I am just new with this APIs of the Nvidia and some expressions are not so clear for me. I was wondering if somebody can help me to understand when and how to use these CUDA commands in a simply way. To be more precise:
Studing how is possible to speed up some applications with parallel execution of a kernel (with CUDA for example), at some point I was facing the problem of speeding up the interaction Host-Device.
I have some informations, taken surfing on the web, but I am little bit confused.
It clear that you can go faster when it is possible to use cudaHostRegister() and/or cudaHostAlloc(). Here it is explained that
"you can use the cudaHostRegister() command to take some data (already allocated) and pin it avoiding extra copy to take into the GPU".
What is the meaning of "pin the memory"? Why is it so fast? How can I do this previously in this field? After, in the same video in the link, they continue explaining that
"if you are transferring PINNED memory, you can use the asynchronous memory transfer, cudaMemcpyAsync(), which let's the CPU keep working during the memory transfer".
Are the PCIe transaction managed entirely from the CPU? Is there a manager of a bus that takes care of this?
Also partial answers are really appreciated to re-compose the puzzle at the end.
It is also appreciate to have some link about the equivalent APIs in OpenCL.
What is the meaning of "pin the memory"?
It means make the memory page locked. That is telling the operating system virtual memory manager that the memory pages must stay in physical ram so that they can be directly accessed by the GPU across the PCI-express bus.
Why is it so fast?
In one word, DMA. When the memory is page locked, the GPU DMA engine can directly run the transfer without requiring the host CPU, which reduces overall latency and decreases net transfer times.
Are the PCIe transaction managed entirely from the CPU?
No. See above.
Is there a manager of a bus that takes care of this?
No. The GPU manages the transfers. In this context there is no such thing as a bus master
EDIT: Seems like CUDA treats pinned and page-locked as the same as per the Pinned Host Memory section in this blog written by Mark Harris. This means by answer is moot and the best answer should be taken as is.
I bumped into this question while looking for something else. For all future users, I think #talonmies answers the question perfectly, but I'd like to bring to notice a slight difference between locking and pinning pages - the former ensures that the memory is not pageable but the kernel is free to move it around and the latter ensures that it stays in memory (i.e. non-pageable) but also is mapped to the same address.
Here's a reference to the same.
I want to prevent a thread switch by Windows XP/7 in a time critical part of my code that runs in a background thread. I'm pretty sure I can't create a situation where I can guarantee that won't happen, because of higher priority interrupts from system drivers, etc. However, I'd like to decrease the probability of a thread switch during that part of my code to the minimum that I can. Are there any create-thread flags or Window API calls that can assist me? General technique tips are appreciated too. If there is a way to get this done without having to raise the threads priority to real-time-critical that would be great, since I worry about creating system performance issues for the user if I do that.
UPDATE: I am adding this update after seeing the first responses to my original post. The concrete application that motivated the question has to do with real-time audio streaming. I want to eliminate every bit of delay I can. I found after coding up my original design that a thread switch can cause a 70ms or more delay at times. Since my app is between two sockets acting as a middleman for delivering audio, the instant I receive an audio buffer I want to immediately turn around and push it out the the destination socket. My original design used two cooperating threads and a semaphore since the there was one thread managing the source socket, and another thread for the destination socket. This architecture evolved from the fact the two devices behind the sockets are disparate entities.
I realized that if I combined the two sockets onto the same thread I could write a code block that reacted immediately to the socket-data-received message and turned it around to the destination socket in one shot. Now if I can do my best to avoid an intervening thread switch, that would be the optimal coding architecture for minimizing delay. To repeat, I know I can't guarantee this situation, but I am looking for tips/suggestions on how to write a block of code that does this and minimizes as best as I can the chance of an intervening thread switch.
Note, I am aware that O/S code behind the sockets introduces (potential) delays of its own.
AFAIK there are no such flags in CreateThread or etc (This also doesn't make sense IMHO). You may snooze other threads in your process from execution during in critical situations (by enumerating them and using SuspendThread), as well as you theoretically may enumerate & suspend threads in other processes.
OTOH snoozing threads is generally not a good idea, eventually you may call some 3rd-party code that would implicitly wait for something that should be accomplished in another threads, which you suspended.
IMHO - you should use what's suggested for the case - playing with thread/process priorities (also you may consider SetThreadPriorityBoost). Also the OS tends to raise the priority to threads that usually don't use CPU aggressively. That is, threads that work often but for short durations (before calling one of the waiting functions that suspend them until some condition) are considered to behave "nicely", and they get prioritized.
I need to write a screencast, and need to detect when window content has changed, even only text was selected. This window is third party control.
Ther're several methods.
(1) Screen polling.
You can poll the screen (that is, create a DIB, each time period to BitBlt from screen to it), and then send it as-is
Pros:
Very simple to implement
Cons:
High CPU load. Polling the entire screen number of times per second is very heavy (a lot of data should be transferred). Hence it'll be heavy and slow.
High network bandwidth
(2) Same as above, except now you do some analyzing of the polled screen to see the difference. Then you may send only the differences (and obviously don't send anything if no changes), plus you may optionally compress the differences stream.
Pros:
Still not too complex to implement
Significantly lower network bandwidth
Cons:
Even higher CPU usage.
(3) Same as above, except that you don't poll the screen constantly. Instead you do some hooking for your control (like spying for Windows messages that the control receives). Then you try learn when your control is supposed to redraw itself, and do the screen polling only in those scenarios.
Pros:
Significantly lower CPU usage
Still acceptable network bandwidth
Cons:
Implementation becomes complicated. Things like injecting hooks and etc.
Since this is based on some heuristic - you're not guaranteed (generally speaking) to cover all possible scenarios. In some circumstances you may miss the changes.
(4)
Hook at lower level: intercept calls to the drawing functions. Since there's enormous number of such functions in the user mode - the only realistic possibility of doing this is in the kernel mode.
You may write a virtual video driver (either "mirror" video driver, or hook the existing one) to receive all the drawing in the system. Then whenever you receive a drawing request on the specific area - you'll know it's changed.
Pros:
Lower CPU usage.
100% guarantee to intercept all drawings, without heuristics
Somewhat cleaner - no need to inject hooks into apps/controls
Cons:
It's a driver development! Unless you're experienced in it - it's a real nightmare.
More complex installation. Need administrator rights, most probably need restart.
Still considerable CPU load and bandwidth
(5)
Going on with driver development. As long as you know now which drawing functions are called - you may switch the strategy now. Instead of "remembering" dirty areas and polling the screen there - you may just "remember" the drawing function invoked with all the parameters, and then "repeat" it at the host side.
By such you don't have to poll the screen at all. You work in a "vectored" method (as opposed to "raster").
This however is much more complex to implement. Some drawing functions take as parameters another bitmaps, which in turn are drawn using another drawing functions and etc. You'll have to spy for bitmaps as well as screen.
Pros:
Zero CPU load
Best possible network traffic
Guaranteed to work always
Cons:
It's a driver development at its best! Months of development are guaranteed
Requires state-of-the-art programming, deep understanding of 2D drawing
Need to write the code at host which will "draw" all the "Recorded" commands.
We have a small embedded system without any video or serial ports (i.e. we can't output text via printf).
We would like to track the progress of our code through the initialization sequence.
Is there some simple things we can do to help with this.
It is not running any OS, and the hardware platform is somewhat customizable.
The simplest most scalable solution are state LEDs. Toggle LEDs based on actions, either in binary form or when certain actions occur if you can narrow your focus.
The most powerful will be a hardware JTAG device. You don't even need to set breakpoints - simply being able to stop the application and inspect the state of memory may be enough. Note that some hardware platforms do not support "fancy" options such as memory watches or hardware breakpoints. The former is usually worked around with constantly stopping the processor and reading memory (turns your 10MHz system into a 1kHz system), while the latter is sometimes performed using code replacement (replace the targeted instruction with a different jump), which sometimes masks other problems. Be aware of these issues and which embedded processors they apply to.
There are a few strategies you can employ to help with debugging:
If you have Output Pins available, you can hook them up to LEDs (or an oscilloscope) and toggle the output pins high/low to indicate that certain points have been reached in the code.
For example, 1 blink might be program loaded, 2 blink is foozbar initialized, 3 blink is accepting inputs...
If you have multiple output lines available, you can use a 7 segment LED to convey more information (numbers/letters instead of blinks).
If you have the capabilities to read memory and have some RAM available, you can use the sprint function to do printf-like debugging, but instead of going to a screen/serial port, it is written in memory.
It depends on the type of debugging that you're trying to do - in particular if you're after a temporary method of tracing or if you are trying to provide a tool that can be used as an indication of status during the life of the project (or product).
For one off, in depth source tracing and debugging an in-circuit debugger (eg. jtag) can be very helpful. However, they are most helpful where your debugging requires setting breakpoints and investigating memory and registers - which makes it of little benefit where you are dealing time critical problems.
Where you need to determine program state without having a significant impact on the execution of your system the use of LEDs connected to spare I/O pins will be helpful. These can also be used as the input to a digital storage oscilloscope (DSO) or logic analyzer.
This technique can be made more powerful by selecting unique patterns of pulses that will be identifiable on the DSO.
For a more versatile debugging tool, though, a serial port is a good solution. To save cost and PCB real-estate you may find it useful to use an plug-in module that contains the RS232 converters.
If you are trying to provide a longer term indication of status as part of the normal operation of your product, LEDs are again a cheap an simple method. However in this situation it is best to choose patterns of pulses that are slow enough to be easily identified by visual inspection. This will all you over time you will learn a particular pattern that represents "normal" behavior.
You can easily emulate serial communications (UARTs) using bit-banging from the IO pins of the system. Hook it to one of the card's pins and attach to a RS232 converter there (TTL to RS232 converters are easy to either buy or build), which goes to your PC's serial port.
A JTAG debugger is also an option, though cumbersome to set up.
If you don't have JTAG, the LEDs suggested by the others are a great idea - although you do tend to end up in a test/rebuild cycle to try to track down the issue.
If you've got more time, and spare hardware pins, and memory to spare, you could always bit-bash a low speed serial interface. I've found that pretty useful in the past.
Others have suggested some pretty good ideas using output pins, so I won't suggest that, although it can be a very good solution, and is very cost effective. If your budget and target processor support it, a hardware trace system, (either an old fashioned emulator, or a fancy BDM with bus snooping trace support) can be great for this type of thing. It's very expensive though.
The idea of using a bit-banged software UART is nice, but there's some effort required in writing one and also you need some free timers and interrupts. If your hardware has any other unused serial interface (SPI, I2C, ..), using them would be easier. With a small microcontroller you could convert the interface to RS-232.
If you have to go for the bit-banging, making a synchronous serial might be a simpler alternative as it wouldn't be critical to timing.
In a typical handheld/portable embedded system device Battery life is a major concern in design of H/W, S/W and the features the device can support. From the Software programming perspective, one is aware of MIPS, Memory(Data and Program) optimized code.
I am aware of the H/W Deep sleep mode, Standby mode that are used to clock the hardware at lower Cycles or turn of the clock entirel to some unused circutis to save power, but i am looking for some ideas from that point of view:
Wherein my code is running and it needs to keep executing, given this how can I write the code "power" efficiently so as to consume minimum watts?
Are there any special programming constructs, data structures, control structures which i should look at to achieve minimum power consumption for a given functionality.
Are there any s/w high level design considerations which one should keep in mind at time of code structure design, or during low level design to make the code as power efficient(Least power consuming) as possible?
Like 1800 INFORMATION said, avoid polling; subscribe to events and wait for them to happen
Update window content only when necessary - let the system decide when to redraw it
When updating window content, ensure your code recreates as little of the invalid region as possible
With quick code the CPU goes back to deep sleep mode faster and there's a better chance that such code stays in L1 cache
Operate on small data at one time so data stays in caches as well
Ensure that your application doesn't do any unnecessary action when in background
Make your software not only power efficient, but also power aware - update graphics less often when on battery, disable animations, less hard drive thrashing
And read some other guidelines. ;)
Recently a series of posts called "Optimizing Software Applications for Power", started appearing on Intel Software Blogs. May be of some use for x86 developers.
Zeroith, use a fully static machine that can stop when idle. You can't beat zero Hz.
First up, switch to a tickless operating system scheduler. Waking up every millisecend or so wastes power. If you can't, consider slowing the scheduler interrupt instead.
Secondly, ensure your idle thread is a power save, wait for next interrupt instruction.
You can do this in the sort of under-regulated "userland" most small devices have.
Thirdly, if you have to poll or perform user confidence activities like updating the UI,
sleep, do it, and get back to sleep.
Don't trust GUI frameworks that you haven't checked for "sleep and spin" kind of code.
Especially the event timer you may be tempted to use for #2.
Block a thread on read instead of polling with select()/epoll()/ WaitForMultipleObjects().
Puts stress on the thread scheuler ( and your brain) but the devices generally do okay.
This ends up changing your high-level design a bit; it gets tidier!.
A main loop that polls all the things you Might do ends up slow and wasteful on CPU, but does guarantee performance. ( Guaranteed to be slow)
Cache results, lazily create things. Users expect the device to be slow so don't disappoint them. Less running is better. Run as little as you can get away with.
Separate threads can be killed off when you stop needing them.
Try to get more memory than you need, then you can insert into more than one hashtable and save ever searching. This is a direct tradeoff if the memory is DRAM.
Look at a realtime-ier system than you think you might need. It saves time (sic) later.
They cope better with threading too.
Do not poll. Use events and other OS primitives to wait for notifiable occurrences. Polling ensures that the CPU will stay active and use more battery life.
From my work using smart phones, the best way I have found of preserving battery life is to ensure that everything you do not need for your program to function at that specific point is disabled.
For example, only switch Bluetooth on when you need it, similarly the phone capabilities, turn the screen brightness down when it isn't needed, turn the volume down, etc.
The power used by these functions will generally far outweigh the power used by your code.
To avoid polling is a good suggestion.
A microprocessor's power consumption is roughly proportional to its clock frequency, and to the square of its supply voltage. If you have the possibility to adjust these from software, that could save some power. Also, turning off the parts of the processor that you don't need (e.g. floating-point unit) may help, but this very much depends on your platform. In any case, you need a way to measure the actual power consumption of your processor, so that you can find out what works and what not. Just like speed optimizations, power optimizations need to be carefully profiled.
Consider using the network interfaces the least you can. You might want to gather information and send it out in bursts instead of constantly send it.
Look at what your compiler generates, particularly for hot areas of code.
If you have low priority intermittent operations, don't use specific timers to wake up to deal with them, but deal with when processing other events.
Use logic to avoid stupid scenarios where your app might go to sleep for 10 ms and then have to wake up again for the next event. For the kind of platform mentioned it shouldn't matter if both events are processed at the same time.
Having your own timer & callback mechanism might be appropriate for this kind of decision making. The trade off is in code complexity and maintenance vs. likely power savings.
Simply put, do as little as possible.
Well, to the extent that your code can execute entirely in the processor cache, you'll have less bus activity and save power. To the extent that your program is small enough to fit code+data entirely in the cache, you get that benefit "for free". OTOH, if your program is too big, and you can divide your programs into modules that are more or less independent of the other, you might get some power saving by dividing it into separate programs. (I suppose it's also possible to make a toolchain that spreas out related bundles of code and data into cache-sized chunks...)
I suppose that, theoretically, you can save some amount of unnecessary work by reducing the number of pointer dereferencing, and by refactoring your jumps so that the most likely jumps are taken first -- but that's not realistic to do as a programmer.
Transmeta had the idea of letting the machine do some instruction optimization on-the-fly to save power... But that didn't seem to help enough... And look where that got them.
Set unused memory or flash to 0xFF not 0x00. This is certainly true for flash and eeprom, not sure about s or d ram. For the proms there is an inversion so a 0 is stored as a 1 and takes more energy, a 1 is stored as a zero and takes less. This is why you read 0xFFs after erasing a block.
Rather timely this, article on Hackaday today about measuring power consumption of various commands:
Hackaday: the-effect-of-code-on-power-consumption
Aside from that:
- Interrupts are your friends
- Polling / wait() aren't your friends
- Do as little as possible
- make your code as small/efficient as possible
- Turn off as many modules, pins, peripherals as possible in the micro
- Run as slowly as possible
- If the micro has settings for pin drive strengh, slew rate, etc. check them & configure them, the defaults are often full power / max speed.
- returning to the article above, go back and measure the power & see if you can drop it by altering things.
also something that is not trivial to do is reduce precision of the mathematical operations, go for the smallest dataset available and if available by your development environment pack data and aggregate operations.
knuth books could give you all the variant of specific algorithms you need to save memory or cpu, or going with reduced precision minimizing the rounding errors
also, spent some time checking for all the embedded device api - for example most symbian phones could do audio encoding via a specialized hardware
Do your work as quickly as possible, and then go to some idle state waiting for interrupts (or events) to happen. Try to make the code run out of cache with as little external memory traffic as possible.
On Linux, install powertop to see how often which piece of software wakes up the CPU. And follow the various tips that the powertop site links to, some of which are probably applicable to non-Linux, too.
http://www.lesswatts.org/projects/powertop/
Choose efficient algorithms that are quick and have small basic blocks and minimal memory accesses.
Understand the cache size and functional units of your processor.
Don't access memory. Don't use objects or garbage collection or any other high level constructs if they expands your working code or data set outside the available cache. If you know the cache size and associativity, lay out the entire working data set you will need in low power mode and fit it all into the dcache (forget some of the "proper" coding practices that scatter the data around in separate objects or data structures if that causes cache trashing). Same with all the subroutines. Put your working code set all in one module if necessary to stripe it all in the icache. If the processor has multiple levels of cache, try to fit in the lowest level of instruction or data cache possible. Don't use floating point unit or any other instructions that may power up any other optional functional units unless you can make a good case that use of these instructions significantly shortens the time that the CPU is out of sleep mode.
etc.
Don't poll, sleep
Avoid using power hungry areas of the chip when possible. For example multipliers are power hungry, if you can shift and add you can save some Joules (as long as you don't do so much shifting and adding that actually the multiplier is a win!)
If you are really serious,l get a power-aware debugger, which can correlate power usage with your source code. Like this