AMD's Entry Level Radeon HD 7790 Revealed
AMD provides details on the Bonnaire XT based HD 7790.
As was rumoured last month, AMD will be releasing the entry-level Radeon HD 7790 graphics card to bridge the gap between the HD 7770 and HD 7850. The HD 7790 will feature 786 shader processors instead of the previously indicated 896, a clock rate of 1075 MHz and 2 GB of GDDR5 memory over a 128-bit interface and support DirectX 11.1 and OpenGL 4.3. According to a now redacted result posted on CL benchmark's website, the HD 7790 can be expected to provide 10 percent lower performance than a HD 7850.
Interestingly, the HD 7790 is actually based on the 28 nm Bonnaire XT chip and uses the GCN 2.0 architecture that will also be used in the upcoming Radeon HD 8000 series. For an explanation of AMD's somewhat confusing naming conventions, refer to both our coverage of the HD 8000 series delay and the overview of AMD's plans for 2013.
The Radeon HD 7790 has a recommended price of £118 and will be available in April 2013.

Cheers!
Looking at the Wiki page comparing AMD graphics cards I see that even the 8XXX series is only 28nm.
What gives?
Nvidia is no less guilty of confusing things than AMD. For example, several of their low end 600 series cards have Fermi GPUs (from the GTX 400 and 500 series) and even more confusing, some of them are die-shrunk Fermi.
Nvidia's GTX 650 is literally just a GT 640 with GDDR5 memory and a higher GPU frequency. The GT 650M is a GT 640. the GTX 660M is a GTX 650. There are three different types of OEM GT 640 cards, at least one of which is a Fermi model desptie having the exact same name as the Kepler models and there are similar situations with some of their higher end cards.
Point is that both AMD and Nvidia do these tricks. A famous example is how many times Nvidia reused their G92 GPU. A similar example for AMD could include their Redwood GPU.
First time ever to release the low level cards first.
no, rebadging it when new lineup comes out doesn't make it any better.
Looking at the Wiki page comparing AMD graphics cards I see that even the 8XXX series is only 28nm.
What gives?
That's possible, but even if true, I highly doubt that poor yields are the only reason for not releasing high end cards yet. AMD has little reason to release high-end cards in a time where they'd be difficult to really stress and most gamers who'd consider buying them would probably wait until there's more reason to upgrade. It'd take more sense to release them later when they'll be in greater demand and that gives AMD more time for improvement too.
It most certainly wouldn't be the first time that AMD introduced lower end cards first.
They been doing that for a while, new architecture on lower end first before higher end.
Cheers!
If this is the case then this could explain them pushing back the release date of the 8000 series.
Nvidia is no less guilty of confusing things than AMD. For example, several of their low end 600 series cards have Fermi GPUs (from the GTX 400 and 500 series) and even more confusing, some of them are die-shrunk Fermi.
Nvidia's GTX 650 is literally just a GT 640 with GDDR5 memory and a higher GPU frequency. The GT 650M is a GT 640. the GTX 660M is a GTX 650. There are three different types of OEM GT 640 cards, at least one of which is a Fermi model desptie having the exact same name as the Kepler models and there are similar situations with some of their higher end cards.
Point is that both AMD and Nvidia do these tricks. A famous example is how many times Nvidia reused their G92 GPU. A similar example for AMD could include their Redwood GPU.
I lol'd
Your actually getting an better product and you _itch.
AMD works exactly the same as nvidia, higher number indicates more powerful card in each respective series. So 7000 series and 7980 is the most powerful single chip card. Nvidia is in fact more confusing since they add in "ti" and other suffixes that mean its a step up from that specific model but not as good as the next one up. Aka 650
bigger numbers are better. same as Nvidia. it's just there are more numbers...
anyways, the first number "_XXX" is the number of the series. the second number "X_XX" is the subseries (this generally depicts the GPU inside. for example, the 7970 and 7950 have tahiti GPU's, the 7870 and 7850 are on pitcairn, and the lower cards are based on the cape verde processor. this card, as mentioned, is based on a new GPU, which is on GCN 2.0 - probably an early run to see real-world performance specs and ready drivers for higher end cards) the third number "XX_X" tells you which cards are in what order in the subseries (7750 is below 7770, 7850 is below 7870, 7950 is below 7970, so on so forth) the "9" digit at the end is reserved for a dual-gpu card (6990, third party 7990s)
a few exceptions to the rules I put out above: the 7870 XT should really be a 7930, as per naming rules. (it's better than the 7870, and runs on the tahiti core, but it's below the 7930)
basically, bigger numbers are better. the first number denotes the series, the second number denotes performance range (7 being an entry-level card, 8 being mid-range, 9 being enthusiast or high-end) and the third number tells you the order in that performance range.
make sense?