When it comes to new gaming products in recent times, NVIDIA is way behind compared to the competitors. In order to resolve that problem finally, NVIDIA held a premier of the GTX 700 series. Aside from focusing the attention on itself in the moment when AMD had interesting developments in terms of APUs, NVIDIA did itself another favor. Namely, considering that GTX 700 doesn’t have a new architecture and doesn’t use a new production process, what we actually got was brand restructuring. Also, improvements do exist in certain areas of the products, so it’s not a case of a simple “rebrand” strategy. In any case, we had the opportunity to test GeForce GTX 780 with which the brand was started, in descending order. Next month, we expect GTX 770, and how the situation will continue to develop, remains to be seen. We assume that NVIDIA will mostly focus on brand reorganization (they’ve become quite the experts in that area). Let’s go back to the GTX 780.

The latest GPU that AMD made was in March of last year. Since then, Southern Islands series consisted of three models, off of which a great many number of Radeon models were based. Still, some changes had to be made after the competition introduced GeForce GTX 650 Ti. Aside from that, Radeon HD 7850 with 1 GB of memory had to be sent in early retirement because of stopped production of GDDR5 chips with 1 GB/256 bits configuration. AMD’s plans were momentarily disrupted, because a great gap was made between Radeon HD 7770 and HD 7850. The gap always existed, because Cape Verde and Pitcairn were too different. Lucky for AMD, that wasn’t as noticeable up until the moment the competition introduced a model that could fill that gap and that was GeForce GTX 650 Ti. The answer wasn’t simple and that’s why we had to wait for a while. Namely, the weakened Pitcairn wasn’t an adequate solution, because it was made for a 256-bit bus, and it probably wouldn’t have been enough of them to satisfy the great demand of the middle class. Not to mention that the price of production couldn’t be lowered enough in order for it to be tempting to customers. On the other hand, Cape Verde reached its peak on HD 7770, and as such couldn’t be further altered.

Sapphire, as AMD’s premium partner, always gets specific models with a particular capability or purpose in mind. The Radeon HD 7950 and HD 7870 cards suffixed FleX Edition, which we’re presenting today, are just such cards. So, what’s it about? While it’s true that we’ve written about FleX cards on multiple occasions, we still think that a reminder is in order. Eyefinity is a well-developed and very useful technology, but with an important limitation in its basic shape. In order for you to connect three monitors into a single large surface (SLS), all three of them need to have digital input, with at least one connected via DisplayPort. Obviously, a casual glance at most markets, especially smaller ones, reveals that many users only have analogue monitors, even in 1080p models. Besides, DisplayPort is still pretty rare and is mostly found on upper-class models such as Dell’s UltraSharp series. This means that Eyefinity is practically unviable for many users, at least without the strenuous “sell old, buy new” activity. However, owing to Sapphire, this problem has been solved in Radeons with the FleX in the name suffix.

As AMD’s premium partner, Sapphire has a certain reputation that it’s been maintaining for very long. Although they rarely venture out to the market with specials models such as ASUS’ Matrix, Ares or MSI’s Lightning, they still always manage to match their competitors. This they accomplish through constant quality, which has admittedly been questioned after the information about them having certain issues with Radeon HD 7870 leaked to the internet. Still, even such rare mishaps aside, Sapphire’s graphics cards have never been on top, but always very close by. Their advantage is based on frequent reality checks, unlike ASUS, MSI and Gigabyte, who tend to wander off into a higher price range in their desire to be the best. Sapphire sticks to the mortals and offers excellent models with an alternative design, with a far more interesting price tag. One direct consequence thereof has been the Dual-X model, always among the cheapest on offer, but with great cooling and excellent performance nevertheless. After the said model from the latest series, it was only logical to expect a card with a Vapor-X cooler. If you aren’t familiar with it, Vapor-X is the best Sapphire has to offer in terms of cooling, since it’s their in-house solution. These coolers rely on a large vapour chamber which is essentially one large heatpipe, lending its name to the entire series. Of course, this cooler has also evolved from one generation to the next, and one could say that this card sees it at its largest and most impressive.

AMD seems to have launched an unofficial competition for which company is going to modify and improve the already brilliant Tahiti chip the best. Well, it seems that ASUS has taken up the challenge with fervour, turning all standards up a notch. Frankly, we believe that other brands will prefer to wait for the new generation for a retry rather than try to compete with the new Matrix, and here’s why.

We all know that NVIDIA has gained a rather negative reputation for its rebranding strategy of selling old products under new names. This has impacted our own views of some of their products in the past, especially in the mobile segment, where a single misinformation in the specs can leave you with a Fermi-based model despite the GT(X) 600 GPU inside. Worst of all, this trend reached the peak of absurdity with NVIDIA announcing that they have decided to base their GeForce GTX 670M on the Fermi chip. However, things have changed with the release of the latest mobile GPU, the flagship GeForce GTX 680M. Luckily for NVIDIA, the surprise wasn’t a nasty one this time around, despite their recent policies. In the past few generations, it’s been something of an unwritten rule that the flagship mobile GPU is always going to be based on the best the company is offering in the desktop mid-range.

NVIDIA's desire to cover as much ground on the market as possible with a relatively low number of chips has resulted in the split of a single chip between desktop and notebook hardware. The first GPU based on the Kepler architecture which we had the chance to test was, indeed, GK107. More interestingly, this chip was housed in Acer’s TimelineU ultrabook, labelled GeForce GTX 640M. Afterwards, GK107 appeared on a few other mobile graphics cards, the most interesting of which is GT 660M. Since then, NVIDIA has transferred exactly the same GPU in discreet graphics cards onto the desktop market.

 If you’re following our website on a regular basis, you must’ve noticed how much we’ve been whining about the fact that NVIDIA has been recycling one and the same GPU under a different name, trying to present it as an entirely new product every time. Essentially, this move isn’t all that bad, since it gave us some really good cards, such as GTX 660 Ti. On the other hand, you’ve got to start wondering if everything’s OK with a company that serves the same product on and on, while other market segments remain starving. We managed to come to terms with the fact that both AMD and NVIDIA had all but given up on the low-end, mostly due to the excellent integrated GPU solutions present in the CPUs of today. However, we couldn’t do the same with the fact that NVIDIA just didn’t have a decent representative in mid-range.

 GeForce GT 640 and GT 630 were launched to little pomp, and NVIDIA didn’t seem to have much faith in these two cards anyway, since no test samples were distributed to the press, and it was difficult to find them in anyone’s stock in general. After a longish wait, we’ve finally got a smaller chip which reminds, at least on paper, of an actual mid-range GPU. Its name is GK106, and when you’re this late to the market, expectations are bound to be high. In order to impress anyone with a product, with the competitors preying on the market niche already, one really has to offer significantly more, or at least the same at a lower price point. Besides, you won’t get sympathy from either media or target market if you bring nothing new to the table. All things considered, GK106 was poised to a long way to success, and we were more than interested to find out how it would cope with such a burden.

The rumours have come true, so just before the “back to school” season, we’re being greeted by the new GeForce, poised to find its place under the sun in the very attractive 250€ segment. Yes, the model in question is GTX 660 Ti, and it’s now finally clear that NVIDIA has decided to bring back the Ti(tanium) suffix for good. We’re glad to see this move, as this suffix has always been synonymous with quality and performance. The speculations surrounding GTX 660 Ti were reaching boiling point, with the latest “real” specs changing from day to day, from the idea that the chip on the card is entirely new, to the more realistic standing that the card will essentially be an impoverished GTX 670. All this journalist hype goes to prove just how much the IT public is interested in this model, especially since NVIDIA has taken so long to start descending into the masses with their latest generation. It’s clear that certain problems with the 28 nm lithography are present, but this doesn’t change the fact that NVIDIA is well behind its competitor, which has already started refreshing the existing latest-gen models, while NVIDIA is merely presenting the first ones, and the completion of the gamma is still not in sight. Yet all this fades into the background now that we finally have this card at hand, especially for those whose pockets aren’t too deep, but still want to have a Kepler GPU inside their enclosure.

NVIDIA has done a great job with Kepler, no doubts about that. The fact that a relatively small chip which doesn’t require a lot of energy contains blazing performance was about as good as it gets, as basis for a dual-GPU solution. Furthermore, this American company has a bit of a hard time in the segment under GTX 670 and GTX 680, so it was quite expected of them to present a graphics card with two GK104 chips and fill the gap while waiting for derivatives to appear. According to gossip, the problems are mostly related to the low yield of graphics chips made in the 28 nm production process. Unfortunately for NVIDIA, lithography problems have caused additional problems in the lacking mid-range and entry-level gamma, a market segment traditionally troublesome for the company. Either way, although things have been rough for NVIDIA lately, they still know they have an ace up the sleeve, and they’re doing their best to properly use it for attracting media and consumer attention.