Posts by Wilco 1 224 publicly visible posts • joined 24 Sep 2009
The screen seems the same. Having just 2GB will save some power. Also it is a dual core Silvermont Atom so it is no surprise it uses slightly less power as you get less CPU/GPU performance than the Exynos variant.
I'm currently eagerly awaiting a Chromebook with the new 20nm 64-bit Exynos octa to replace the Exynos dual Chromebook I use for travelling, benchmarking and GCC development. That should give a 2-4x speedup to large builds and benchmarking runs!
Cortex-A57 is actually pretty quick for a 3-way OoO design - about 30-50% faster than Cortex-A15 (similar gain vs Jaguar and significantly faster than Silvermont). So no, it won't beat Xeon on single-threaded performance, but that was not its goal. You could replace a low-power quad Xeon with the 8-core Seattle and get similar performance at lower power and cost. As Calxeda showed, you should be able to cram many sockets onto a server board and get significant gains in sever density.
I agree the custom ARMs coming from AMD, QC, Broadcomm, Cavium, NVidia, Samsung, AMCC soon are very exciting. Broadcomm for example is claiming 90% of single-threaded performance of Haswell and something like 150% of Xeon performance per socket. If they pull it off, that changes the landscape a bit - Intel beware, ARM is coming!
Arctic summer ice extent was smallest ever in 2012. See http://climate.nasa.gov/key_indicators/
So what exactly is compensating sea level rise or increase in CO2 levels already observed? If there was any compensation, then there could have been no rise at all, right? So there is no compensation.
Or are you claiming there will be a tipping point where some unknown negative feedback will kick in? Go ahead, tells us all about it - it will make you rich beyond your dreams if you were right. 7 Billion of us are depending on your ideas. We're waiting...
Simples. The huge quantities of glacial meltwater, having no salt, reduce the salinity of the sea. That means the melting point increases, leading to more sea ice forming. So the more land ice melts, the more sea ice is formed locally.
In other news, Antarctica land mass is at its minimum since recording began, and sea level rise is at 3.6mm per year (that's 3.6m in 100 years, assuming the rate doesn't increase any further). Eg. http://climate.nasa.gov/key_indicators/
Sure, DC distribution in homes would require a major overhaul. It wouldn't replace 240V AC for high power appliances of course, however there is no reason why any appliance below 100W couldn't use 12V or 24V by default. You can take inbetween steps, for example replace the dozens of inefficient DC chargers and transformers that a typical home has with larger, more efficient ones that can safely supply multiple appliances with a standard voltage. That sort of thing would be useful just to reduce the amount of cable clutter from all the different incompatible chargers.
While off shore windfarms are certainly more expensive, they also produce far more electricity than an equivalent on-shore windfarm (40% capacity factor for London Array and Greater Gabbard, the 2 largest off-shore windfarms in the world - on-shore wind farms like Whitelee are about 25%). You avoid the NIMBY argument, loss of land and produce more power, so I would say it is more than worth it. And with several large off-shore windfarms running for over a decade, the idea that they rust away in a few years seems crazy - ships last many decades as well after all. Also there are still major improvements being made to reduce costs further and improve efficiency.
Improving the grid is certainly a good plan. DC would be great even in homes, most electric appliances use 12V DC internally, so all the conversion steps (with 60-70% efficiency transformers) are just crazy. All mains LED lights for example have a built-in transformer. Adding a single 90+% efficient 12V or 24V transformer in all homes would avoid a lot of wasted energy.
As for Thorium and fusion, those will be interesting in the future - there is still a lot of research needed before they become practical reality. Ignoring the whole nuclear=dangerous debate, today modern nuclear has a hard time competing with renewables (check out Hinkley C cost per MWh).
The applications will still be familiar, they are simply touch based ports of the desktop versions. For example Surface RT and Surface 2 run all of Office, so it's not like you're being forced to use something completely different instead and relearn everything.
The key issue is that applications originally designed for desktop use (mouse & keyboard) won't run as well on a tablet with a touch based GUI. So even if you were able to run your desktop applications, you'd probably still prefer the Metro versions of them.
iPhone/iPad can't execute OSX binaries either. Android could run Linux binaries (the kernel does exactly that) but doesn't allow users to do it by default.
So I don't see your point. As long as you can run applications, and read/edit all the common file formats, does it really matter at all whether you can run the exact same binaries across different devices? Do most people want to install SQL server or Visual Studio on their phone/tablet or do they just want to display/edit powerpoint slides? I suspect the latter, not the former.
launched back in February and not mentioned much besides in the recent AnTuTu benchmark cheating debacle... There was talk that ZTE would launch phones based on it but nothing has happened since. I haven't seen recent x86 phone market share figures, but I very much doubt they are gaining share.
"Intel also claims that the power needs of its next-gen SoC should be comparable to its ARM-based competition"
Maybe, I haven't seen detailed power consumption tests. However it's performance is certainly lacklustre compared to the Galaxy S4: http://browser.primatelabs.com/geekbench2/compare/1979365/1970335
What is new though is that Intel managed to optimize their compiler for an independent closed-source benchmark and convince the developers to use their benchmark busting compiler without publicly saying so. Then they paid ABI Research to do the dirty work and make the false claims. Presumably the "leaks" of Bay Trail AnTuTu scores were done on purpose as well.
It means that we have to take all future benchmark results where Intel appears to do well as possibly cheated in the same way, even if the benchmark appears to be independent initially. AnTuTu is quite tainted now, but I suppose that never stopped anyone before from relying on a bad benchmark...
GCC's vectorizer is not perfect indeed, but none of the benchmark in question benefits from vectorization - it's 20 year old code! It would have been better and more accurate to use Dhrystone, at least that's a benchmark that cannot be broken with dubious compiler tricks.
Well to completely counter your point with overwhelming evidence, my phone lasts at least a week despite its battery dying (it used to be 2 weeks). Yes it has an ARM. So ARM wins again. QED.
Obviously battery life depends on how you use your phone, so either anecdotal data point is useless. Anand's battery life tests showed Intel phones like Lava XOLO having mediocre battery life.
As for Intel not even trying - well if they have to cheat benchmarks and power consumption tests to pretend to be the fastest/most efficient then clearly they are losing the battle. If they were really faster and more efficient then their phones would sell like hot cakes. In 2012 Intel had just 0.2% of the smart phone market, or in total about 1.8 million devices. To put that in perspective, total phone sales were 1.7 billion, and Samsung sold over 1 million phones every single day of 2012.
Even comparing Watts is meaningless if the CPUs don't run the same task. You really need to compare energy to complete a task in order to do a fair comparison. A CPU that runs twice as fast while using 50% more power will complete any given task using 25% less energy. This is especially relevant in the ABI comparison as both the A15 and Krait cores used are significantly faster than Atom (unless you rig the benchmark of course).
In this particular case it was unreasonable to switch to ICC as GCC is the only compiler that is used on Android. So even if ICC was genuinely better than GCC (and that is questionable - ICC is regarded as a benchmarking compiler, not as a production compiler as it doesn't beat GCC or VC++ on typical code), then it still wouldn't make any difference as nobody uses it to build Android and its applications.
There was also the issue that Intel somehow persuaded AnTuTu to change the compiler for x86, and allow them to tune to the benchmark, including better compilation options and special optimizations - the ones that turned out to optimize much of the benchmark away. The ARM version of AnTuTu is compiled -Os and without inlining on an old version of GCC, so changing to -O3 and a newer GCC version should make a huge difference.
So to be fair, has ARM been given the opportunity to do the same optimizations as Intel? If not, then it was effectively cheating. If money changed hands then this is something that should be referred to the FTC as it would count as anti-competitive practices - something Intel did in the past.
Definitely avoid AnTuTu, even Dhrystone is better, more accurate and less easy to cheat...
Geekbench is far more trustworthy as it is done professionally and uses a large set of fairly standard benchmarks. While it is not SPEC, my experience is that it correlates reasonably well with actual CPU performance. It is not without issues however, 2 of the FP benchmarks accidentally use denormals which causes the scores to be slower on some CPUs with slow denormal handling. This will be fixed in the upcoming Geekbench 3.
In my experience every single benchmark has flaws, and that includes Dhrystone, CoreMark, SPEC, EEMBC. Most of them are easy to spot, others require more analysis, but over the last 20 years I've come to the conclusion that the perfect benchmark doesn't exist. Relying on a single benchmark number is foolish, someone may already have broken the benchmark. It invariably happens...
No 3.7V is not standard. There are different kinds of Li-Ion batteries, and commonly used ones vary from 3.6V to 3.8V nominal. It's important to understand what nominal voltage means - it is simply the average between the minimum and maximum voltage. Actual voltage varies from ~4.2V when full to ~3V when empty. Also the battery age, temperature and current draw affect the voltage.
So no, one cannot just measure the current and assume voltage remains a constant. I'd say showing just current is admitting you're an amateur. To measure power consumption accurately you need thousands of samples per second.
For Coremark the most recent certified scores are 9.36 Coremark/MHz for a dual A15 and 6.61 for a dual core / 4 thread Atom N2800.
In addition to what you said, check the Atom die size: http://chip-architect.com/news/2013_core_sizes_768.jpg It seems to me Atom is a bit more than 1% larger than A15.
That's the "ISA doesn't matter" myth debunked once and for all.
The benchmarks appear to show that a dual Atom can beat a quad A15 on CPU performance. That's quite a feat considering Atom is a 5-year old 2-way in-order CPU while the A15 is a modern 3-way aggressive out-of-order CPU!
However independent benchmarks show a completely different picture: a Galaxy S4 leaves the K900 in the dust as you'd expect from the microarchitecture comparison: http://browser.primatelabs.com/geekbench2/compare/1979365/1970335
So that suggests something is going on with the chosen benchmarks. There are a million ways to cheat with benchmarketing. If the quad A15's somehow have to do more work then it is no surprise they burn more power doing so...
Also the results only show current, which means nothing. For a power efficiency comparison you'd have to measure Watts, and even more importantly Joules (ideally just the CPU, not the whole phone as in this case). Total energy to complete a given task is what matters.
Yes comparing quad 1.6GHz A15 with 2GHz Z2580 is quite reasonable:
1. Both support 4 threads of execution
2. Hyperthreading gives a significant speedup on Atom, around 40-50% extra performance, so a single core can behave almost like a dual core.
3. Most ARM cores are quad cores, while Atom is primarily single core with some dual core variants, so we compare with the rare latter ones - the Z2580 as is the fastest mobile/tablet SoC.
4. Current Atom cores are huge, 3.5 times larger than A15, so we'll never see a quad-core version of Z2580: http://chip-architect.com/news/2013_core_sizes_768.jpg
You can also compare only single threaded performance if you prefer core-for-core IPC comparisons but then Atom looks even worse.
Antutu is a rubbish benchmark, no native code is used, it's worse than Dhrystone or Sunspider, so I wouldn't conclude anything from it. Again if you want to compare CPUs, use native benchmarks.
You've got that the wrong way around. The only benchmark that makes Atom look vaguely competitive is Sunspider. The reason is due to Intel adding software optimizations to make the score look better. When you look at actual CPU benchmarks such as Phoronix or Geekbench then Atom trails far behind Cortex-A9 and even further behind Krait and Cortex-A15.
A9-based Calxeda server chips have been available for a while. But for a rough comparison with modern ARM cores, Tegra 4 scores 4582 on Octane, so a Cortex-A15 is quite a bit faster than Centerton on single-threaded code. At 8.5W the current Centerton has no chance of winning the power efficiency battle.
The x86 complexity certainly wastes power, which is more noticeable in low power designs and less so in high performance cores. This shows the relative sizes of ARM and x86 cores: http://chip-architect.com/news/2013_core_sizes_768.jpg. Jaguar achieves similar performance as Cortex-A15 but needs twice the area. Atom needs a lot more area in order to achieve low voltage / low power operation, and it's huge die size explains why we won't see quad core versions until 22nm. So yes there is definitely a big cost to x86 - few companies have succeeded making competitive designs. With ARM it is far easier to design a high-end CPU.
There is no "other 60%". Capacity factor means you get on average 40% of the max capacity throughout a year. On average. Ie. it's not 100% for 40% of the days and 0% for 60% of the days but far more averaged. Obviously there are low-wind days across the UK, but in those cases you simply turn CCGT stations a bit higher. These are required anyway in order to follow the daily demand (nuclear and coal can't do this). Wind power simply reduces the amount of CCGT required (and to a lesser extent coal).
You can clearly see this happening in the last few days: http://www.gridwatch.templar.co.uk/index.php
See how the monthly nuclear/coal/ccgt graph shows much lower CCGT/Coal is from April 13 onwards? Next check out what wind power did in those days. Also remember the recent El-Reg scare story how we almost ran out of gas? Now you know why we didn't...
Talking about recent cooling: check
AGW must be false because of the recent cold weather: check
Conflating weather and climate: check
From http://climate.nasa.gov/key_indicators/
CO2 increasing: check
Land ice melting: check
Sea ice melting: check
Sea level rising: check
Temperature rising: check
Checkmate.
Yes, without any subsidies solar PV would become competitive eventually, but it would take much longer that way. If you dislike renewable subsidies, do you also dislike all the oil, gas, coal and nuclear subsidies? The tax payer will be paying countless billions just for the cleanup of the current nuclear generation...
What source of energy can generate cheap and plentiful energy? Certainly not nuclear (not cheap, and we never got the promised "too cheap to meter"). What else? Fusion is still at least 50 years off...
We're going to face increased bills no matter what. Coal, gas, oil are all becoming scarcer and more expensive. You can't blame increased bills on renewable energy when Russia decides to charge twice as much on the gas we use.
Ledswinger, it's very simple really - wind, solar and nuclear always want to sell all electricity they can produce irrespectively of the market price. If you disagree with that, you simply don't have a clue despite trying to claim to know better. And no, I don't care who you work for. But can you show us some of your publications? Now that would show you know your stuff.
Nuclear is NOT competitive on the grid without subsidy. EDF tries to get a subsidy for 40 years at TWICE the current market rate for Hinkley C. That just says it all.
You're wrong there, wind power is very reliable (it's unlikely wind is suddenly going to disappear forever is, it?), and quite predictable days and weeks in advance. Yes you cannot 100% guarantee it is windy all the time, but a recent study showed that just placing wind farms all around the UK reduced intermittency significantly.
Interconnects will play an increasing role in the future, and they don't need to be very far either. Connecting to Iceland would open up a large amount of geothermal power, Sweden/Norway/Finland gives you hydro and pumped storage.
It's not just the anti-nuclear parties or the NIMBY's, EDF wants twice the market rate for electricity for 40 years for the new nuclear power station Hinkley C. That's a price where even the current government is saying that's a pretty huge subsidy to a private company (especially given wind subsidies last for just 15 years).
Renewable technology has become far far cheaper in the last few decades due to increased volume and efficiencies. For example solar PV panels are now just £1/W in the UK! That's becoming cheap enough that they pay for themselves through reduced electricity bills even without any subsidies. And it appears the relentless downward curve has no sign of stopping any soon. Efficiencies are constantly increasing and new discoveries (even thinner films) are being made regularly .
When there isn't much wind obviously production is low. At the moment every GWh produced is consumed immediately, basically no storage is required unless capacity is larger than consumption (and we're a long way off reaching that). For storage solar thermal looks like a better option, but it's not really feasible in the UK climate.
In what way is the number of wind farms vs other plants relevant? Wind farms are typically smaller, even the largest is currently 0.5GW.
Nuclear plants have their downtime as well. It can take many months to refuel for example and during that time you must have a backup.The exact same argument applies to all power generation, all have downtime for expected maintenance and unexpected failures, so you always have to have some kind of backup.
As one adds more windpower to the grid and more interconnects, it becomes more reliable - it is typically always windy somewhere!
No, it's a fact. Wind power costs are coming down, nuclear costs seem to only go up. As I already mentioned, the electricity price EDF wants for nuclear power is twice that of the market rate - higher than on-shore wind power. And if you are going to mention subsidies, what about all the nuclear subsidies, cleanup and storage costs? Nuclear subsidies are an order of magnitude higher and have been going on for many decades.
Wind power and nuclear power are actually fairly similar in the way they work. Both are expensive to build but have low "fuel" costs compared to gas and coal. As a result, both want to sell as much electricity as they can generate. That's logical, if you think it is bizarre for wind power, do you also believe it is bizarre for nuclear?
The fact is that with modern generation, your "merit curve" no longer applies. It only works for fuel based generation.
If we aren't already past peak oil, we certainly will be soon. Even shale oil will be a short fad, as each well only lasts a few years. But the real issue people always ignore is the cost of extraction, which is rising fast. Just because we've got large reserves of dirty tar sands in Canada for example doesn't mean it doesn't take huge amounts of gas and electricity to actually extract it (and refine it if you want low-sulfur petrol). So not only does it take a lot more effort to extract it, but in the process of extracting it, we actually waste a significant proportion of it as well. The end result is we get an ever smaller amount of actual useful oil for every barrel of oil extracted. That's typically not shown in the production figures...
Actually nuclear is more expensive than wind power. The £14B new nuclear power station Hinkley C to be built in the UK will produce power that is twice as expensive as the market price, more expensive than on-shore wind power and similar priced as off-shore. Yes, modern, efficient off-shore windfarms like Greater Gabbard (40% capacity factor) are cost competitive with nuclear power. That's the hard reality.
Note wind turbines have been producing over 10% of total UK electricity over the last week: http://www.gridwatch.templar.co.uk/index.php
There has not been any pause in warming. All the measurements show the temperature increase is accelerating in the last 3 decades. Temperatures haven't gone down either - every decade has been consistently warmer than the previous decade. And all other climate indicators show steady loss of land and sea ice, retreating glaciers, increasing sea levels and lower pH levels. See for example: http://climate.nasa.gov/key_indicators/
You're entitled to stick your head in the sand and pretend nothing is going on. But I will laugh at you for being a denialist, especially when the reality turns out to be worse than any predictions so far.
Anyone with any sense of intellect will understand that spewing out an ever increasing amount of pollution is going to do serious long-term harm to our environment and health.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
