modified to support realtime operation too.

Well, full translation would be "hardware support for loops, including pipelining". How exactly is pipelining implemented is not clear.

It's stated on official site [1] that it's based on Linux 2.6.33, and it looks like kernel and userspace are compiled for Elbrus ISA and run natively, without x86 emulation.

There are no links to sources on their site, and they don't provide datasheets. To request sources under GPL you need to get binaries first. I live in Russia, and I've never seen Elbrus in real use anywhere. It's not marketed or sold to general public. I think target market is government's security agencies. Of course they get sources anyway for audit and have no incentive to publish them.

[1] (russian) http://www.mcst.ru/os_elbrus

Here is more information from official site [1,4,5]:

- native "Elbrus" ISA or x86 ISA,

- Ebrus ISA is VLIW, can dispatch 23 operations per cycle (33 with SIMD), in-order execution,

- it's stated that x86 code translation + register allocation is done in HW, but later they write about a software translator and full-system emulator,

- 6 ALUs (all support integer operations, 4 can do FP),

- 256 x 84-bit register file,

- hardware support for loops, including pipelining,

- some kind of module for async mem preloading,

- speculative execution and branching predicates,

- "4S" model has 4 cores,

- 800 Mhz core clock,

- 64 KB L1, 128 KB L2, 8 MB L3 (shared between cores),

- 3 DDR3-1600 interfaces, ECC support,

- 3 x 12GBytes/s inter-CPU links, support for up to 4 sockets,

- 65nm process, 380 mm^2 die size, 986e6 transistors,

- software is based on Linux 2.6.33 and Debian 5.0 with more than 3000 packages.

There are some benchmarks for older chip model "2S" (overclocked to 500MHz, 2 cores) [2,3]. FP performance is about 1-5x of Pentium M 1GHz (1 core?) depending on benchmark, integer performance is about 1x. New CPU, "4S", should be 3 times faster than "2S".

[1] http://www.elbrus.ru/arhitektura_elbrus

[2] http://www.elbrus.ru/files/535269/9f0cd8/50606f/000000/2014-...

[3] http://www.elbrus.ru/files/535269/0e0cd8/50586f/000000/2014-...

[4] http://www.mcst.ru/mikroprocessor-elbrus4s

[5] http://www.mcst.ru/mikroprocessor-elbrus4s-gotov-k-serijnomu...

--

Edit: loop pipeling, OS information

In near field zone (low frequences, mains and DC-DC converters) the device acts like a secondary coil of a transformer, so it drains power from some other device.

In far field zone (high frequences, bluetooth, wifi) it captures part of already emmited wave, so it doesn't place additional load on emitter directly. But since device blocks wave propagation through itself if placed between transmitter and receiver it will degrade reception. And that in turn can result in automatic increase of transmitter power (in case of GSM phones at least). Which will increase transmitter consumption.

I've got my Crossover 27Q a month ago from this seller: http://www.ebay.com/itm/CROSSOVER-27Q-LED-27-2560X1440-PC-Mo...

Fully working, no dead pixels, slightly non-uniform backlight (it could be seen only on black background). The display is really great for the price.

The only downside for me is glossy screen, but it seems there are no cheap 27'' korean matte displays.

I've been wondering what's the point of online courses, at least in disciplines, where there are good books available? E.g. you could very well get the same (or deeper) knowledge of AI by reading Russell & Norwig book, or algorithms from D. Knuth books. And text is mostly always is a better format than video lectures.

As I understand, participating in a course will put some pressure to complete most exercises and actually finish the course. You would also get some automated submission checks and could extend CV in the end. Also courses might mention recent discoveries that didn't find their way in books yet. Is that all or I'm missing something?

I doubt that you could use NAND Flash as RAM substitute as it is right now. Basically it's a block device with page size about 8KB and erase block ~ 2MB. Data could be read and written only in pages, in-chip page load times are on the order of 100us, page write 1ms, block erase -- 5ms, and it takes about 100us to transfer page data in either direction. So, random read/write latencies are 10^2 / 10^1 times lower than HDD's, but still about 10^3 / 10^4 times larger than DRAM's. I guess internal architecture could be redisigned to reduce block sizes and allow parallel operations inside a sigle chip, but it's unclear how much it will affect density and price.

But it would be useful to put NAND controller closer to CPU, remove block device emulation layer and let OS do wear-leveling and block allocation. We could get high-performance swap or put there specialized MTD filesystems like UBIFS.

Actually it's quite easy to bruteforce by hand if you mention that point set has multiple symmetries. For example, you only need to start drawing lines from (0,0), (0,1) and (1,1). Any other combinations are derived by rotation and mirroring. After several minutes you can prove that solution in impossible under implicit presumption that every line must terminate at a dot. And after that puzzle is really easy.

So, as I understand, this is also a main principle of most modern sciences: outcome of a proper experiment must be reproducible. In other words, given enough detailes that matter, it's possible to predict behaviour of the system.

Also if an Enlightened being sees the world as it is and gets perfect understanding directly, he(she) can make what is generally considered scientific breakthroughs with ease. Does it really happen?