Five days ago, AMD launched a newbie expected to aim performance-price niche between the fastest of R7 series and the cheapest of R9 series where exactly its two-year old 7850 lies in, namely R7 265 based on Curacao Pro GPU. By introducing R7 265 AMD intends both to expulse oldies out of its product portfolio and to give an impressive first encounter to another newbie of its very own competitor, NVIDIA, namely GTX 750 Ti at the same time. R7 265 itself is, however, nothing new but its nomenclature : the first product employing '-5' numbering in Radeon R-series family. Basically it's identical to 7850 in its core config, 1024 GCN-based SPs / 64 TMUs / 32 ROPs out of full 1280 / 80 / 32 Pitcairn GPU though clock speed is increased in both GPU and memory, by 7.5% and 16.7% respectively, as well as TDP is increased from 130W to 150W.


As new leader of the order of R7, its main virtues are not only to win the previous leader R7 260X in terms of absolute frame-per-second but also to maintain good reputation in terms of performance-per-price. To achieve these two goals concurrently, AMD put a de-facto 7850 with increased clock speed with $150 price tag. To tell the truth, I already was seeing a valid conclusion (plus bored) on R7 265 even before starting any test : "Lies somewhere between 7850 and R9 270(7870), cheaper than 7850, thus better perf/price ration than 7850". But I'm a responsible man who knows what he has to do. Hereby I dedicate this article to my vocation.


Test system I used for benchmark consists of follow components:





Speaking of methodology, I ran all games 3 times repeatedly for each resolution under same routines as similar as I could play, and choose the median value as the final result. Graphics quality of each game is the highest possible in-game settings except few minorities : for example, anti-aliasing and PhysX. Further details regarding this will be noted in each game's chapter, respectively.


Before we go for tests, I want to give you some theoretical background. Let's see this table.





Forget about the codenames : Pitcairn and Curacao refer exactly a same silicon comprises of 2800 million transistors materializing GCN architecture and occupies 212 mm2 (square milimeter) of area. A full-blown Pitcairn/Curacao has a suffix "-XT" while a minor version gets "-Pro" with 20% fewer SP / TMU counts. As a filler-of-gap between 7850 and R9 270, R7 265 employs clock profiles of R9 270 and hardware specifications of 7850 at the same time. Anyone wondering how a mixture of R9 270 and 7850 performs against R9 270 and 7850?


(※ Detailed results are skipped for image # limit. Full review is on my personal blog : http://iyd.kr/625)


Here are performance summary graphs:









R7 265 wins GTX 750 Ti (indeed, GTX 650 Ti Boost) with solid gap over 10%. In the comparison group, only R9 270 and GTX 660 dominate R7 265 by less than 10% and obviously less budget-friendly. We'd better to play a funeral for 7850 in memory of its two-year-long life. (literally, two-year is quite long in this industry after all) At least AMD aims the right position whether it will sustain competitiveness in that niche or not.


So far we have covered how fast (or slow) R7 265 is in compared with its neighborhoods. I already was sure that I know everything about this card even before starting any test, and now I'm sure that I was f**king right. Basically, hierarchy of frame-per-second throughputs just follow each graphics card's numbering : 270 is faster than 265, 265 is faster than 260X. Nothing special.


Speaking of performance-per-price criteria, however, price tag of $150 is quite attractive for this niche of market. Despite AMD first put its R9 270 in $180 and R7 260X in $140, currently they are being sold at prices pretty higher than first introduced because of basic economy : demand-and-supply. Assume that R7 265 will be tangible at that price, there is no doubt that it will be new perf/price king. Unfortunately, however, it seems the contrary is more likely : there is no rationality to believe that AMD suddenly becomes capable of manufacture tons of R7 265 to meet the demand curve yet incapable for 7850(and indeed, any of other R series family). For short, the key to stabilize the market is manufacturing, not declaring an old product as new brand with brand-new box and sticker.


In spite of critique above, it seems reasonable that AMD fill the gap between R9 270 and R7 260X. I think it will compete GTX 750 Ti well in the same price level with better performance unless AMD fails again to achieve the demand curve. ("It's economy, stupid!")


Well... the article is over.
Thanks for reading. Have a nice day!