Slashdot Mirror

2 phoronix studies combined prove that there is no real world advantage of SSD. Advantage is only in synthetic benchmarks. Note that they used one of the fastest SSDs available (Intel X25-E), and a 7200 rpm laptop hard disk. For a slower SSD, difference might be even lower or hard disk might beat SSD in performance. This SSD was not even available at the time of this MS study.

1. http://www.phoronix.com/scan.php?page=article&item=intel_x25e_ssd_linux&num=1 : Uses ext3 filesystem and compares a hard disk and an SSD . Negligible performance difference in real world tests, huge difference in synthetic tests.

2. http://www.phoronix.com/scan.php?page=article&item=intel_x25e_filesystems&num=1: Using SSD, compared different file systems (JFS, ext4, ext3, XFS). Again, negligible difference in real world tests, but different filesystems did well in different synthetic tests.

2 phoronix studies combined prove that there is no real world advantage of SSD. Advantage is only in synthetic benchmarks. Note that they used one of the fastest SSDs available (Intel X25-E), and a 7200 rpm laptop hard disk. For a slower SSD, difference might be even lower or hard disk might beat SSD in performance. This SSD was not even available at the time of this MS study.

1. http://www.phoronix.com/scan.php?page=article&item=intel_x25e_ssd_linux&num=1 : Uses ext3 filesystem and compares a hard disk and an SSD . Negligible performance difference in real world tests, huge difference in synthetic tests.

2. http://www.phoronix.com/scan.php?page=article&item=intel_x25e_filesystems&num=1: Using SSD, compared different file systems (JFS, ext4, ext3, XFS). Again, negligible difference in real world tests, but different filesystems did well in different synthetic tests.

Multipointer X has been in the mainline development branch of the Xorg X server for a year but wasn't released with X server 1.6.

http://www.phoronix.com/scan.php?page=article&item=xorg_server_16&num=1

It should be released in X server 1.7

http://en.wikipedia.org/wiki/MPX

Well in that case it's TFS and TFA that are misleading - do not take this as representative of open source. E.g., check out http://www.ogre3d.org/gallery/ for some decent screenshots from an open source engine.

The technology is also more comparable to Doom 3 than Quake 2 (it's just that the screenshots on TFA mostly don't seem to show them very well - even the screenshots from version 1 4 years ago showed realtime shadows for example, but for some reason these are disabled in most of the screenshots on the latest article!)

Although it was originally Quake 1 they started with, it looks like they've made major changes to it, adding in newer technology (e.g., dynamic lighting and realistic shadows, which is more comparable to Doom 3 than Quake 1).

I must admit, I was a bit confused. A lot of the screenshots in TFA look rather crappy, no different to the Quake 1 or 2, and don't seem to show off what the engine's really capable of. There's no evidence of anything newer such as dynamic shadows (the lighting on the walls could easily be precomputed lightmaps).

But go to the older article one year ago, and there are much better graphics - e.g., here you have realtime shadows, and what looks like bloom effect, and possibly bump mapping on the walls. There is one screenshot in the latest article that shows realtime shadows from a character, but it's rather hard to see in the screenshot. It seems odd why they released this latest batch in order to show its "impressive graphics".

There is also a lot more in the way of open source graphics engines - e.g., projects like OGRE (check out the Gallery).

http://www.phoronix.com/scan.php?page=article&item=ubuntu_fedora_q209&num=1

This might be useful to someone like me, Phoronix just reviewed the 4890 on Linux with the ATI catalyst drivers:
http://www.phoronix.com/scan.php?page=article&item=amd_radeon_hd4890&num=1 :)

Re: 3, RandR, ... currently doesn't support multiple GPUs. So basically, if you have more than 2 monitors on Linux, you're screwed.

You can manually setup a Screen for each display adaptor in your xorg.conf. If your video card driver supports Xinerama, enable that as well, and you're good to go. If it doesn't, it's up to your Window Manager to support moving windows from one Screen to another. (The odds are low that you have one that does.)

Also, XRandR 1.3 will allow CRTCs on multiple video cards to be combined into a single Screen. [0] It was supposed to be released in April of last year. Who knows what the current schedule is, though? [0]

[0] http://www.phoronix.com/scan.php?page=news_item&px=NjI5OA [1]
[1] Yes, I know... its Phoronix. Not *everything* that they publish is sensationalist tripe! :)

Where they compare 5-year-old 32-bit Solaris x86 GCC to new 64-bit Linux GCC, then crow about how fast Linux is when compared to Solaris:

Our build of Ubuntu 9.04 was depending upon the Linux 2.6.28 kernel, GNOME 2.25.5, X Server 1.6.0 RC 1, GCC 4.3.3, and we opted to use the EXT4 file-system. OpenSolaris 2008.11 is based upon Solaris Nevada Build 101b, uses the 5.11 kernel with X Server 1.3, GCC 3.4.3, and the ZFS file-system.

Heaven forbid that Phoronix would benchmark an up-to-date version of Sun Studio compiler on Solaris. No, they use an old version of GCC.

As I said, Moronix.

The Phoronix benchmarking is intended to provide you the answers as to why. It is to highlight the stuff that has happened.

If performance management is going on within the kernel community, then this shouldn't come up as a shock. The whole purpose of independent testing is that you see something that looks out of place, investigate and resolve. A perfect example is http://www.phoronix.com/scan.php?page=article&item=intel_atom_four&num=1 phoronix article, that showed that SuSE was trailing. This causes this http://lizards.opensuse.org/2008/12/16/comments-on-phoronix-benchmarking-opensuse-111/ discussion.

The question and answer don't need to be provided by the same voice. It is when you have someone questioning, and then someone answering, then you have a discussion, then finally you have progress.

To make it worse, there is virtually no reason that any number of the organizations supporting the leading developers can't invest a small amount of infrastructure and run the tests themselves. Phoronix Test Suite is absolutely trivial to use. The amount of "software development in autopilot" is frightening, this applies equally to Open Source as it does to Proprietary.

If you have an Asus motherboard it's branded as Express Gate. Some models have it in the flash bios, some require a 512 MB image file to be located on an NTFS partition (also the installer is windows). Either way, it boots really fast, 5-10 seconds.

It has Firefox and Skype, Pidgin and a photo viewer. When you exit, the system boots from the hard disk.

Ext4 *is* better, and probably because it benefits from the wiggle room provided by the specifications. The question is if you accept the tradeoff between performance and security. I choose performance, because my system doesn't crash that often.

Take a look at this article linked to in TFA. Seems like a pretty big boost.

And in this case we are entitled to it.

Apparently, the drivers are supposed to be GPL, but no actual source was released.

Beta versions of Virtual Box support DirectX just fine.
http://www.phoronix.com/scan.php?page=news_item&px=NzAyNA

Basically they use a WINE wrapper that runs in your Virtual Machine. It'll run DirectX games at pretty much the same level of performance and compatibility as Wine runs natively.
It's not perfect but it beats the pants off any of the commercial solutions.

In an update defined by new graphics chipsets that were build specifically to accelerate high definition video playback these geniuses are testing the processor performance.

If you are interested in some useful numbers, anandtech did some good competitive tests on the current generation of integrated graphics chipsets. No, these are not inside a Mac Mini, but it provides much more relevant information than this ridiculous article.

Those Anandtech high-def video benchmarks might be worthless for the Mac mini because, AFAIK, the OS X drivers have not yet enabled full hardware decode for MPEG-2, VC-1 and H.264 on the GPU. This had been a Windows-only feature for a long time and was just recently enabled in a Linux beta driver (NVIDIA binary).

The Linux driver looks like a good sign (for eventual OS X support), but high-bitrate 1080p H.264 playback will continue to be a CPU-hog on the Mac mini (with a whirring CPU fan) until the OS X driver gets full hardware decode support.

A Mac Mini looks to be a decent media center if you get a wireless keyboard+mouse and download HandBrake+VLC. A better AppleTV than the AppleTV, since it comes with a DVD player.

Except the Mac mini doesn't have an HDMI port, which cuts down on unnecessary cable clutter between the media center and home theater. The Apple TV has one, of course. All media centers (or "home theater PCs") should have HDMI.

Also, does anybody know if NVIDIA's PureVideo hardware acceleration (for high-def H.264 and WMV/VC-1) work in OS X yet? Linux didn't get support until November 2008 (works great, BTW).

If you're looking for accelerated MPEG4 and HD video playback you won't find that on the Intel board. While they support XvMC fairly well that only does MPEG2.

Last month they released some drivers for the VA-API but that's in their closed source binary blob driver which works very poorly on Linux.

NVIDIA has VDPAU support which will already allow you to play back HD streams without having to fork over for a more expensive, and hotter running CPU.

Phoronix has several Articles about this:

http://www.phoronix.com/scan.php?page=article&item=xorg_vdpau_vaapi&num=1

http://www.phoronix.com/scan.php?page=article&item=nvidia_vdpau&num=1

If you're looking for accelerated MPEG4 and HD video playback you won't find that on the Intel board. While they support XvMC fairly well that only does MPEG2.

Last month they released some drivers for the VA-API but that's in their closed source binary blob driver which works very poorly on Linux.

NVIDIA has VDPAU support which will already allow you to play back HD streams without having to fork over for a more expensive, and hotter running CPU.

Phoronix has several Articles about this:

http://www.phoronix.com/scan.php?page=article&item=xorg_vdpau_vaapi&num=1

http://www.phoronix.com/scan.php?page=article&item=nvidia_vdpau&num=1

I'm currently on a hold for funding reasons, but I've intentionally sought out a motherboard with Intel integrated graphics based on a review that suggests it would be suitable, and possibly the best low-budget option, for watching HDTV under MythTV.

I've heard that full fs encryption on higher end computers has a negligible performance impact (cpu can generally keep up with the hdd) but on lower end machines esp. netbooks, the performance impact can be appreciable. Here is an article with benchmarks

Yes. The article you cited is from 2004. Good morning, Rip van Winkle! There have been a lot of changes. :-)

That's all nice and good, Bruce, but in more recent times (circa 2008) the driver for their Poulsbo chipset is a mess (with a closed-source blob sprinkled on top). Seems rather strange to see in these days of Ati going open source, but there you go. Change, as requested :(

It's been done before and the results weren't pretty. Not only slower in every regard, but when it comes to more advanced features like Pixel Sharers, WINE gets completely smoked.

Linux isn't suited or intended for games. Results for this system simply aren't interesting to gamers.

Interesting theory.

It doesn't seem to be borne out by benchmarks though, at least for Linux.

it does look like an kde4 rippoff

http://www.phoronix.com/forums/showthread.php?t=13515

There are plenty of good reasons why cross platform is an impossibility right now. One is that major game developers have already invested in making a game engine which relies heavily in windows API calls. Oblivion's Engine was used in Fallout 3. Valve has spent millions on their Source engine which they license to other game companies.

Oblivion and Fallout 3 both use the cross-platform Gamebryo engine, and Source is coming to Linux. Source is derived from Goldsrc, which was derived from Quake, which itself rendered both Direct3D and OpenGL. Given the current climate of releasing for PC and console platforms, not retaining some amount of cross-compatibility is a dumb design move.

As for Wine, every time I've tried running something meaningful, it's usually a subpar experience. I don't understand why you argue in favor of an API translation layer over a direct port that could fully take advantage of the target platform (for example, the Mac version of World of Warcraft has iTunes keybinding support and built-in Quicktime video recording).

I'm sure Phoronix will come up with one, otherwise you can download the benchmark and try it our for yourself.

I agree, this article would be a much better candidate (even though they have some of the same points).

Check out this list:
http://www.phoronix.com/scan.php?page=article&item=great_linux_innovations_2008&num=1

I actually think the article "Great Linux Innovations Of 2008" on http://www.phoronix.com/scan.php?page=article&item=great_linux_innovations_2008&num=1 was much better.

Well, I started a rant about ffmpeg not supporting purevideo, but then I did a Google search and found this article which says that as of a month ago that has finally been resolved. Unfortunately it requires the closed drivers so it won't work on Atom but you should be good to go with Nano.

Intel staff were the ones mainly responsible for implementing GEM, so their driver supports it. The open-source ATI drivers recently got a layer of glue to use GEM on the outside without changing much of the TTM-based code that was on the inside. I don't know what nouveau is up to, but the nvidia blob has had a lot of memory management stuff implemented independently for a while now in their X driver.

Phoronix follows a lot of this stuff well.

Just do a few searches for tutorials. Visit a few forums and see what gets stickied as solutions to common problems.

Every time the solution involves editing some file and/or opening a terminal window... you're on to something that is not user friendly.

Windows tutorials start with "click" and Linux tutorials start with "At the sh prompt type..."

In a recent survey the a text editor is still the number one way to configure X. Sure the survey may not reflect all users, but if I were to guess I'd say that there are still quote a few users who end up using a text editor (or terminal) and not just for configuring X.

Think about all the times you've seen the solution "sudo apt-get". Why wasn't the user asked to go click something? These are the type of things that are still kicking Linux in the face.

I'm telling you, if you are curious what sucks about a product, work the general broad access tech support. Places that noob users find easily like linuxquestions.org and Ubuntu forums are a good place to look.

Reality is benchmarks. And Wine sucks at 3DMark06 which is the de facto standard for DirectX. In fact while I was searching for numbers I found lots of apologists like you trying to tell people that they shouldn't run the benchmark.

I wonder why

http://www.phoronix.com/scan.php?page=article&item=882&num=5
While the 3DMark series are synthetic benchmarks, they had worked well under WINE 0.9.46 and served their function of comparing Windows XP versus the WINE performance with Ubuntu 7.10. When using the GeForce 8600GT in these benchmarks comparing Ubuntu 7.10 with WINE to Windows XP, Windows was the clear winner by a landslide. Windows XP was noticeably faster and in some cases was nearly five times faster.

http://linuxhelp.blogspot.com/2006/02/wine-vs-windows-xp-benchmarks.html
And not surprisingly, Wine lags behind Windows XP in the graphics test suite which uses DirectX instead of OpenGL.

Plus new games like Crysis don't work.

http://ubuntuforums.org/showthread.php?t=658415

The only games that work properly on Wine are OpenGL ones, and that's because OpenGL is hardware accelerated (well so long as you use the closed source NVidia drivers not the freetard ones) on Linux.

So much for everything working in reality. Unless by reality you mean what loudmouth evangelists like you post whenever anyone points out flaws with Wine's approach of emulating DirectX on OpenGL when the graphics card hardware is optimized for DirectX.

I found this comment to be amusing. I recalled reading their 4870 X2 review not too long ago. You'll notice cards of vastly different performance levels showing similar speeds until AA & AF get cranked up, a clear indication of the benchmarks being CPU bound. This isn't commented on at all in their summary!

Admittedly there aren't any games in Linux that could stress a card like that but things like that should at least be acknowledged. Whilst it's great that they're testing these cards in Linux they really do need to improve their methodology.

Remember, this is the site that's decided that Ubuntu 7.04 is twice as fast as any other version of Ubuntu. Take what they say with a good healthy dose of skepticism.

In the case of taxes, there is more or less a gun held to your head that forces you to do such a thing. Not so with Apple.

It is a gun, or rather more like Dr. Evil with the finger on the nuclear launch button. Let me quote you a little from AMDs John Bridgman on the WTF state of DRM:

If the worst case was as minor as not being able to release any more specs we wouldn't be worrying so much. The kind of risks we are worrying about are much larger, ie things which would either kill or cripple our graphics business.

Worst case is that we lose the ability to sell our products into the Windows market as a result of releasing info which results in our DRM implementation no longer being considered sufficiently robust. Without the Windows market (which is >90 % of our revenues) we would, for all practical purposes, cease to exist as a GPU manufacturer, especially since we would probably lose the Mac market at the same time.

Next worst case is that we find a way to continue shipping into the Windows market but get sued under one or more of the DRM-related agreements we have signed. These all have high dollar-value penalties, again enough to significantly impact our ability to continue operating.

There are a bunch of smaller risks but we spend proportionally less time worrying about them. What makes all this complicated is that we have to consider not just the information we release but the information which is likely to be reverse-engineered and published. Each time we release information we simply raise the bar for where reverse engineering starts, and it's the combination of released plus "likely to be reverse-engineered" info that we need to consider.

If we tripped any of these risks then the impact would not only affect the GPUs we are shipping today but anything we have in the pipe. Best guess is that we would lose the next generation (ie the one after 7xx) and see significant delays in the one after that.

I don't think that's enough of a reason. Most other video decoding doesn't use much hardware acceleration either -- my Linux + NVidia system might be able to accelerate MPEG-2 decoding or so, but that's pretty irrelevant because that's pretty fast to decode anyway.

Support for hardware accelerated H.264 decoding is beginning to appear (perhaps also on Linux, with NVidia releasing VDPAU support for Linux), but to the best of my knowledge, most systems aren't hardware accelerating it anyway.

It might be, though, that the slow part is video output rather than decoding the input stream, and the reason you suggest might be significant for that. I think I recall reading that at least Flash 9 for Linux uses OpenGL for scaling etc., and that might well be slower than using hardware accelerated video overlay like most video player software.

I thought the plan was to move to DRI2 and Gallium3D http://www.phoronix.com/scan.php?page=news_item&px=NjcyMA

elanthis at phoronix explained how Wayland was never intended for normal desktops running Gnome/KDE:
http://www.phoronix.com/forums/showpost.php?p=51097&postcount=6

It will take something different to replace Xorg.

And then out of no where...

http://www.phoronix.com/scan.php?page=article&item=nvidia_180_vdpau&num=1

crazy

You might want to read a blog post I wrote about why nVidia rocks when x.org does not. It's likely to give you more reasons to move over to nVidia over ATi.

I don't find your arguments compelling.

For one thing, you assert that "because of vocal powers in the foundation that demand that things should stay compliant to a specification and they should work around the architecture rather than strip out certain pieces and implement them, add proper new features (memory management and API functions to go with it)" -- yet my reading of the Xorg mailing lists suggests that is exactly what is being done with the GEM memory manager and API's previously there was the TTM memory manager, but the APIs were not satisfactory, so they ripped it out and started again.

The bulk of your argument seems to be that Nvidia's got a much more complete OpenGL implementation than does anyone else. Nevermind that almost all of it is simply code duped from their MS Windows driver, your argument is really the ages-old "if it works, then who cares if it is closed source" argument we've heard time and time again.

Of course the fallacy of that approach becomes obvious the second it stops working and you are helpless to do anything about it.

That happened to a guy I know, he spent about $600 on a pair of top-end nvidia cards a few years back. All based on nvidia's highly touted support for linux. Except the cards did not work with his IBM T220 monitor. It wasn't anything to do with the ultra-high resolution. It was a trivial bug in the nvidia drivers - if the card could not read an EDID, the drivers assumed the card had a single-link DVI transmitter. A stupid, stupid bug because the actual nvidia chip had the DVI transmitters onboard and they were always dual-link, there was no way for any card in that generation to even be single link, and of course no matter what directives we specified in the config file, the driver "knew better."

He had to go out and spend another ~$150 for two Gefen DVI Detectives just to enable the nvidia card to see an edid so that the driver would correctly turn on the chip's DVI transmitter.

Nvidia's vaunted customer support? Totally clueless and useless, they completely dropped the ball, just ignoring the issue once they realized it was more than a "did you plug in the power cord" level issue.

And don't think that problem was unique to an odd-ball monitor - the same lack of edid is an issue for anyone using unidirectional fibre DVI extender cables.

So, while it is great for you personally that Nvidia's drivers work perfectly with the hardware you own, I'm pretty sure your tune would change right quick if you had to just bend over and take it due to such a trivial bug, the kind that could easily be fixed with a single line or two of code, if you just had the source.

VDPAU sounds like some sort of local hawaiian cure for venereal disease
- VD Pau! Fo wen you ala-alas stay too itchy.

Meanwhile, it is interesting that after many years, Nvidia finally starts to support video decode/playback acceleration just days after ATI ships a driver with similar hardware acceleration support. Of course neither vendor uses any sort of common standard - although ATI claims their stuff is almost identical to the Direct X Video Acceleration (DXVA) API that MS has enforced on Windows.

You might be interested to know that ATI's equivalent was also revealed a short while ago.

With a little luck, all of them will, pretty soon.

NVidia just released their HW accelerated video decoding API
http://www.phoronix.com/scan.php?page=article&item=nvidia_180_vdpau&num=1
http://www.nvnews.net/vbulletin/showthread.php?t=123091

AMD pretty close to doing the same with their alternative
http://www.phoronix.com/scan.php?page=article&item=amd_xvmc_xvba&num=1

And Intel is backing VAAPI as well as apparently considering extending XvMC to support more recent video codecs as a stop-gap measure.

It's rather unfortunate that they're all separate and mutually incompatible (and presumably proprietary, for the first two), which will no doubt hinder adoption by application developers, but it's a lot more than what we had until now.

With a little luck, all of them will, pretty soon.

NVidia just released their HW accelerated video decoding API
http://www.phoronix.com/scan.php?page=article&item=nvidia_180_vdpau&num=1
http://www.nvnews.net/vbulletin/showthread.php?t=123091

AMD pretty close to doing the same with their alternative
http://www.phoronix.com/scan.php?page=article&item=amd_xvmc_xvba&num=1

And Intel is backing VAAPI as well as apparently considering extending XvMC to support more recent video codecs as a stop-gap measure.

It's rather unfortunate that they're all separate and mutually incompatible (and presumably proprietary, for the first two), which will no doubt hinder adoption by application developers, but it's a lot more than what we had until now.

http://www.phoronix.com/scan.php?page=article&item=xorg_wayland&num=1

It is a new (very experimental) graphics server that would end up replacing the current X server in Linux.

Overall the benchmarks suggests that Linux (not just Ubuntu) needs some work on the graphics system and the Intel drivers.

Since they only tested on a Mac Mini, what the results actually suggest is that an operating system distribution that's been finely tuned for a very small set of hardware beats a generic distribution that's currently running on thousands of different hardware configurations. If they actually wanted to draw some generic conclusions about Ubuntu versus Mac OS X, then they should've installed both on as many different hardware platforms as possible, and then run their benchmarks and aggregated the results.

So what might be happening is that Ubuntu Linux isn't as optimised to run on a Mac Mini as OS X - imagine the surprise.

Another problem with the Phronix methodology is that they've made no effort to identify exactly why they're seeing differences. For example, see page 8 of the results. The "Bork file encrypter" should either be limited by CPU and memory bandwidth for the actual encryption, or by the speed of the hard disk if it's reading files, encrypting, and writing them back. Given that the limiting factors here are hardware, there's no way that MacOS X should be 27% faster than Linux on this benchmark. Or on the same page, the Java Scimark v2.0 benchmark shows OSX being 370% faster than Ubuntu x86. Given that the performance of Java code is dominated by the JVM performance, this is indicating a massive regression between java 1.5 (OSX) and 1.6 (Ubuntu). Has Sun really allowed this to happen? Or is the Phronix testing methodology broken? My money's on the latter.

Based on this: http://www.phoronix.com/scan.php?page=article&item=ubuntu_bench_2008&num=1

For example you can read an Ubuntu/Mandriva/Fedora comparison. You can find other benchmarks by scouring the Phoronix articles page. I have a feeling the result is usually a wash with no distro coming out on top on all tests though...