Search Results

Search found 955 results on 39 pages for 'gpu'.

Page 24/39 | < Previous Page | 20 21 22 23 24 25 26 27 28 29 30 31  | Next Page >

  • Why does OpenGL's glDrawArrays() fail with GL_INVALID_OPERATION under Core Profile 3.2, but not 3.3 or 4.2?

    - by metaleap
    I have OpenGL rendering code calling glDrawArrays that works flawlessly when the OpenGL context is (automatically / implicitly obtained) 4.2 but fails consistently (GL_INVALID_OPERATION) with an explicitly requested OpenGL core context 3.2. (Shaders are always set to #version 150 in both cases but that's beside the point here I suspect.) According to specs, there are only two instances when glDrawArrays() fails with GL_INVALID_OPERATION: "if a non-zero buffer object name is bound to an enabled array and the buffer object's data store is currently mapped" -- I'm not doing any buffer mapping at this point "if a geometry shader is active and mode? is incompatible with [...]" -- nope, no geometry shaders as of now. Furthermore: I have verified & double-checked that it's only the glDrawArrays() calls failing. Also double-checked that all arguments passed to glDrawArrays() are identical under both GL versions, buffer bindings too. This happens across 3 different nvidia GPUs and 2 different OSes (Win7 and OSX, both 64-bit -- of course, in OSX we have only the 3.2 context, no 4.2 anyway). It does not happen with an integrated "Intel HD" GPU but for that one, I only get an automatic implicit 3.3 context (trying to explicitly force a 3.2 core profile with this GPU via GLFW here fails the window creation but that's an entirely different issue...) For what it's worth, here's the relevant routine excerpted from the render loop, in Golang: func (me *TMesh) render () { curMesh = me curTechnique.OnRenderMesh() gl.BindBuffer(gl.ARRAY_BUFFER, me.glVertBuf) if me.glElemBuf > 0 { gl.BindBuffer(gl.ELEMENT_ARRAY_BUFFER, me.glElemBuf) gl.VertexAttribPointer(curProg.AttrLocs["aPos"], 3, gl.FLOAT, gl.FALSE, 0, gl.Pointer(nil)) gl.DrawElements(me.glMode, me.glNumIndices, gl.UNSIGNED_INT, gl.Pointer(nil)) gl.BindBuffer(gl.ELEMENT_ARRAY_BUFFER, 0) } else { gl.VertexAttribPointer(curProg.AttrLocs["aPos"], 3, gl.FLOAT, gl.FALSE, 0, gl.Pointer(nil)) /* BOOM! */ gl.DrawArrays(me.glMode, 0, me.glNumVerts) } gl.BindBuffer(gl.ARRAY_BUFFER, 0) } So of course this is part of a bigger render-loop, though the whole "*TMesh" construction for now is just two instances, one a simple cube and the other a simple pyramid. What matters is that the entire drawing loop works flawlessly with no errors reported when GL is queried for errors under both 3.3 and 4.2, yet on 3 nvidia GPUs with an explicit 3.2 core profile fails with an error code that according to spec is only invoked in two specific situations, none of which as far as I can tell apply here. What could be wrong here? Have you ever run into this? Any ideas what I have been missing?

    Read the article

  • Is there a simple way to let a layer throw an smooth shadow?

    - by mystify
    I was drawing a path into a layer. Lets say I can't access that drawing code in any way, because it comes from a compiled lib. Now I want to let that layer throw a shadow which matches the shape of its irregular content shape. Is there an easy way to do it? Or must I draw like 20 of those layers and scale them up on every iteration, adjusting their alpha and letting the GPU do the extraordinarily heavy compositing?

    Read the article

  • How to diagnose computer lockup/freezing problem

    - by Scott Mitchell
    I built a desktop computer a couple years back with the following specs: CPU: Intel Core 2 Quad Q9300 Yorkfield 2.5GHz 6MB L2 Cache LGA 775 95W Quad-Core Processor BX80580Q9300 Motherboard: EVGA 122-CK-NF68-T1 LGA 775 NVIDIA nForce 680i SLI ATX Intel Motherboard Video Card: Two EVGA 256-P2-N758-TR GeForce 8600GT SCC 256MB 128-bit GDDR3 PCI Express x16 SLI Supported Video Card PSU: SeaSonic S12 Energy Plus SS-550HT 550W ATX12V V2.3 / EPS12V V2.91 SLI Certified CrossFire Ready 80 PLUS Certified Active PFC Power Supply Memory: Two G.SKILL 4GB (2 x 2GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit Desktop Memory Model F2-6400CL5D-4GBPQ Since its inception, the machine has periodically locked up, the regularlity having varied over the years from once a day to once a month. Typically, lockups happen once every few days. By "lockup" I mean my computer just freezes. The screen locks up, I can't move the mouse. Hitting keys on my keyboard that normally turn LEDs on or off on the keyboard (such as Caps Lock) no longer turn the LEDs on or off. If there was music playing at the time of the lockup, noise keeps coming out of the speakers, but it's just the current frequency/note that plays indefinitely. There is no BSOD. When such a lockup occurs I have to do a hard reboot by either turning off the computer or hitting the reset button. I have the most recent version of the NVIDIA hardware drivers, and update them semi-regularly, but that hasn't seemed to help. I am currently using Windows 7 x64, but was previously using Windows Server 2003 x64 and having the same lockup issues. My guess is that it's somehow video driver or motherboard related, but I don't know how to go about diagnosing this problem to narrow down which of the two is the culprit. Additional information re: cooling Regarding cooling... I've not installed any after-market cooling systems aside from two regular fans I scavenged from an older computer. The fan atop the CPU is the one that shipped with it. One of the two scavenged fans I added it located at the bottom tower of the corner, in an attempt to create some airflow from front to back. The second fan is pointed directly at the two video cards. SpeedFan installation and readings Per studiohack's suggestion, I installed SpeedFan, which provided the following temperature readings: GPU: 63C GPU: 65C System: 76C CPU: 64C AUX: 36C Core 0: 78C Core 1: 76C Core 2: 79C Core 3: 79C Update #3: Another Lockup :-( Well, I had another lockup last night. :-( SpeedFan reported the CPU temp at 38 C when it happened, and there was no spike in temperature leading up to the freeze. One thing I notice is that the freeze seems more likely to happen if I am watching a video. In fact, of the last 5 freezes over the past month, 4 of them have been while watching a video on Flickr. Not necessarily the same video, but a video nevertheless. I don't know if this is just coincidence or if it means anything. (As an aside, each night before bedtime my 2 year old daughter sits on my lap and watches some home videos on Flickr and, in the last month, has learned the phrase, "Uh oh, computer broke.") Update #4: MemTest86 and 3DMark06 Test Results: Per suggestions in the comments, I ran the MemTest86 overnight and it cycled through the 8 GB of memory 5 times without error. I also ran the 3DMark06 test without a problem (see my scores at http://3dmark.com/3dm06/15163549). So... what now? :-) Any further suggestions on what to check? Is there some way to get a stack trace or something when the computer locks like that? Thanks

    Read the article

  • What would make a noise in a PC on graphics operations on a passively-cooled system?

    - by T.J. Crowder
    I have this system based on the Intel D510MO motherboard, which is basically an Atom D510 (dual-core HT Atom w/built-in GPU), an Intel NM10 chipset, and a Realtek Gigabit LAN controller. It's entirely passively cooled. I noticed almost immediately that there was a kind of very, very soft noise that corresponded with graphics operations, sort of the noise you'd get if you had a sheet of flat paper and slid something really light across it — but more electronic than that. I wrote it off as observation error and/or disk activity triggered by the graphics operation (although the latter seemed like a lot of unnecessary disk activity). It isn't. I got curious enough that I finally did a few controlled experiments, and here's what I've determined: It isn't the HDD. For one thing, the sounds the HDD makes (when seeking, when reading or writing, when just sitting there spinning) is different. For another, I used sudo hdparm -y /dev/sda (I'm using Ubuntu 10.04 LTS) to temporarily put the disk on standby while making sure that non-disk graphics op was happening in a loop. The disk spun down, but the other sound continued, corresponding perfectly with the timing of the graphics op. (Then the disk spun up again, but it takes long enough that I could rule out the HDD.) It isn't the monitor; I ensured the two were well physically-separated and the sound was definitely coming from the main box. It isn't something else in the room; the sound is coming from the box. It isn't cross-talk to an audio circuit coming out the speakers. (It doesn't have any speakers.) It isn't my mouse (e.g., when I'm trying to make graphics ops happen); the sound happens if I set up a recurring operation and don't use the mouse at all, or if I lift the mouse off the table slightly (but enough that the laser still registers movement). It isn't the voices in my head; they never whisper like that. Other observations: It doesn't seem to matter what the graphics operation is; anything that changes what's on the screen seems to do it. I get the sound when moving the mouse over the Chromium tab bar (which makes the tab backgrounds change); I get it when a web page has a counter on it that changes the text on the page: I get it when dragging window contents around. The sound is very, very slightly louder if the graphics op is larger, like scrolling a text area when writing a question on superuser.com, than for smaller operations like the tick counter on the web page. But it's very slight. It's fairly loud (and of good duration) when the op involves color changes to substantial surface areas. For instance, when asking a question here on superuser and you move the cursor between the question box and the tag box, and the help to the right fades out, changes, and fades back in. (Yet another example related to the web browser, so let me say: I hear it when operations completely unrelated to the web browser as well.) It doesn't sound like arcing or anything like that (I'd've shut off the machine Right Quick Like if it did). Moving windows does it. Scrolling windows (by and large) doesn't. I have the feeling I've heard this sort of thing before, when all system fans were on low and such, with other systems — but (again) written it off as observational error. For all the world it's like I'm hearing the CPU working (as opposed to the GPU; note the window scroll thing above) or data being transferred somewhere, but that just seems...unlikely. So what am I hearing? This may seem like a very localized question, but perhaps other silent PC enthusiasts may be interested as well...

    Read the article

  • How to diagnose computer lockups and freezes?

    - by Scott Mitchell
    I built a desktop computer a couple years back with the following specs: CPU: Intel Core 2 Quad Q9300 Yorkfield 2.5GHz 6 MB L2 Cache LGA 775 95W Quad-Core Processor BX80580Q9300 Motherboard: EVGA 122-CK-NF68-T1 LGA 775 NVIDIA nForce 680i SLI ATX Intel Motherboard Video Card: Two EVGA 256-P2-N758-TR GeForce 8600GT SCC 256 MB 128-bit GDDR3 PCI Express x16 SLI Supported Video Card PSU: SeaSonic S12 Energy Plus SS-550HT 550W ATX12V V2.3 / EPS12V V2.91 SLI Certified CrossFire Ready 80 PLUS Certified Active PFC Power Supply Memory: Two G.SKILL 4 GB (2 x 2 GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit Desktop Memory Model F2-6400CL5D-4GBPQ Since its inception, the machine has periodically locked up, the regularity having varied over the years from once a day to once a month. Typically, lockups happen once every few days. By "lockup" I mean my computer just freezes. The screen locks up, I can't move the mouse. Hitting keys on my keyboard that normally turn LEDs on or off on the keyboard (such as Caps Lock) no longer turn the LEDs on or off. If there was music playing at the time of the lockup, noise keeps coming out of the speakers, but it's just the current frequency/note that plays indefinitely. There is no BSOD. When such a lockup occurs I have to do a hard reboot by either turning off the computer or hitting the reset button. I have the most recent version of the NVIDIA hardware drivers, and update them semi-regularly, but that hasn't seemed to help. I am currently using Windows 7 x64, but was previously using Windows Server 2003 x64 and having the same lockup issues. My guess is that it's somehow video driver or motherboard related, but I don't know how to go about diagnosing this problem to narrow down which of the two is the culprit. Additional information re: cooling Regarding cooling... I've not installed any after-market cooling systems aside from two regular fans I scavenged from an older computer. The fan atop the CPU is the one that shipped with it. One of the two scavenged fans I added it located at the bottom tower of the corner, in an attempt to create some airflow from front to back. The second fan is pointed directly at the two video cards. SpeedFan installation and readings Per studiohack's suggestion, I installed SpeedFan, which provided the following temperature readings: GPU: 63C GPU: 65C System: 76C CPU: 64C AUX: 36C Core 0: 78C Core 1: 76C Core 2: 79C Core 3: 79C Update #3: Another Lockup :-( Well, I had another lockup last night. :-( SpeedFan reported the CPU temp at 38 C when it happened, and there was no spike in temperature leading up to the freeze. One thing I notice is that the freeze seems more likely to happen if I am watching a video. In fact, of the last 5 freezes over the past month, 4 of them have been while watching a video on Flickr. Not necessarily the same video, but a video nevertheless. I don't know if this is just coincidence or if it means anything. (As an aside, each night before bedtime my 2 year old daughter sits on my lap and watches some home videos on Flickr and, in the last month, has learned the phrase, "Uh oh, computer broke.") Update #4: MemTest86 and 3DMark06 Test Results: Per suggestions in the comments, I ran the MemTest86 overnight and it cycled through the 8 GB of memory 5 times without error. I also ran the 3DMark06 test without a problem (see my scores at http://3dmark.com/3dm06/15163549). So... what now? :-) Any further suggestions on what to check? Is there some way to get a stack trace or something when the computer locks like that? Resolution I have never did figure out the particular problems, but based on the suggestions here and elsewhere, I'm presuming it was a motherboard issue. In any event, I recently upgraded my system, buying a new motherbeard, PSU, CPU, and RAM, and that new rig has been working splendidly the past several weeks. I am using the same graphic cards as in the old setup, so I think it's safe to reason that they weren't the cause of the problem.

    Read the article

  • BSOD Code 16, artifacts all over the place (gtx 260)

    - by belinea
    I have a following, quite dated rig E8400 Core 2 Duo cpu Intel Dragontail Peak DP35DP motherboard on Intel Bearlake P35 chipset 4GB ram Geforce 260gtx Corsair 650W PSU Windows 7 64 bit The following things has happened in the last few days. I first decided to update my Nvidia drivers to the latest version. That was 4 days ago. PC worked fine for 2 days and I was able to play few games as well without any problems. Then 2 days ago a first crash happened while playing the new XCOM game. BSOD code 16. Just the blue screen, no artifacts. PC rebooted and worked well again, I continued playing this game for another 2 hours and went to sleep. Next evening I tried to play some BF3 multiplayer (use to play on LOW settings). Approx. 10 minutes into the game red/pink-ish artifacts appeared on the screen and game quit to desktop. Restarted the game and another 3-4 minutes afterwards another crash to desktop but this time followed shortly by BSOD Code 16. From that moment I started to seeing artifacts on random startups, including Windows loading screen and the BIOS itself. Windows would still load but soon enough it would BSOD on a simple task like opening a Internet browser. Today I get tons of artifacts (little small red dashes all over the screen) on BIOS, loading screen, normal Windows mode as well as safe mode. I suspect it wouldn't be drivers but I tried removing and sweeping them entirely in the safe mode. PC would still start with artifacts all over the place but would load the normal mode, just without the driver, in the default lowest resolution. As soon as proper Nvidia drivers are installed though and PC rebooted, Windows doesn't load at all as BSOD now appears on loading screen. However, again, if I go to safe mode and remove drivers, normal mode launches fine. So obviously crash happens only on high resolution. I opened my machine this morning and gave it a proper cleaning even though it wasn't heavily dusted. It didn't help and number of artifacts seems to increase with every PC restart. I write these words in safe mode, which works, but I have to look through all the red dots and dashes. I don't have built in GPU chipset so I can't try removing my Geforce card nor can I borrow A GPU from anyone else. What are my options? I was looking into getting a completely new rig around Christmas so I'm not freaking out about this. If everything points to hardware issue I may simply decide to get the new machine earlier and don't bother with fixing this one. However it would be great to learn more if this is indeed situation that has slim chances of getting sorted. I realize BSOD Code 16 is rather popular topic online but every story seems a bit different and there can be number of issues with it. Hence a new thread.

    Read the article

  • Thursday Community Keynote: "By the Community, For the Community"

    - by Janice J. Heiss
    Sharat Chander, JavaOne Community Chairperson, began Thursday's Community Keynote. As part of the morning’s theme of "By the Community, For the Community," Chander noted that 60% of the material at the 2012 JavaOne conference was presented by Java Community members. "So next year, when the call for papers starts, put-in your submissions," he urged.From there, Gary Frost, Principal Member of Technical Staff, AMD, expanded upon Sunday's Strategy Keynote exploration of Project Sumatra, an OpenJDK project targeted at bringing Java to heterogeneous computing platforms (which combine the CPU and the parallel processor of the GPU into a single piece of silicon). Sumatra entails enhancing the JVM to make maximum use of these advanced platforms. Within this development space, AMD created the Aparapi API, which converts Java bytecode into OpenCL for execution on such GPU devices. The Aparapi API was open sourced in September 2011.Whether it was zooming-in on a Mandelbrot set, "the game of life," or a swarm of 10,000 Dukes in a space-bound gravitational dance, Frost's demos, using an Aparapi/OpenCL implementation, produced stunningly faster display results. He indicated that the Java 9 timeframe is where they see Project Sumatra coming to ultimate fruition, employing the Lamdas of Java 8.Returning to the theme of the keynote, Donald Smith, Director, Java Product Management, Oracle, explored a mind map graphic demonstrating the importance of Community in terms of fostering innovation. "It's the sharing and mixing of culture, the diversity, and the rapid prototyping," he said. Within this topic, Smith, brought up a panel of representatives from Cloudera, Eclipse, Eucalyptus, Perrone Robotics, and Twitter--ideal manifestations of community and innovation in the world of Java.Marten Mickos, CEO, Eucalyptus Systems, explored his company's open source cloud software platform, written in Java, and used by gaming companies, technology companies, media companies, and more. Chris Aniszczyk, Operations Engineering,Twitter, noted the importance of the JVM in terms of their multiple-language development environment. Mike Olson, CEO, Cloudera, described his company's Apache Hadoop-based software, support, and training. Mike Milinkovich, Executive Director, Eclipse Foundation, noted that they have about 270 tools projects at Eclipse, with 267 of them written in Java. Milinkovich added that Eclipse will even be going into space in 2013, as part of the control software on various experiments aboard the International Space Station. Lastly, Paul Perrone, CEO, Perrone Robotics, detailed his company's robotics and automation software platform built 100% on Java, including Java SE and Java ME--"on rat, to cat, to elephant-sized systems." Milinkovic noted that communities are by nature so good at innovation because of their very openness--"The more open you make your innovation process, the more ideas are challenged, and the more developers are focused on justifying their choices all the way through the process."From there, Georges Saab, VP Development Java SE OpenJDK, continued the topic of innovation and helping the Java Community to "Make the Future Java." Martijn Verburg, representing the London Java Community (winner of a Duke's Choice Award 2012 for their activity in OpenJDK and JCP), soon joined Saab onstage. Verburg detailed the LJC's "Adopt a JSR" program--"to get day-to-day developers more involved in the innovation that's happening around them."  From its London launching pad, the innovative program has spread to Brazil, Morocco, Latvia, India, and more.Other active participants in the program joined Verburg onstage--Ben Evans, London Java Community; James Gough, Stackthread; Bruno Souza, SOUJava; Richard Warburton, jClarity; and Cecelia Borg, Oracle--OpenJDK Onboarding. Together, the group explored the goals and tasks inherent in the Adopt a JSR program--from organizing hack days (testing prototype implementations), to managing mailing lists and forums, to triaging issues, to evangelism—all with the goal of fostering greater community/developer involvement, but equally importantly, building better open standards. “Come join us, and make your ecosystem better!" urged Verburg.Paul Perrone returned to profile the latest in his company's robotics work around Java--including the AARDBOTS family of smaller robotic vehicles, running the Perrone MAX platform on top of the Java JVM. Perrone took his "Rumbles" four-wheeled robot out for a spin onstage--a roaming, ARM-based security-bot vehicle, complete with IR, ultrasonic, and "cliff" sensors (the latter, for the raised stage at JavaOne). As an ultimate window into the future of robotics, Perrone displayed a "head-set" controller--a sensor directed at the forehead to monitor brainwaves, for the someday-implementation of brain-to-robot control.Then, just when it seemed this might be the end of the day's futuristic offerings, a mystery voice from offstage pronounced "I've got some toys"--proving to be guest-visitor James Gosling, there to explore his cutting-edge work with Liquid Robotics. While most think of robots as something with wheels or arms or lasers, Gosling explained, the Liquid Robotics vehicle is an entirely new and innovative ocean-going 'bot. Looking like a floating surfboard, with an attached set of underwater wings, the autonomous devices roam the oceans using only the energy of ocean waves to propel them, and a single actuated rudder to steer. "We have to accomplish all guidance just by wiggling the rudder," Gosling said. The devices offer applications from self-installing weather buoy, to pollution monitoring station, to marine mammal monitoring device, to climate change data gathering, to even ocean life genomic sampling. The early versions of the vehicle used C code on very tiny industrial micro controllers, where they had to "count the bytes one at a time."  But the latest generation vehicles, which just hit the water a week or so ago, employ an ARM processor running Linux and the ARM version of JDK 7. Gosling explained that vehicle communication from remote locations is achieved via the Iridium satellite network. But because of the costs of this communication path, the data must be sent in very small bursts--using SBD short burst data. "It costs $1/kb, so that rules everything in the software design,” said Gosling. “If you were trying to stream a Netflix video over this, it would cost a million dollars a movie. …We don't have a 'big data' problem," he quipped. There are currently about 150 Liquid Robotics vehicles out traversing the oceans. Gosling demonstrated real time satellite tracking of several vehicles currently at sea, noting that Java is actually particularly good at AI applications--due to the language having garbage collection, which facilitates complex data structures. To close-out his time onstage, Gosling of course participated in the ceremonial Java tee-shirt toss out to the audience…In parting, Chander passed the JavaOne Community Chairperson baton to Stephen Chin, Java Technology Evangelist, Oracle. Onstage in full motorcycle gear, Chin noted that he'll soon be touring Europe by motorcycle, meeting Java Community Members and streaming live via UStream--the ultimate manifestation of community and technology!  He also reminded attendees of the upcoming JavaOne Latin America 2012, São Paulo, Brazil (December 4-6, 2012), and stated that the CFP (call for papers) at the conference has been extended for one more week. "Remember, December is summer in Brazil!" Chin said.

    Read the article

  • tile_static, tile_barrier, and tiled matrix multiplication with C++ AMP

    - by Daniel Moth
    We ended the previous post with a mechanical transformation of the C++ AMP matrix multiplication example to the tiled model and in the process introduced tiled_index and tiled_grid. This is part 2. tile_static memory You all know that in regular CPU code, static variables have the same value regardless of which thread accesses the static variable. This is in contrast with non-static local variables, where each thread has its own copy. Back to C++ AMP, the same rules apply and each thread has its own value for local variables in your lambda, whereas all threads see the same global memory, which is the data they have access to via the array and array_view. In addition, on an accelerator like the GPU, there is a programmable cache, a third kind of memory type if you'd like to think of it that way (some call it shared memory, others call it scratchpad memory). Variables stored in that memory share the same value for every thread in the same tile. So, when you use the tiled model, you can have variables where each thread in the same tile sees the same value for that variable, that threads from other tiles do not. The new storage class for local variables introduced for this purpose is called tile_static. You can only use tile_static in restrict(direct3d) functions, and only when explicitly using the tiled model. What this looks like in code should be no surprise, but here is a snippet to confirm your mental image, using a good old regular C array // each tile of threads has its own copy of locA, // shared among the threads of the tile tile_static float locA[16][16]; Note that tile_static variables are scoped and have the lifetime of the tile, and they cannot have constructors or destructors. tile_barrier In amp.h one of the types introduced is tile_barrier. You cannot construct this object yourself (although if you had one, you could use a copy constructor to create another one). So how do you get one of these? You get it, from a tiled_index object. Beyond the 4 properties returning index objects, tiled_index has another property, barrier, that returns a tile_barrier object. The tile_barrier class exposes a single member, the method wait. 15: // Given a tiled_index object named t_idx 16: t_idx.barrier.wait(); 17: // more code …in the code above, all threads in the tile will reach line 16 before a single one progresses to line 17. Note that all threads must be able to reach the barrier, i.e. if you had branchy code in such a way which meant that there is a chance that not all threads could reach line 16, then the code above would be illegal. Tiled Matrix Multiplication Example – part 2 So now that we added to our understanding the concepts of tile_static and tile_barrier, let me obfuscate rewrite the matrix multiplication code so that it takes advantage of tiling. Before you start reading this, I suggest you get a cup of your favorite non-alcoholic beverage to enjoy while you try to fully understand the code. 01: void MatrixMultiplyTiled(vector<float>& vC, const vector<float>& vA, const vector<float>& vB, int M, int N, int W) 02: { 03: static const int TS = 16; 04: array_view<const float,2> a(M, W, vA); 05: array_view<const float,2> b(W, N, vB); 06: array_view<writeonly<float>,2> c(M,N,vC); 07: parallel_for_each(c.grid.tile< TS, TS >(), 08: [=] (tiled_index< TS, TS> t_idx) restrict(direct3d) 09: { 10: int row = t_idx.local[0]; int col = t_idx.local[1]; 11: float sum = 0.0f; 12: for (int i = 0; i < W; i += TS) { 13: tile_static float locA[TS][TS], locB[TS][TS]; 14: locA[row][col] = a(t_idx.global[0], col + i); 15: locB[row][col] = b(row + i, t_idx.global[1]); 16: t_idx.barrier.wait(); 17: for (int k = 0; k < TS; k++) 18: sum += locA[row][k] * locB[k][col]; 19: t_idx.barrier.wait(); 20: } 21: c[t_idx.global] = sum; 22: }); 23: } Notice that all the code up to line 9 is the same as per the changes we made in part 1 of tiling introduction. If you squint, the body of the lambda itself preserves the original algorithm on lines 10, 11, and 17, 18, and 21. The difference being that those lines use new indexing and the tile_static arrays; the tile_static arrays are declared and initialized on the brand new lines 13-15. On those lines we copy from the global memory represented by the array_view objects (a and b), to the tile_static vanilla arrays (locA and locB) – we are copying enough to fit a tile. Because in the code that follows on line 18 we expect the data for this tile to be in the tile_static storage, we need to synchronize the threads within each tile with a barrier, which we do on line 16 (to avoid accessing uninitialized memory on line 18). We also need to synchronize the threads within a tile on line 19, again to avoid the race between lines 14, 15 (retrieving the next set of data for each tile and overwriting the previous set) and line 18 (not being done processing the previous set of data). Luckily, as part of the awesome C++ AMP debugger in Visual Studio there is an option that helps you find such races, but that is a story for another blog post another time. May I suggest reading the next section, and then coming back to re-read and walk through this code with pen and paper to really grok what is going on, if you haven't already? Cool. Why would I introduce this tiling complexity into my code? Funny you should ask that, I was just about to tell you. There is only one reason we tiled our extent, had to deal with finding a good tile size, ensure the number of threads we schedule are correctly divisible with the tile size, had to use a tiled_index instead of a normal index, and had to understand tile_barrier and to figure out where we need to use it, and double the size of our lambda in terms of lines of code: the reason is to be able to use tile_static memory. Why do we want to use tile_static memory? Because accessing tile_static memory is around 10 times faster than accessing the global memory on an accelerator like the GPU, e.g. in the code above, if you can get 150GB/second accessing data from the array_view a, you can get 1500GB/second accessing the tile_static array locA. And since by definition you are dealing with really large data sets, the savings really pay off. We have seen tiled implementations being twice as fast as their non-tiled counterparts. Now, some algorithms will not have performance benefits from tiling (and in fact may deteriorate), e.g. algorithms that require you to go only once to global memory will not benefit from tiling, since with tiling you already have to fetch the data once from global memory! Other algorithms may benefit, but you may decide that you are happy with your code being 150 times faster than the serial-version you had, and you do not need to invest to make it 250 times faster. Also algorithms with more than 3 dimensions, which C++ AMP supports in the non-tiled model, cannot be tiled. Also note that in future releases, we may invest in making the non-tiled model, which already uses tiling under the covers, go the extra step and use tile_static memory on your behalf, but it is obviously way to early to commit to anything like that, and we certainly don't do any of that today. Comments about this post by Daniel Moth welcome at the original blog.

    Read the article

  • updating from windows 8 to windows 8.1 give me error 0x101 0x2000c and driver irql not less or equal on clean install

    - by user1143745
    i'm trying to update my hp dv6-6156sl running win 8 to win 8.1. When i try to update the os with all kept when it configures devices (around 55%) i get a bsod with driver irql not less or equal error and windows revert back to my previous windows 8 showing at the end 0x101 0x2000c error. So i tried a clean install to see if i could fix it but i still get the same driver irql not less or equal error (on ntoskrnl.exe) I tried to edit bios settings to see if there's something to disable but like almost all hp laptops it is locked so i can only enable virtualization, select boot order and change gpu between fixed or dinamic (none of these worked) i found the dmp file with the error of the update: https://dl.dropboxusercontent.com/u/2219086/setupmem.dmp How can i fix this?

    Read the article

  • Remove automatic Aero disabling in Windows 7

    - by Jani Hartikainen
    Sometimes when I'm playing games which are heavy on the GPU, Windows decides to helpfully disable aero, causing everything to freeze for a bit and in the worst case, combined with ATI's brilliant drivers, causes the game to crash. So, How do I stop Windows from automatically disabling Aero when playing games? It has absolutely no effect on the performance of the game itself when it does that. Also, I'd like to get rid of the "You should disable Aero to improve performance" helpful hint popup which sometimes shows up. But I suppose getting rid of the first will get rid of the second, assuming anyone knows how.

    Read the article

  • Dual monitors through 1 HDMI port

    - by Carlos
    I currently have a Dell Studio XPS 13 laptop connected to a 24" HP monitor (w2448hc). Im thinking on getting a second one, however am wondering what i need for the setup (hardware wise). Also I was wondering it there is any down side to it, or something i should be aware of. For example, image quality loss, GPU overloading, or anything important I should know. More than anything Im interested in your advice. Also the monitors do have built in speakers (HDMI sound output), is the sound going to be reproduced by only one monitor or both? Specs Model: Dell Studio XPS 13 OS: Genuine Windows® 7 Home Premium 64-Bit CPU: Intel® CoreTM 2 Duo P8600 (2.4GHz, 3MB L2 Cache, 1067MHz FSB) Chipset: NVIDIA® GeForce® MCP79MX RAM: 4GB 1067MHz DDR3 SDRAM Graphics: SLi NVIDIA® GeForce® 9500M - 256MB Thanks for your advice, if there is anything additional i need to buy an you have a personal preference pass the brand name so i can check it out. Thanks!

    Read the article

  • Dual monitors though 1 HDMI port

    - by Carlos
    I currently have a Dell Studio XPS 13 laptop connected to a 24" HP monitor (w2448hc). Im thinking on getting a second one, however am wondering what i need for the setup (hardware wise). Also I was wondering it there is any down side to it, or something i should be aware of. For example, image quality loss, GPU overloading, or anything important I should know. More than anything Im interested in your advice. Also the monitors do have built in speakers (HDMI sound output), is the sound going to be reproduced by only one monitor or both? Specs Model: Dell Studio XPS 13 OS: Genuine Windows® 7 Home Premium 64-Bit CPU: Intel® CoreTM 2 Duo P8600 (2.4GHz, 3MB L2 Cache, 1067MHz FSB) Chipset: NVIDIA® GeForce® MCP79MX RAM: 4GB 1067MHz DDR3 SDRAM Graphics: SLi NVIDIA® GeForce® 9500M - 256MB Thanks for your advice, if there is anything additional i need to buy an you have a personal preference pass the brand name so i can check it out. Thanks!

    Read the article

  • HP Pavilion dv6000 graphics issue

    - by Tony
    A few days ago my laptop seized up on me while I was playing a game. When I restarted the computer, it sent me back to 640x480 resolution with 16 colors (clearly a graphical issue). In device manager, it shows an error, stating "Windows has stopped this device because it has reported problems. (Code 43)" regarding the video card. Upon removing and re-installing the drivers using both Windows Update and my own searching through nvidia's site, it's given me zero results. For kicks, I ran my antivirus (you never know!) and registry cleaner to see if anything would turn up. The antivirus found nothing, but CCleaner found 27 registry errors all dealing with my video card. What could be the cause of this, and what course of action would you recommend in this situation? Thanks in advance for the help! OS: Windows 7 Professional GPU: nvidia GeForce 8400M GS

    Read the article

  • Macbook Pro suddenly lagging video playback + Flash sites

    - by Mathias
    I have a Macbook Pro, OSx Lion, Intel Core2 Duo, 4GB Ram, NVidia Geforce 8600M GT 128 MB Ram, Intel x25m SSD. Approximately 4 years old. I've been running Flash sites and playing videos without any problems for years. Then suddenly 3 months ago, a flash site like http://thefwa.com is lagging in all browsers. Even mouseover animations - anything. Also video playback in e.g VLC and Quicktime is now lagging. Same videos I used before, I tried installing an older version of VLC without any luck. Playing back video in VLC utilizes the CPU almost 100%, and Flash sites like thefwa.com easily takes up 50-60%. It's as if the hardware acceleration stopped working, or the GPU lost its magic. UPDATE: Same issues also occurred on Snow Leopard Has anyone experienced something similar, or do you know what might be wrong?

    Read the article

  • How to connect 3 monitors into an Asus 9800GT

    - by aliasbody
    I don't know if this is possible but, I've heard that there exist some DVI connections that allow and recognize the connection os 2 Monitores in the same port (I can be wrong tho). I have an Asus 9800GT 512MB on my computer that runs Ubuntu 12.04 and Arch Linux, and I am trying to connect 3 monitors. I already have 2 (17" each), and want to connect a new one with 19", but on my 9800GT I only have 2 DVI connections. Is there anyway to have 3 monitores extended without having to buy a new GPU ? Thanks in Advance,

    Read the article

  • What is the best position for power unit?

    - by guest86
    I would like to buy new computer case. Last time I bought a computer was in 2008 and many things have changed up to day. Many new computer cases have power unit placed down, on bottom. I'm thinking about buying some of those cases, but i'm not sure about something - if power unit is placed on the bottom it can't take away hot air from the case and pump it out right? All my PC parts are silent - CPU (E8200, placed below 12cm Nochtua fan of power unit) has heat-pipe cooler with Nochtua fan spinning at only 800rpms, GPU has cooler powered by 7V instead 12 and that's why i don't want to HAVE TO place another fan to pump out hot air instead of PU placed on top. That might make some noise. So i ask someone more experienced: if i buy some computer case with PU placed down, do i HAVE TO place some fan to pump out hot air?

    Read the article

  • 2 ATI Radeon HD 5870 (crossfire): Intermittent loud fan and non-functioning secondary card

    - by Merritt
    Installed 2 5870s. Left off the left side panel of my case(cooler master haf x), started my computer, installed the drivers and set up the crossfire... yada yada yada... everything worked fine. Then turned off computer, put left side panel back on, turned on computer, and heard a very loud buzzing (asssuming it was a gpu fan). Checked status of cards (in Windows 7) and the secondary card did not appear. Turned off the computer, took off left side panel, jiggled power connectors into secondary card, then turned the computer back on, and it worked again. Put the panel on, turned on computer, worked. Sometime later, turned on computer, and there's that loud fan sound again. What gives?

    Read the article

  • High RAM usage, not seen in task manager

    - by r4dk0
    Hi! I am using Windows 7 64-bit 7600, with 4Gb of RAM. I have a serious problem, since something uses a lot of RAM(3.94Gb) and I see "stairs" in taskmanager, it rises to +3Gb RAM and it drops to about 2Gb and then rises slowly again, and suddenly drops. I tryed installing this version again and other versions, newer ones, but no effect. Ive even tryed disconnecting other harddrives while I installed it, and then installed NOD32 and updated it. How could I know what is using that much RAM? P.S.: I was suspecting superfetch service, I disabled it, restarted pc, and it didnt work, since the memory is the highest point when I login with password, it is really annoying since I need about 1minute to see my desktop, neither alone try anything else. After loging in it slowly drops and after random time it starts rising again. That doesnt happen immediately after a fresh windows install. And how the drivers go, I tryed older drivers for GPU, and newest ones.

    Read the article

  • High RAM usage, not seen in task manager

    - by r4dk0
    Hi! I am using Windows 7 64-bit 7600, with 4Gb of RAM. I have a serious problem, since something uses a lot of RAM(3.94Gb) and I see "stairs" in taskmanager, it rises to +3Gb RAM and it drops to about 2Gb and then rises slowly again, and suddenly drops. I tryed installing this version again and other versions, newer ones, but no effect. Ive even tryed disconnecting other harddrives while I installed it, and then installed NOD32 and updated it. How could I know what is using that much RAM? P.S.: I was suspecting superfetch service, I disabled it, restarted pc, and it didnt work, since the memory is the highest point when I login with password, it is really annoying since I need about 1minute to see my desktop, neither alone try anything else. After loging in it slowly drops and after random time it starts rising again. That doesnt happen immediately after a fresh windows install. And how the drivers go, I tryed older drivers for GPU, and newest ones.

    Read the article

  • Firefox 3.5.6 Uninstaller not working

    - by Wesley
    Hi all, I want to uninstall Firefox 3.5.6 in favor of Opera 10.10. However, I've went into Control Panel Add or Remove Programs and tried clicking Remove from there, but nothing happens. I go into the Program Files for FIrefox and find the Uninstaller folder and click the helper.exe found in there. Still, nothing happens. Is there some possible way to uninstall Firefox without screwing up the rest of my computer? Computer is a EMACHINES T2482: http://emachines.com/support/product_support.html?cat=Desktops&subcat=T%20Series&model=T2482 RAM has been upgraded to 1 GB PC3200 DDR, GPU has been upgraded to 128MB GeForce 6200, HDD has been upgraded to 160 GB, PSU has been upgraded to 350W and Floppy and DVD Reader have been disconnected from mobo and power. OS is XP Pro SP3

    Read the article

  • Dual-headed graphics card choice: DVI & VGA with 512MB _or_ DVI & DVI with 256MB

    - by TimH
    Which would you choose? Some more detail: I can choose between: A dual-headed card with both heads DVI but only 256MB of memory A dual-headed card with one VGA and one DVI, with 512MB of memory. Both monitors are 1600x1200 I'll be doing mostly business app development on the computer. No gameplay or advanced graphics work. It's running Win7 and is a quad-core i5. I'm thinking of going with 256MB one, just so both displays are DVI and I don't have to shift between sharp & blurry when I look from one screen to the other. But I'm not sure if the additional RAM would be a huge boon for some reason (Win7 GPU acceleration, for example? But with a quad-core, who cares?).

    Read the article

  • How small is too small (of laptop for a programmer)

    - by defn
    Im thinking of replacing my HP Pavilion dv5-1004nr with Acer Aspire AS5738-6444, I'm a CS student... so im torn between screen size (liking the 15in and decent gpu.. but it weight nearly 8 pounds and battery can barely last 2 hours in power saver mode) and wanting something portable to carry to classes/take notes ect (3 pounds and up to 8 hour battery life of the acer) what would you guys suggest? acer any good? i love my hp keyboard.. (or have links to other good laptops, im aiming for 600$ range?)

    Read the article

  • How do I get VDPAU working with Ubuntu 9.1?

    - by Brad Robertson
    What do I need to do to get MKV HD videos playing with VDPAU and also Blu-ray discs? Lots of people say you need to compile the latest MPlayer (which I haven't had luck doing) for VDPAU. I found an mplayer ppa that says it has VDPAU compiled into it so I'd like to use that. What packages do I need for playing MKV files and Blu-ray with the video decoding offloaded to my GPU? So far I haven't had any luck with any of the tutorials I've found. I'm just looking for a quick synopsis that will tell me what I'm looking for as I'm kind of shooting in the dark. (I didn't know what VDPAU was until a few days ago.)

    Read the article

  • Can the NVIDIA ION chipset handle streaming and gaming reasonable well?

    - by true_gritt
    I'm considering getting a small-footprint "nettop" computer to use as a home theater PC with my Samsung LN40A550 HDTV. I've been looking at systems like the AS Rock ION HT330, the Acer AspireRevo 3610, or the Asus EeeBox PC EB1501. These are all systems with NVIDIA ION chipset (Intel Atom N330 dual core CPU + NVIDIA GeForce 9400 GPU). Is the NVIDIA ION chipset powerful enough to support media streaming at HD resolutions (e.g. via Boxee, Hulu, Netflix) and casual gaming (e.g. World of Warcraft, Madden NFL) reasonably well without herky-jerky video output?

    Read the article

  • 1080p playback on external monitor

    - by xibalban
    My system (netbook) specs: 1.6 GHz dual core atom N2600 2 GB DDR3 RAM Integrated GMA3600 GPU Win 7 Professional 32 Bit Daum Pot Player V-1.5 The issue: I downloaded a 1.33 GB sized 1920 x 1080 HD documentary (file extension: mp4) and it plays without a hitch using pot player. However, when I connected an external display (an LED Full HD 24" monitor) using D-Sub, the video plays fine but the player controls become very unresponsive. For instance, it takes about 40 seconds to exit full screen while playing. I tried: Installed/updated all the latest monitor drivers Updated the VGA drivers for GMA3600 No other programs ran while the video played, with most unwanted services and background processes disabled/killed. CPU usage goes only up to 49% when the video plays Question: What may I do in order to improve the player response/video playback experience? Do I need to play with (or change) Pot Player's default video decoders, etc?

    Read the article

< Previous Page | 20 21 22 23 24 25 26 27 28 29 30 31  | Next Page >