Slashdot Mirror

Primarily, it shows that the free drivers made by mostly unpaid people are so good that even the producer of the hardware uses them rather than their own code

Most people working on the Radeon driver are paid, as far as I know, even if not all by AMD.

Did they take the code for free because they could? Yep. Did they offer anything back? Not necessarily money, but what about saying "really good work guys, we will use it as well"? Nope. Why bother, it is free, isn't it?

Not true. AMD has offered a bunch of specs which were invaluable for the development of the drivers and contributes code too.
Not to mention their contributions to a bunch of other open source projects: http://developer.amd.com/zones/opensource/pages/default.aspx

While not perfect, they're better than most other companies, which don't contribute anything at all.

We should remember this when someone from the large business starts babbling about "appropriate award for our large investments and the hard work of our engineers" when explaining why we all have to pay again and more for their crappy software.

When did AMD start selling software? I paid them for hardware.

The point of Bulldozer and pairing it with AMD GPGPUs is to leverage all the wealth of work AMD has put into OpenCL 1.1 with OpenGL 4.x.

When more and more apps leverage OpenCL 1.1 [and the list is growing rapidly] using the likes of LLVM/Clang where AMD has worked hard at leveraging you'll begin to see a lot of these ``benchmarks'' being truly useless and tuned specifically for Intel.

The work AMD is putting in with that marriage should be obvious: http://developer.amd.com/pages/default.aspx

Until applications truly leverage the architecture in conjunction with the AMD 6000/7000 GPGPUs talking about game benchmarks is truly juvenile. I mean seriously folks. Grow up, sit back and watch this platform advance and those same games become tuned for that CPU/GPGPU marriage.

And why would I want that when I can get an AMD APU + motherboard combo for that same price, but with 80 Radeon cores for the same price as the underpowered g620 (*without* the motherboard)?

http://www.amd.com/us/products/desktop/apu/mainstream/Pages/mainstream.aspx#3

http://www.guru3d.com/article/amd-brazos-platform-tested-e350-apu-review/1

Heck, for your Intel's price and TPD, I could get a triple core A6-3500 with Radeon HD 6530D integrated on the die.

If someone is not limited to PCI due to external factors, there is little reason to get g620 over AMD's APU.

90C is the max, so clearly less is better than 90C.

60's are well below 90C.. quite a bit of headroom there.

AMD's factories are in China, Malaysia, and Singapore. Intel's are in the US, Ireland, Israel, China, Malaysia, Vietnam and Costa Rica.

I seriously doubt you can claim any moral high ground by using AMD.

http://en.wikipedia.org/wiki/List_of_Intel_manufacturing_sites
http://www.amd.com/US/ABOUTAMD/CONTACT-US/Pages/locations-type.aspx

You'd better be calling The Good Lord for help right now, because while you weren't looking specialized chips like that DID become the dominant platform, and GPGPU is going to make it more so. The primary processor (which is also from a duopoloy) will just become a system management hypervisor.

Read up on the technological cycle of reincarnation. "All this has happened before. All this will happen again."

Two 5870 running at full will be 350~400 Watts Each.

Add in the motherboard and other basics you're talking 1000 Watts constantly.

Nice job pulling those numbers out of your ass.

Here's the real power consumption of a 5870 right off of AMD's spec sheets: http://www.amd.com/us/products/desktop/graphics/ati-radeon-hd-5000/hd-5870/Pages/ati-radeon-hd-5870-overview.aspx#2

I'll pull the relevent part out for you: Maximum board power: 188 Watts

Assuming people who bitcoin mine use at least a decent power supply that is 80% efficient PSU at given load (realistically most decent ones are 82%+ in optimal load range), you're going to be pulling 235 watts from the wall per card, max.

235 watts is way less than 350-400 watts, by a long shot.

The rest of the system isn't going to be pulling huge amounts of power, since nobody who is mining bitcoin for real cash does it on a CPU, they do it on GPUs, and the amount of power a motherboard, RAM, disk drive, CPU use while they aren't really working is pretty low, usually in the 30-60 watt range, depending on your CPU, but nowhere near 200 watts of draw

No, AMD does not make a C++ compiler.

what about Open64?

So in other words you get the "choice' of cripple ALL CPUs, or just Intel's competitors? Wow their douchebaggery just gets better and better don't it? Especially when it would be trivial to check for the SSE flag (which EVERY CPU that has SSE has implemented for nearly 8 years, which is a lifetime in CPUs) and "If CPU = SSE then run optimized' end of story.

The sad part is AMD does have their own compiler which is completely FOSS and unlike Intel doesn't try to hamstring a CPU because it isn't made by them. Pretty damned sad when a company feels they can't just compete on their merits, nope, they gotta pull dirty backhanded bullshit, just to make sure they can't lose.

This is why I sell nothing but AMD in my shop now and my customers couldn't be happier. They like the low prices on triples and quads, I like supporting a company that isn't a douchebag, its a win/win.

...and make a compiler.

They did. It's even GPL licensed.

http://developer.amd.com/tools/open64/Pages/default.aspx

Um, they do have their own compiler: http://developer.amd.com/tools/open64/Pages/default.aspx

Seems to be both free to download, and comes with source code so you can go over it if you wish.

Someone above has already posted this : AMD64

That's great! Why doesn't AMD go and write a compiler of thier own and give it away for free?

Its called Open64.

ATI doesnt exist any more you must mean AMD http://sites.amd.com/us/game/products/graphics. Graphics card manufacturers tend to come and go and nowdays with the push to make the GPU a co-processor on the core CPU chip, graphics cards are looking at the end of their existence.

So Nvidia would be looking to partner with a CPU manufacturer, M$ is just getting in the mix to retain some control. Once you have high powered computer in a chip, with the price saving inherent in that, the software licence becomes a real burden.

Originally M$ had intended to make a shift to an internet company via MSN as the closed source software monopoly dried up, catch was they proved a failure ta making money out of MSN, strangling the chicken, in terms of trying to hard to squeeze out profits and attacking the creative types in meetings, was their undoing.

So they might be making a jump to hardware using their software monopoly to leverage out a hardware advantage, based upon their past performance, not only will they get sued all over the place they will also make a complete balls up of it.

That's funny. I have poclbm running on my Mobility Radeon 5830, and both my desktop's Radeon 6950's (unlocked to 6970's and overclocked).

Perhaps you need to install the OpenCL dev kit or something? Because if the consumer level parts can run the miners, the pro level cards should have no problem. Try downloading the drivers from AMD and not HP: http://support.amd.com/us/gpudownload/windows/Pages/radeonmob_win7-64.aspx

I don't know for sure, though.

So you're telling me the AMD Radeon HD 6770M listed on the iMac's tech specs is in fact not at 6770M but a 6770? Is this an error on Apple's site? Where are these benchmarks you speak of? This was one of the things that stopped me from getting an iMac, but if you're right it changes my mind.

So you're telling me the AMD Radeon HD 6770M listed on the iMac's tech specs is in fact not at 6770M but a 6770? Is this an error on Apple's site? Where are these benchmarks you speak of? This was one of the things that stopped me from getting an iMac, but if you're right it changes my mind.

Thought I'd write up a quick 'getting started' guide for anyone that wants to give bitcoin mining a go:

#1 - Download the bitcoin application from bitcoin.org, install and fire it up. It will connect and sync with the p2p network, downloading approximately 114700 blocks.

#2 - Download and install the OpenCL driver for your graphics card / OS.
You might also need the full SDK, my drivers were supposed to include OpenCL support, but the GPU miner still didn't work. For AMD/ATI cards, this link should work:
http://developer.amd.com/gpu/AMDAPPSDK/downloads/Pages/default.aspx

#3 - Download and unpack "PyOpenCL bitcoin miner" somewhere. You'll find windows binaries here (7zip-compressed):
https://github.com/m0mchil/poclbm/downloads

#4 - Using the bitcoin client, create a new 'receiving address' which you call 'mining income' to track payments.

#5 - Sign up for a mining pool. You'd rather have a few cents an hour than wait months for a random shot at 50 BTC. I'd go with:
http://www.bitcoinpool.com/newuser.php
as they're free, while the others charge a fee of 2-3%. Wallet ID is the thing you created in step 4.

You'll find the other pools here:
http://www.bitcoin.org/smf/index.php?board=14.0

#6 - Stuff the following into a .bat file and run it. Might want to try from the console first, to make sure all is ok.

start /DD:\bin\bitcoin\poclbm poclbm.exe -f 60 --host=bitcoinpool.com --port=8334 --user=username --pass=password -d0 -v -w 128

This of course assumes you're on windows, and installed to a directory named d:\bin\bitcoin\poclbm..

Setting the f options to a higher value will cause less stress on your system. 30 is the default, shoot for 120 if your screen is lagging too much.

The d option is the device id of your graphics card. Mine's device 0, it could also be 1, 2 or whatever.

If the above worked, you should see a console window containing output like this:

23/03/2011 17:18:55, long poll: new block
23/03/2011 17:19:27, b15bbc4d, accepted
23/03/2011 17:19:47, 97f98213, accepted
23/03/2011 17:20:04, 2a8d658f, accepted
23/03/2011 17:20:15, 96fd6e6e, accepted
160772 khash/s

Or he could just work at Best Buy, as when I filled in there for a few days for a friend I found damned near all the guys in the back had USB drives with .bat files to rip every tune, video, and naughty pics that crossed their desks.

As for TFA the AMD Fusion Media Explorer is free and pretty nice, or you can create all kinds of custom libraries in Windows 7 pretty easy. That said it would be a royal PITA to fill in all the blanks but it sounds like what you are wanting is a media DB. Making your own DB isn't hard (just the filling in the shitload of data is hard) and with your own DB you can add as much or little info as you like and sort any way you want.

So there are several ways you can go about it and none are more "right" than another, it is more of a personal preference thing really. I use WMP 12 for my tunes and Windows 7 MC software for my video, works for me, but it may be too much or too little for you just depending on how deep you want to go here.

Okay, I see... IGP means to you that it has dedicated memory, according to wikipedia, that's even correct.

Personally, we bought that iMac mainly for surf-duty, and I never intensively studied its graphics capacities. I am surprised to learn it has its own graphics memory. Good, I'm wrong on that... Still, I doubt that a ATI Radeon HD 4200 (which definitely is an integrated graphics chipset) couldn't drive a 256x1600 screen... Yup, it can... : "Primary supports 18-, 24-, and 30-bit digital displays at all resolutions up to 1920x1200 (single-link DVI) or 2560x1600 (dual-link DVI) "

You have been living under a rug in terms of computer gaming I guess.

All games support these new multi-monitor setups because it's been built into the video card drives for about a year now. The drivers present the game software a single combined resolution which I've never seen any game not support. Yes even stuff like Quake 1 works across 3 monitors like this.

Also some people want to run 3D setups where it must render the game twice at different angles and it takes roughly double the GPU power.

http://www.amd.com/US/PRODUCTS/TECHNOLOGIES/AMD-EYEFINITY-TECHNOLOGY/Pages/eyefinity.aspx
http://www.nvidia.com/object/3d-vision-surround-technology.html

In your experience does it work well (crashing etc?)?

Last time I worked with it was in 2002.

I'm somehow inclined to believe the complaints :). I'm also having difficulty finding the mythical more up to date version (that's mentioned in that thread) of Hydravision on AMD which works with more apps

Select your driver here, then click on the "optional downloads" tab. Supposedly it will give you the current version.

Nview has been around since 2002 or so, so I guess they've had time to figure stuff out :).

HydraVision was originally developed in late 90s by Appian Graphics and bought by ATI in July 2001 along with the development team (2 or 3 people, as I recall). I joined in September and left some 10 months later after figuring out that it was a dead end for me. By that time one of the developers was let go and a short time later, the team lead (who actually knew the code) passed away. I can tell you that the source was quite a mess and I felt that a fair bit of it should have been rewritten. Of course I did not get to do that although I did fix several bugs. That said, it was nine years ago so a lot may have changed since then.

In your experience does it work well (crashing etc?)? I see complaints about it only working with some stuff: http://forums.amd.com/game/messageview.cfm?catid=279&threadid=119851

I'm somehow inclined to believe the complaints :). I'm also having difficulty finding the mythical more up to date version (that's mentioned in that thread) of Hydravision on AMD which works with more apps despite looking in:
http://sites.amd.com/us/game/downloads/Pages/downloads.aspx
http://support.amd.com/us/Pages/AMDSupportHub.aspx

In my experience Nview works with chrome etc, it only has problems moving "full screen" stuff to another screen, but can I understand why that might not be supported (except it makes showing powerpoint viewer stuff on a different monitor difficult). Nview has been around since 2002 or so, so I guess they've had time to figure stuff out :).

In your experience does it work well (crashing etc?)? I see complaints about it only working with some stuff: http://forums.amd.com/game/messageview.cfm?catid=279&threadid=119851

I'm somehow inclined to believe the complaints :). I'm also having difficulty finding the mythical more up to date version (that's mentioned in that thread) of Hydravision on AMD which works with more apps despite looking in:
http://sites.amd.com/us/game/downloads/Pages/downloads.aspx
http://support.amd.com/us/Pages/AMDSupportHub.aspx

In my experience Nview works with chrome etc, it only has problems moving "full screen" stuff to another screen, but can I understand why that might not be supported (except it makes showing powerpoint viewer stuff on a different monitor difficult). Nview has been around since 2002 or so, so I guess they've had time to figure stuff out :).

In your experience does it work well (crashing etc?)? I see complaints about it only working with some stuff: http://forums.amd.com/game/messageview.cfm?catid=279&threadid=119851

I'm somehow inclined to believe the complaints :). I'm also having difficulty finding the mythical more up to date version (that's mentioned in that thread) of Hydravision on AMD which works with more apps despite looking in:
http://sites.amd.com/us/game/downloads/Pages/downloads.aspx
http://support.amd.com/us/Pages/AMDSupportHub.aspx

In my experience Nview works with chrome etc, it only has problems moving "full screen" stuff to another screen, but can I understand why that might not be supported (except it makes showing powerpoint viewer stuff on a different monitor difficult). Nview has been around since 2002 or so, so I guess they've had time to figure stuff out :).

Just found this:
http://www.amd.com/us/products/technologies/ati-xgp/Pages/ati-xgp.aspx

Sounds cool. I'd like that.

See http://www.amd.com/us/press-releases/Pages/amd-appts-seifert-2011jan10.aspx

Some very interesting analysis can be found at:
http://www.brightsideofnews.com/news/2011/1/10/coup-at-amd-dirk-meyer-pushed-out.aspx
"Remember, Dirk Meyer’s three deadly sins were:

1) Failure to Execute: K8/Hammer/AMD64 was 18 months late, Barcelona was deliberately delayed by 9 months, original Bulldozer was scrapped and is running 22 months late -I personally think this is not true; Dirk Meyer was AMD's CEO from July 18, 2008 until January 10, 2011; he could not be responsible for K8 nor Barcelona, however Bulldozer...-
2) Giving the netbook market to Intel [AMD created the first netbook as a part of OLPC project] and long delays of Barcelona and Bulldozer architectures -this is interesting, after Intel has a serious failure with the Pentium 4, it's mobile division is the one who changes everything with Intel Core 2, designed from a mobile perspective-.
3) Completely missing the perspective on handheld space - selling Imageon to Qualcomm, Xilleon to BroadCom -I think this is the key; no one expected this market to be as successful as it is at the moment-"

I didn't see any mention of VAAPI or XvBA Acceleration for playing media? How about OpenCL support?

Granted the HD 6000 looks more like a gamers card than something you'd stick in a home theater pc, but I'd think that OpenCL support would interest quite a few people doing massive number crunching. Especially since there's even PyOpenCL available.

Things have changed. As of like, 6 months ago. Try to keep up.

I believe AMD and ATI were working on that?

Got that right! I've also heard tale of the APU. I think Intel might be coming out with one as well, but as far as I know AMD is the first. Here's a link to site with some information on it. http://sites.amd.com/us/fusion/apu/Pages/fusion.aspx

Yep. The name's pretty obvious too. ( http://fusion.amd.com/ )

Yeah it is truly crazy how much "bang for the buck" from AMD and lets be honest here: VERY few of us are gonna have the kind of day to day work lined up that is gonna pound the dog snot out of the CPU hard enough to make the price difference worth it. For less than $530 after MIR I got an AMD 925 quad with 8Mb of Lvl 2 cache, 8GB of DDR 2 800MHz RAM, 2 500GB HDDs, a HD4650 1Gb GPU, a 20x DVD burner, and a nice case to put it all in. You really can't beat that.

And everyone here on /. is always talking about "voting with your dollars, now here is your chance. We have Intel being caught in bribery, rigging their compiler, and bribing OEMs so badly that nearly 40% of Dell's "profits" some quarters were nothing but Intel kickbacks. If you value a free market you might want to look at something like what I linked to below. And don't forget FOSS guys that AMD has been good to the FOSS community, as they have been opening up the ATI specs as fast as they can crank out the docs and they also support the x86 Open64 Compiler which unlike Intel's accelerates BOTH Intel and AMD CPUs.

So if anyone is shopping for desktops this Xmas? How would you like a quad desktop fully loaded for $199? That is with a Phenom X4 at 2.4GHz, 2Gb of RAM, DVD burner, 500Gb HDD, and case. Just buy a CPU fan and put whatever OS suits your fancy. Hell with prices so cheap when my dad decided it was time to retire his old P4 I went ahead and got him almost the same deal, just went with 4Gb of RAM instead of 2. Will my dad ever need that kind of power? I doubt it, but with C&Q it drops down to just 800MHz and is whisper quiet, and I know that for the foreseeable future he will not need anything done to this PC. Well, except for adding more USB ports. I swear that man can go through USB ports like crap through a goose.

APU doesn't standard for Applications Processing Unit, it's an acronym for Accelerated Processing Unit.

http://sites.amd.com/us/fusion/apu/Pages/apu.aspx

"The GPU, with its massively parallel computing architecture, is increasingly being leveraged by applications to assist in these tasks. AMD software partners today are taking advantage of the GPU to deliver better experiences to across an ever-wider set of content, paving the way to breakthrough experiences with the upcoming AMD Fusion Family of Accelerated Processing Units (APU)."

This is a photo of Leslie Sobon Alright, that's pretty good, but then these are also photos of Leslie: one two three four That's quite a range there, never know what you're going to get.

Are you new to dealing with human beings (I was going to say "women" but then realized even that was too narrow)?

Every single one of those pix looks like the same attractive woman. But people look a bit different from time to time, like when they are working vs when they are posing for an "image" shot. It is exactly the same with guys as with gals, except that you don't have the same "appearance police" mentality scrutinizing the guys for imperfections like you do with gals.

This reminds me of the candid paparazzi snaps of, say, Jennifer Aniston picking up some hygiene products on a midnight run to 7-11 at midnight in her sweats. Someone who is possibly the world's most beautiful woman (sorry Ashwariya) looks kind of dowdy under those conditions but she is still the same person and could look absolutely dazzling later in the morning. If she doesn't pass your "always must look beautiful test", its your problem not hers.

Guys who expect woman to be glamorous 24/7/365 are either very rich or idiots. Both sets are jerks.

The submitter linked to an site that links to the actual posting. The real posting it at: http://blogs.amd.com/home/2010/09/22/getageek/

• My 3D Love Affair
• Microsoft Office 2010 is Sexy
• What Women Want

It seems like she would rather be a VP of Marketing at a tabloid than a technology company.

This is a photo of Leslie Sobon

Alright, that's pretty good, but then these are also photos of Leslie:
one
two
three
four

That's quite a range there, never know what you're going to get.

Well, if you read the actual blog post instead of that jackass written article, you will find the blog is actually pretty light and makes little, if any offensive comments or insinuations. Really, it's pretty cut and dry and somewhat silly. So, after three posts, I've decided Gareth Halfacree is a total douchebag, and the author of the blog itself, Leslie Sobon comes off as pretty date-able. Also, I posted three separate times because everything is better in triplicates, and work is extremely slow today.

Well, if you read the actual blog post instead of that jackass written article, you will find the blog is actually pretty light and makes little, if any offensive comments or insinuations. Really, it's pretty cut and dry and somewhat silly. So, after three posts, I've decided Gareth Halfacree is a total douchebag, and the author of the blog itself, Leslie Sobon comes off as pretty date-able. Also, I posted three separate times because everything is better in triplicates, and work is extremely slow today.

After researching some more, a socket G34 motherboard will support Bulldozer, so that's another huge plus (in my mind).

Sometimes it is harder to get an OEM computer to use AMD (like apple) but according to AMD's website: Powering ultrathin notebooks to blade servers, all AMD processors shipped are designed to use AMD-V features. Where as Intel has been a little less free and more cumbersome. For instance most Atom processors by Intel do not support virtualization but all shipping AMD (and it has been a while) do. Also computer models such as the sony viao (undercapitalized for a reason) use the "feature" provided by Intel to disallow virtualization through the BIOS, meaning that you have to turn in on before booting. Along with other technology that AMD has developed makes you wonder why Intel is so dominant in the space. So for an informed geek, switching to AMD was already a good move, if only the manufacturers would follow.

If the course will involve teaching 3D graphics programming, I highly recommend you look at RenderMonkey
I've found it to be really useful for prototyping shaders. One of its most useful features is that it allows developers to mark variables as "artist variables", which are then mapped onto sliders / colour-pickers etc. This allows a user to tweak parameters in the shader and view the results immediately, without having to recode it.
Assuming the engine you ultimately decide upon supports loading custom HLSL / GLSL shaders, your students could potentially apply their own shaders to assets in the games they create, allowing them full creative control over things such as lighting and blending.

If the course is going to include 3D graphics, I highly recommend looking at RenderMonkey.
I've found it very useful for prototyping shaders, and it allows developers to define 'artist variables' that are mapped to sliders, colour-pickers etc., allowing users to change various shader parameters and see the effects instantly without needing to recode.
This also helps open up the possibility of allowing your students to add their own shaders to games they create, assuming the engine / framework you eventually choose supports this.

You mean something like Fusion?

L1 CPU caches are shamefully stuck with the laughable 20-year old 640K meme in rarely noticed ways. Everyone's first thought is about RAM memory, but remember that CPU's are less change friendly and benefit more from tech like 128K * 5 size at the new density improvement.

Our supposedly macho CPU's have only 128K L1 sizes and comparably, absurdly high L2 and L3 sizes to make up.

The current excuse is that cost and die-space constraints keep size-improvements mostly on the L2 and L3 side. Sadly, someone tagged the article "tenyears" and we'll be dealing with different research by then, like utilizing today's 64 bit, multi-core technology to its fullest.

There is also this on the AMD site. It has a slightly different take on the core/module semantics.

To suggest it is anything OTHER than mathematics is to prove you have absolutely no idea how computers actually work. In the real world- every computer is a universal turing machine.
If you have any real doubt - just consider this: any program COULD be written in lisp.
Lisp is DIRECTLY based on lambda-calculus - in fact the ONLY (minor) difference as small syntactical changes designed to make it easier to TYPE lambda on a computer (it was after-all designed for writing in).

That may satisfy the folks who can only recognize math if it is syntactically nearly identical to the math they would do with a pencil and paper. And it's a fine comparison to make, but as I'm sure you know it's not necessary.

For folks who deal with computers at a lower level and aren't confused by divergent syntax, consider a computer ISA.

An ISA is really just a description of a bunch of valid mathematical operators and the operations they perform. There's no reason that the ISA requires hardware that directly implements that ISA, or for that matter a computer of any kind, as long as the math is faithfully performed. That's what makes simulators like Simnow possible, because everything a computer does is just math and can be translated into any equivalent math and you'll get the same result. Even including the IO, since even the behavior of the hard drive, at least as it appears to a program, is simply a mathematical operation.

It's true that currently the most popular languages do not follow the lisp "look like the function you are" structure

I'm not sure I follow you there... the mathematical operations that comprise a function is the function it is, and nothing looks more like that than the underlying assembly. I do get the advantages of functional programming for multithreaded programming, just not sure what you're saying there.

Psst... http://shop.amd.com/US/Pages/ShopHome.aspx . You're welcome. ;)

ATi's numbering has been pretty easy to follow lately.

I agree, but the key word is lately. I despised product names like Radeon X600 (based on Radeon 9600) and Radeon 9200 (based on Radeon 8500).

The first number is the series, this tells you the basics in terms of features, process, and so on. 5 series are DX11 40nm parts, 4 series are DX 10.1 40nm parts, etc.

Still mostly true. My heart broke a little when I read about Mobility Radeon HD 5165/5145, which are DX 10.1 parts.

Meanwhile, AMD competes where it can on the processor front (but ruled the previous 6 months on the performance graphic front).

Ahem -- still rules. http://www.amd.com/us/products/desktop/graphics/ati-radeon-hd-5000/hd-5970/pages/ati-radeon-hd-5970-overview.aspx

(And don't say it cheats because it has two processors, as Nvidia has been doing the same thing for their last two model lines, as well.