ShaneBaxtor.com

KitGuru broke a story only recently that the GTX 470 could become discontinued; for the moment it seemed that the model which has become a bit of a favourite of mine is becoming hard to get a hold off.

Throwing on the detective cap it was time to see what’s going on with this story and if it holds any truth. Chasing up our own contacts it seems that NVIDIA are making way for a dual GPU based Fermi card as KitGuru later updated in their post. It seems this isn’t coming at the cost of the GTX 470 but at the cost of GTX 470 supply.

What it seems is that these will be going towards the new dual core model coming out; while no official name has been said the mumbles are it’ll be called the GTX 490. At the moment though we find ourselves a little disappointed.

It seems that the GTX 470 PCB has gone over to the GTX 465 while the cores go over to a new PCB which will create the GTX 490. Unlike ATI though who opted for the best core which was used on the HD 5870 and then clocked the card down to HD 5850 speeds it seems that NVIDIA are going to opt for the second best core (GTX 470) and we’re sure they’ll also clock the model slightly down which means it’ll be slower than a GTX 470 SLI setup.

Of course you throw overclocking into the mix but the same can be said about the HD 5970; at least that was based on the higher spec chip when it came to stream processors and other goodies.

With all that said it doesn’t seem like the GTX 470 is going anywhere; not in the short term anyway. It seems that NVIDIA probably just want to move the cores temporarily over to its new high end model; as yield increases we’ll no doubt see the GTX 470 back in force.

I try to be careful when I organise my video card samples, it’s hopeless getting in a bunch of cards that carry with it the same clocks and all that changes is the cooler. For the most part why should I waste my time benching a card that is going to give the same results, waste your time with results that are going to be similar and TweakTowns money with a review that’s not going to bring traffic.

Now there are some exceptions to this rule, if that model is part of a higher profile series like IceQ and iCooler from HIS or Vapor-X from Sapphire; due to the nature of these cards they get happily tested across all games even though they may carry with it the same or similar results in FPS related tests. People will Google these particular variations of models though and they shouldn’t have to look at a standalone review on a reference card just to know the performance.

On the other hand if Sapphire and HIS send two cards in at launch, slightly different coolers but both don’t slot into one of the aforementioned categories the excitement level isn’t high for card number two. So we do two reviews with these cards, the first is a standard single card review with card one. Who is card one? Well it’s generally the person who arrives first, in the case they both arrive at the same time it’s the person that told me they were sending it first.

For card number two though since it’s reference clocks and follows the reference PCB design to mix it up a bit I’ll make that second article a CrossFire or SLI one. This does a few things, one it makes the review different from the first, second we still get cooling numbers and noise levels off the card and finally it’s something a bit more interesting for you to read.

Recently though I had a company complain to me that the review looked like it was a CrossFire one and not on their card, to be honest I must’ve missed the part where I wrote about their package, card, cooler, included their names in all the graphs, wrote about the temperature and heat difference and wrapped it all up in a conclusion which covered both the technology and the card.

This company will now have a few options, the first is, unless they can be the first company to send a card before NDA they won’t get a standalone review, secondly they can just not send a card that has reference clocks if they’re not interested in being tested in a different fashion, instead they can just wait till OC models come out. Thirdly they can step back a second and realize I know what I’m doing after seven years when making sure that I give each company as much exposure as possible.

If the company doesn’t want to be included in a CF article many other companies will, after I have two reference cards I ultimately stop trying to organise any more until OC ones are available. Sure they can send an OC model which will get a standalone review but that company now misses out on a review on TweakTown, instead of getting two they now only have one. Since I make sure I don’t organise crap that also means the company will more then likely miss out on an award, that’s fine for me someone else can pick it up instead.

I’ve been doing TweakTown for over 7 years and this writing gig for even longer. These companies need to know that how I represent their product in a review is the best way for them. CrossFire and SLI articles are a way to mix it up a bit, it means I learn more about the performance and you do as well. It means that if someone on a forum goes, I’m thinking about having CF or SLI *insert mid range model* that forum goers can link to the article and say this is what you can expect.

After reading this I hope the company realizes that I do what I feel is the best for them. If they don’t though there’s someone else who is always happy to take the spot and get the exposure. As for other readers of my blog, this just gives you a bit of an idea of what we deal with behind the scenes, it’s one thing whinging because they don’t like the score, but to get an award, a high score and then complain that they don’t like how it’s tested; come on!