ShaneBaxtor.com

Recently we made mention about Galaxy working on a Dual GPU GTS 250 Video Card, while we made a slight jest at it due to the fact we figured Galaxy was probably only working on it because they where board it seems the model is very real.

While the card may look like a Dual PCB model now if you look carefully the second PCB is nothing more than a card which holds the three fans that the card uses. It also gives us a fair idea that the card going to take up at least three motherboard slots. When installed you can see the three fans run a blue LED setup and looks fairly nice. See images at the bottom of the page.

As for performance we can see that under 3DMark Vantage it manages to perform a little better than the older single GPU GTX 280. We can see that the card is overclocked to 675MHz on the core and the Shader Clock has been bumped to 1696MHz which helps improve performance overall. Having a look at the Xtreme Tuner HD software which is Galaxys own overclocking program we can see that the GU temp is only 45c, no doubt this is just idle though.

Since the GTS 250 is the highest end GPU that NVIDIA partners are currently working with it comes as no surprise that someone is trying to mix it up a bit in an effort to dump some cores.

We still have no information on any release date or retail pricing but as always we’ll keep you up to date if anything changes.

We’re not sure how popular a dual GPU GTS 250 would be but it’s always nice to see companies thinking outside the box.

It doesn’t matter what you call it, be it GF 100, Fermi or 300 series the new line up of NVIDIA cards are coming, albeit a bit later then we had hoped. What NVIDIA has done though over the past week is added some information on their website about the upcoming series. To have a look at what exactly is going on you can see the page here.

What we’re seeing are two bits of information, you can check out the white papers that are linked towards the bottom which will give you information that you’ll probably not understand or not be interested in. After reading it you’ll end up with still no idea what kind of performance you can expect from the series.

The other piece of information is normal marketing talk to make the new series sound exciting. To be honest if you’re into computers you don’t need this as you’re probably already getting excited about what NVIDIA is going to offer.

The reason behind this post today though is to quickly talk about NVIDIAs new surround gaming feature and the issues I think it’s going to have when compared to the competing technology from ATI, EyeFinity.

I’m only going to give you a quick run down here for two reasons, one I don’t have the card, and two information on the whole thing is still limited. What I will say though is based on the information we have, of course by the time the card launches this could change.

The biggest pro behind the way NVIDIA have implemented the setup is that it seems you don’t need to have a DisplayPort monitor. With few companies really offering the connectivity, read only Dell really! this is pretty cool. It means you should be able to get into the technology for less from the monitor side of things as the chances are you already have a monitor that doesn’t have DisplayPort. Instead of having to buy three new monitors with DisplayPort so they all look the same you can buy just two more of the monitors you have.

The other pro is that the setup will support 3D Vision, only recently I wrote a big piece about 3D Vision and how I felt about it on TweakTown. To put it in just a few words; “I loved it!” For more information about the technology though I would highly recommend you check out the article I did at TweakTown here.

For the most part though this is where the pros stop, the lack of DisplayPort and the fact the NVIDIA cards will only offer two DVI ports means that you will require a SLI setup to make use of the technology. From my understanding it also seems that the technology will only support up to 1920 x 1080 @ 120Hz displays. This is the feeling I get from reading the white paper but really there’s no reason why you shouldn’t be able to run three 2560 x 1600 @ 60Hz displays off two cards.

I’m not sure if NVIDIA are choosing to just push the 1920 x 1080 @ 120Hz display support because these ones will support 3D Vision or because this is the limitation of the technology. I have a real hatred for 1920 x 1080 16:9 computer monitors after using 16:10 ones for a long time. For most people though this won’t be a deal breaker, the chances are from a cost perspective 1680 x 1050 @ 60Hz and 120Hz are going to be the most popular option.

One of the biggest cons for 3D Vision is the fact that when you’re using the technology the video card is essentially working twice as hard at that resolution. What this means is that a 1920 x 1080 screen with 3D Vision on will be more intensive then a 2560 x 1600 display without 3D Vision. Now if you’re running three monitors with 3D Vision at 1920 x 1080 the amount of power needed to run that setup would be more than a setup of three 2560 x 1600 monitors without 3D Vision.

We don’t know performance numbers for the next generation series from NVIDIA but you’re going to need some serious juice to push this kind of setup.

To quickly just recap the pros for NVIDIAs surround gaming experience the lack of DisplayPort monitors needed means that you can buy two more of your current monitors and have a really nice clean looking setup, the other big pro is that you’ll be able to use 3D Vision technology with the setup, from an immersion perspective this is just going to be insane.

The con list includes the fact you’re going to need an SLI setup to run surround gaming, be that with 3D Vision or not. This in turn will mean SLI compatible motherboard and bigger power supply which is going to hit the bank account. Now there’s also the fact that if you go for a 1920 x 1080 @ 120Hz Tri-Screen setup to use 3D Vision the amount of raw graphics power needed is going to be insane and exceed that of three 2560 x 1600 @ 60Hz displays, getting playable framerates isn’t going to be cheap.

One last con would be the fact that we’ve got this feeling that the maximum resolution for the technology is 1920 x 1080 @ 120Hz. What this means is that 1920 x 1080 @ 60Hz, 1680 x 1060 @ 120Hz / 60Hz and lower should be supported but not 1920 x 1200 @ 60Hz and above. I write this separate to the last paragraph because I think as we get more information on the technology this won’t hold true but at the moment it doesn’t change the fact that this is how it feels reading the white paper.

For so many people this isn’t going to be an issue as multi monitor display setups are still a rarity; it’s clear the reason NVIDIA has done it though is due to the simple fact ATI offer EyeFinity, unfortunately the solution from NVIDIA feels like a bit of a bandaid one, it’s simply done because the competition offers it, and since NVIDIA didn’t expect it and it was too late to go back to the Fermi drawing board the technology is how it is because it was added last minute.

The reason I’m speaking about it so much thought is that I personally use a 3x 1920 x 1200 setup via EyeFinity. Buying new monitors wasn’t an issue for me as I was moving from a single 30″ and I didn’t want to buy two more 30″ monitors due to the huge amount of desk space I would need.

At the end of it all though for most people the most important feature Fermi will offer is performance, if they’re able to beat ATI, offer stock and offer us a reasonable price it will be a huge win. For this reason I can’t wait to attach a Fermi or three to my 2560 x 1600 testbed monitor and let it tear up some of the latest games available.

You have to think what many of the NVIDIA only partners are doing at the moment, speaking to a few as of late it seems that the highest end product being sold is the GTS 250 with most other models like the popular GTX 275, GTX 285 and dual GPU GTX 295 all becoming discontinued.

With all that free time on hand though it seems like some companies are choosing to experiment and have a bit of fun. The latest wacky incarnation of a NVIDIA graphics card seems to be coming from Galaxy and will carry with it two GTS 250 cores.

Information we know of at the moment is, two G92-426 cores will be placed on a single PCB with the help of the BR04 bridge chip. The particular 426 version of the G92 core are low power versions meaning that this card should use less power than that of two non 426 based GTS 250 cards.

2GB of memory is present thanks to 16 .8NS DRAM chips with both running on a 256-bit memory interface. This particular model will have 6-Phase power and Bussmann high-frequency inductance which will boost the frequency up to 2MHz. What that bit means exactly we’re not sure but the general word is that due to the low power nature of the GPUs and the high quality power supply components overclocking should be impressive.

As for clocks or a release date information is extremely limited at the moment, it’s possible that boredom could’ve just set in at Galaxy while waiting for Fermi and the company is just killing some time. If we get any more information though we’ll be sure to keep you in the loop.

With ATI set to launch the new Radeon HD 5500 and 5600 series at CES which starts on the 7th of January and the HD 5400 series later that month, the company is about to flood the market with a huge amount of low-end cards.

Word has come through that in the HD 5600 series alone ATIs partners will have the option to use as many of the six variances they want. Why we don’t know 100% what the different variances include we’ve been told it’s a combination of connectivity and memory. It’s probably fairly safe to assume that we’ll end up with an Eyefinity version along with 512MB and 1GB versions. Within the series we’ll no doubt see two model names which should consist of the HD 5650 and HD 5670 simply by going off past trends. 

As for the HD 5500 and HD 5400, we have been told that ATI will offer between two and three variances which means a total of six variances in those bottom two series and another six variances in the HD 5600 series bringing us to a grant total of 12 new video cards.

While I had planned to get stuck back into reviewing early January it has becoming extremely clear that the option to be a bit slack is quickly fading away.