Slashdot Mirror
Overclocked Radeon Card Breaks 1 GHz
dacaldar writes "According to Yahoo Finance, noted Finnish over-clockers Sampsa Kurri and Ville Suvanto have made world history by over-clocking a graphics processor to engine clock levels above 1 GHz. The record was set on the recently-announced Radeon® X1800 XT graphics processor from ATI Technologies Inc."
What a day for world history! It will be remembered forever!
why this announcement would come out on Yahoo! Finance
No, seriously, I just come here for the articles.
I didn't have Slashdot in a full screen window, so the headline read:
:p
Overclocked Radeon Card Breaks
1 GHz
Was wondering why an overclocked card breaking is such a big deal
ClutterMe.com - easiest site creation on the Net. Just click and type.
Without the pretty graphs how will I know what's going on?!
><));>
I always knew ati would finnish first.
Coding projects blog - Code Slim
have made world history
I think that's going a bit far. Good for them and everything, but world history? V-E day, Einstein's 1905, Rosa Parks refusing to give up her seat on the bus- these events impact world history (sorry for the all-Western examples); making a chip oscillate faster than an arbitrary threshold does not.
Simple Unexpected Concrete Credible Emotional Stories
I think we've found the source of global warming.
Finance tutorials and more! Understandfinance
where were you when the first video card was overclocked to 1GHz. And most people will respond "huh?".
Seriously, "world history"? There's no historical significance here. It was inevitable, and no big deal.
NVidia will make a competitve model, with blackjack, and hookers.
And as for being the first people in the world to do this... the chances of that are small. I'm sure there are people at Radeon (and other companies) who have done things far more bizarre, but didn't announce it to the world.
The team, optimistic that higher speeds could ultimately be achieved with the Radeon X1800 XT, attained the record speeds using a custom-built liquid nitrogen cooling system that cooled the graphics processor to minus-80 degrees Celsius.
It seems we may have a ways to go before it can be done with standard air cooling. I actually didn't think that operating temperatures for these processors went down to -80C.
Things you think are in the Constitution, but are not.
If I could only go back in time and add liquid nitrogen to my 8088 processor. I know I could have gotten it up to 5.33 mhz, no problem. NetHack benchmarks would have been off the chart.
Sampsa Kurri told in a Finnish forum that it was over 1 GHz only in 2D mode. They are trying to run it with same clocks later. ATI left some tiny details away from their press release... ;P
What I've been waiting for is some sort of mathematics program (I used to use Mathematica in college) that could utilize this concentrated power, rather than hampering the CPU.
Does anyone know if any researchers have gone as far as to utilize a GPU for anything like this?
This sig isn't original enough, it's time to come up with something witty...
Sampsa and Ville already broke their own record by overclocking the same setup to over 1GHz for both the GPU and memory. See pictures over at Muropaketti.
Cuz you know it's like way better to play Quake IV at 953 Frames Per Second. Totally!
That's just sad... that video card now has more clockspeed and more memory than my own main computer.
http://www.gpgpu.org/
LL
Go here for several examples of this -- far from simply having been proposed, it's been done a fair number of times.
The thing to keep in mind with this is that while the GPU has a lot of bandwidth and throughput, most of that is due to a high degree of parallelism. Obviously 1 GHz hasn't been a major milestone for CPUs for quite a while, but CPUs are only recently starting to do multi-core processing, while GPUs have been doing fairly seriously parallel processing for quite a while.
Along with that, the GPU has a major advantage for some tasks in having hardware support for some relatively complex operations that require a fair amount of programming on the CPU (e.g. multiplying, inverting, etc., small vectors, typically has a single instruction to find Euclidean distance between two 3D points, etc.)
That means the GPU can be quite a bit faster for some things, but it's a long ways from a panacea -- you can get spectacular results applying a single mathematical transformation to a large matrix, but if you have a process that's mostly serial in nature, it'll probably be substantially slower than on the CPU.
Along with that, development for the GPU is generally somewhat difficult compared to development on the CPU. Writing the code itself isn't too bad, as there are decent IDEs (e.g ATI's RenderMonkey) but you're working in a strange (though somewhat C-like) language. Much worse is essentially a complete lack of debugging support. Along with that, you have to take the target GPU into account in the code (to some extent). I just got a call in the middle of a meeting this morning from one of my co-workers, pointing out that some of my code works perfectly on my own machine, but not at all on any his. I haven't had a chance to figure out what's wrong yet, but I'm betting it stems from the difference in graphics controllers (my machine has an nVidia board but his has Intel "Extreme" (ly slow) graphics).
--
The universe is a figment of its own imagination.
The universe is a figment of its own imagination.
Since DirectX 8 (I think), the color values have been floating point numbers, this is to avoid loosing a lot of possible values through all blending with multi-texturing and effects (fog, lighting ect) which are of course much slower than very simple integer calculations. Even on the Athlon64's FP add and muls are 4 cycles, you'd have to make the top end A64 about 700mhz if you make them single cycle execution. (multi-cycle instructions aren't as bad a thing on the CPU as there are plenty of other things to do while you wait, not so in GPUs).
GPUs have also tended to focus on parallel execution - at least over the last few years - increasing the number of pixels done at the same time, to compensate for not being able to hit multi-ghz speeds, so yes they have many more transistors than typical CPUs (the 7800GTX might break 300 million, well over 250 million) - and of course heat is an issue if you push the voltage and / or clock speeds to far. The last few generations of GPUs have been up around 65-80W real world draw, more than most CPUs out there. And of course GPUs have very little room for cooling in those expansion slots.
...the pictures of the rig : here they are, 3DMark05 included.
http://www.bfgtech.com/7800GTX_256_WC.html
BFG GeForce(TM) 7800 GTX OC(TM) with Water Block. Factory overclocked to 490MHz / 1300MHz (vs. 400MHz / 1000MHz standard), this built-to-order card will feature a water block instead of a GPU fan for those wanting to purchase or who may already have an existing liquid-cooled PC system. BFG will hand-build your card using Arctic Silver 5 Premium Thermal Compound. Easily hooked up to any existing 1/4" tubing system or to 3/8" tubes with the included adapters, this card runs cool and silent. BFG Tech is proud to offer their true lifetime warranty on this graphics card. (Card with water block requires internal or external water cooled system, sold separately.)
Pain lasts, kid. Its how you know you're alive. Sometimes I think this growing up thing is just pain management-TheMaxx
It'll just about be able to handle Windows Vista... :-)
PocketGamer.org - For the gamer on the go!
Well, while the CPU people are finally doing dual core processors (essentially, two instruction pipelines in one die, plus cache et al), the GPU people have something like 24 pipelines in a single graphics chip. Why is it that the CPU people have such lame parallelism?
To answer both questions. Graphics are trivial to parallelize. You know to start with that you'll be doing essentially the same code for all pixels, and each pixel is essentially independent from its neighbours. So doing one or twenty at the same time is mostly the same, and since all you need is to make sure the whole screen is rendered, each pipeline just needs to grab the next unhandled pixel. No syncronization difficulties, no nothing. Since pixel pipelines don't stop each other doing syncing, you effectively have a 24 GHz processor in this beast.
On the other hand, you have an Athlon 64 X2 4800+ (damn, that's a needlessly big, numbery name). It has two cores, each running at 2.4 GHz (2.4 * 2 = 4.8, hence the name, I believe). However, for safe use of two processors for general computing purposes, lots of timing trouble has to be handled. Even if you do have those two processors, a lot of time has to be spent making sure they're coherent, and the effective performance is well below twice that of a single processor at twice the clock speed.
So, if raising the speed is easier than adding another core, and gives enough performance benefits to justify it, without the added programming complexity and errors (there was at least one privilege elevation exploit in linux that involved race conditions in kernel calls, IIRC), why go multiple processor earlier than needed? Of course, for some easily parallelized problems, people have been using multiprocessing for quite a while, and actually doing two things at the same time is also a possibility, but not quite as directly useful as in the graphics card scenario.