Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?
DEAL: For $25 - Add A Second Phone Number To Your Smartphone for life! Use promo code SLASHDOT25. Also, Slashdot's Facebook page has a chat bot now. Message it for stories and more. Check out the new SourceForge HTML5 internet speed test! ×

Comment Re:Overview of Apple connectors (Score 0) 791

You are only repeating what the others already said. I thought Apple-users were original (think different) and found the environment important, but it seems neither is true - you all think the same different and don't give anything about the environment. You only care about design-decisions.

Doesn't matter. Just updating my image of Apple-users.

Submission + - Google blocked OpenCL on Android 4.3 (streamcomputing.eu)

Vincent77 writes: Thanks to the work of ARM in the beginning of this year, we could use OpenCL on the Nexus 4 and 10 and starting to port the various libraries to Android, initiating the same revolution of accelerated software on smartphones and tablets as we had on the desktop. Google was not happy with the competition for their RenderScript Compute and abruptly blocked access to the OpenCL-driver in Android 4.3. Noteworthy is that Google did not have the choice to simply remove the driver itself, as ARM implemented the RenderScript-compiler on top of OpenCL.

Comment Re:Not that fast at all (Score 1) 81

Thanks for the link. Release date of iPad 4 is December 2012 (!) and has a much higher resolution to handle. Snapdragon 800 was designed for use in a (high-end) phone. The active cooling suggests Tegra 4 is not fit for mobile devices - I still think the reason for this device is to dump their unsold processor tech.

I have to say I would welcome mini-computers in the range 15 - 30W. For notebooks this Tegra 4 would be interesting.

Comment Not that fast at all (Score 1) 81

According to Anandtech only 74.8 GFLOPS - comparable to an iPad 4. Other sources say 96 GFLOPS, but only when in power-hungry overclock mode: image. The real winner for Q4 2013 will be the Qualcomm Snapdragon 800 - 129 GFLOPS. That leaves Tegra 4 completely in the dust.

The main reason the Tegra 4 is in no tablet/phone, is because Tegra 3 real performance and power usage was worse than advertised/marketed, and therefore the tablet/phone makers did not trust Tegra 4 would be a good bet. Another (smaller) reason was that NVIDIA is quite pushing their own agenda and brand, whereas other vendors do not meddle with their customer's business so much. Unluckily they did not learn from their experience and suggest in their latest video (the face-demo) that Tegra 5 uses 2 to 3 Watts when under full load - truth is that the load was not given. NVIDIA knows a little too much about marketing...

Comment A marketeer wrote the article (Score 4, Informative) 195

There are several languages that are written on top of OpenCL - that is the whole idea of this API. But if your read the article, it seems this guy was the actual inventor of the wheel.

Same response happened when some guy made Rootbeer and let some marketeer write an alike article. It was suggested that you could just run existing Java-code on the GPU, but that was not true at all - you had to rewrite the code to the rootbeer-API. This Harlan-project is comparable: just beta-software that has not run into the real limits of GPU-computing - but still making big promises that in contrary to their peers they actually will fix the problem.
I'm not saying it can be in the future, but just that this article is a marketing-piece with several promises on future advancements.

Check out Aparapi and VexCL to name just two. There are loads and loads of these solutions - many of these wrappers slowly advance to higher level languages, and have been in the field a lot longer.

Submission + - NVIDIA abandons open standard OpenCL (ipetitions.com)

Vincent77 writes: "To keep NVIDIA from undermining the open standard OpenCL for parallel computing on GPUs, CPUs (via AVX and SSE), FPGAs and ARM, a petition has been initiated to call out for support for the standard:

Nvidia is not including OpenCL samples in the latest CUDA SDK, focusing more on CUDA instead. As a Khronos member with an excellent record in implementing and promoting standards like OpenGL, this is a surprising and unacceptable behavior from Nvidia.

We, OpenCL developers, humbly request that OpenCL samples again be included in the SDK. Industry standards like OpenCL help in building up a bigger market for GPU computing, and will be beneficial to Nvidia in the long term.

In a presentation for a small group the unofficial statement was that they want to sell chips *now*, and not wait for a standard to mature: "For NVIDIA OpenCL is not the answer". The remark that OpenCL can be extended to support these needs, was answered with that you don't want unportable code. Funny is that you can use #IFDEF in such cases, but not if CUDA-code needs to be mixed with OpenCL.

This is just a matter of increasing their profits, while making it hard for coders who want to make performance-code for all modern hardware."

Comment That's easier than maintaining Unix boxes (Score 1) 454

When there is a problem with Linux, you fix it. When there is a problem with Windows, you replace it. Make sure you have updated Windows images - this is possible by making a daily image of the system in VMWare is alike. Tweaking of software like databases is possible, but it is cheaper (in terms of hourly price) to just buy more memory and more processing power. Actually these tricks do pretty well on Unix-systems too as an extra solution. As you cannot look into the code, you have to trust others. So make sure you have a good firewall and antivirus-software. I'm not tryng to make fun of windows here; Windows is a black box, so you need to treat it as such.

Slashdot Top Deals

C'est magnifique, mais ce n'est pas l'Informatique. -- Bosquet [on seeing the IBM 4341]