As to TitanX, like every other highest end component, you pay a premium for it, regardless. I should know, I just bought a 5960x!
Congrats!! As crazy as it sounds, I can actually understand the guy who buy a 5960X.If you need many cores for specific workloads (rendering, encoding, etc.), then you actually save time with this processor. Time is A LOT more valuable than FPS because with more free time you can do anything you want - that means you are paying not only for performance but time savings which brings more flexibility in your daily life.
Secondly, a processor such as 5960X will probably last 5+ years at 4.4Ghz for gaming and it will still command a solid resale value even in 2-3 years from now. As a result, the overall cost of ownership of 5960X will likely be much lower than a $1000 graphics card unless one perfectly times the resale of a $1000 card right before the next best card launches. With a $1000-1500 GPU like Titan X / R9 295X2 at launch, you aren't going to feel like you have top-of-the-line or near top-of-the-line GPU for 2+ years, but with a 5960X you will. In a matter of 1 year, there is a card as fast for nearly half the price, and in 2 years a $700 card will be 40-50% faster than the Titan X. Since the day 5960X launched, there was no faster CPU overall under $1k.
R9 390/390X will have very high power draw and TDP and there will be countless comparisons against GTX Titan X and GTX 980 Ti where AMD may not gain much if the cost HBM drives the price up toward GTX 980 Ti price. The efficiency vs Maxwell will be what will be dragging them down, and if performance is the same as 980 Ti or worse, with the price similar, AMD will have huge problems regaining any market share.
Titan X already uses nearly 250W of power on its own. At this point who cares if 1 card uses 250W of power and another 290-300W. It'll come down to features, noise levels, availability of AIB offerings, overclocking and price/performance. When an informed gamer is buying a card in the $600+ range, 50W of extra power usage is meaningless in a system that already uses 390-400W power.
R9 360, 370 and 380 (+X) will all be rebrands, renamed to fit the "Pirate Islands" theme, to try to trick people in to buying rebrands. With the current news and the latest roadmaps and the driver leak, I am not very hopeful for AMD.
You keep repeating this over and over. Ok, tell us what benefit would it give AMD to take an R9 290 that today sells for $250 and R9 290X that today sells for $290-300 and re-release the
exact same chip for $249 and $299?
Do you ever think about this point?
MSI Gaming 290 =
$250
HIS IceQ 290 =
$250
PowerColor R9 290 =
$250
XFX R9 290X =
$290
Unless you think AMD will sell R9 380 for $199 and R9 380X for $249, what exactly would they gain by re-branding R9 290/290X as R9 380/380X?
Also, as others already pointed out, Pitcairn does not even support FreeSync and Huddy officially stated that all GPUs moving forward will support FreeSync. He also said that FreeSync support can't just be added on a PCB/via a display controller as it requires a redesign of the ASIC itself. So what you are saying is AMD will just redesign Pitcairn to fit a new display controller and UVD to support FreeSync and H.264 hardware decoding but make 0 other changes to the chip in terms of clock speeds due to improved 28nm node, and any other components such as geometry performance, colour compression, etc.?
I just bought a 295x2 because it was so cheap and wanted to check out the AIO. Definitely not going with the WCE. Runs cool and quiet but raises the temps on all my other components due to exhaust restrictions. Losing that 140mm rear fan has a huge deficit under idle conditions.
Why not set the GPU exhaust at the front of the case or the top of the case and leave the rear for your CPU heatsink tower?
Titan X is 'OK' but I don't feel the $1k price is justified. The 4xx series really looks to be bringing some good stuff for AMD, so maybe I will just wait it out until we see Pascal vs. 4xx.
That's a good point. If you have a card like 970 OC or similar, you can probably coast to next year's 14nm/16nm GPUs unless you absolutely need 60 fps in every title with ultra quality.
Who in their right minds would call GTX Titan cards a failure? The only failure was AMDs inability to compete with the cards when they launch. Hence the price of the cards.
The price of these cards is high, I totally agree that $999 is very high for any GPU, but you are totally out of your mind if you think they bunked.
Sales do not always indicate that the product is actually good. There are plenty of products that sell well but aren't worth the money. The point is 780Ti doubled 580's performance for $700 in about 3 years. Titan X doubles HD7970Ghz performance in a similar 3 year period but it costs $1000. If it was $700, it would be a lot more acceptable/reasonable. With your logic, should NV raise the next Titan's price to $1200-1300 because they can? They certainly could because Titan X seems to be selling even better than the original Titan based on what Computerbase reported.
The Titan X GPU itself is very good but the price is out to lunch. In a span of 4.5 years NV convinced some gamers that a flagship shouldn't cost $500-650 (280/480/580) but $1000. If you think that's perfectly acceptable, that's fine. There are plenty of PC gamers who find the Titan X's price acceptable and I am sure even if it was $1300-1500, they would also buy it. A lot of us view it as NV arrogance and a marketing money grab because NV keeps removing features like DP but keeping prices the same, not to mention Titan X's performance advantage over 980 is less than the original Titan had over the 680. All that bashing the 7970 received, the Titan X gets almost none of it on our forum.
AMD once again is absolutely horrible at marketing, they seem to live in their little la-la land. What they should have done the minute GTX Titan X launched was to release a statement to the press about an upcoming card(s) and/or have many reviewer sites do a preview of R9 390X.
No, absolutely not. That's called the Osborne Effect in business. Remember Nokia announcing that they are ditching Symbian in favour of Windows OS but it was still a long time away until they would even release the first Windows phone? The minute they did this, sales of all Symbian phones tanked as consumers knew Nokia would not support them in the future. The result was excess inventory of Symbian phones that had to be sold at highly reduced prices, which resulted in very little profits and even losses. Similar mistakes have been made by many firms such Sega when they killed Sega Saturn just 3 years in the US by prematurely announcing the Dreamcast, or Nintendo this month when they already announced the Nintendo NX for 2016! Goodbye Wii U sales.
These mistakes are made over and over in business and that's why they are taught at the top business schools in the world as case studies later on.
AMD reduced its inventory in the channel to clear all R9 200 series of cards as was reported by many sites. It's clear to anyone who follows the news that most of NV's market share gains came NOT from them taking away AMD's sales but from AMD simply stopping to ship new products in the channels for the last 2 quarters. Whether or not you agree with AMD's inventory reduction strategy is a point of discussion but it's obvious AMD needs to sell all of those old R9 200 cards before launching the new series.
If we look at NV's GPU sales, they are still at about 9-10 million, which is no change at all based on historical data or last 8 quarters (except that dip to 7). Had NV gained AMD's sales, their sales would have jumped to 12-13 million. That's not the case. It's obvious AMD is not shipping product to OEMs/channel, which means AMD bleeding market share is their own decision as a consequence to clear inventory. Another way to think about it is the overall Discrete GPU market has fallen from 14-17 million to just 11.5-12.5 million! If it was NV gaining market share at the expense of AMD's sales, the overall discrete GPU market would still have hovered at 14-17 million. It isn't.
Also, the Titan X is $1000 so for the majority of the market it's a card that might as well not even exist. That's why AMD doesn't care to release info on its cards prematurely because it knows that the $400-500 GPU segment will probably sell 10-20X more. However, releasing information risks cannibalizing sales of existing 200 series.
As to your comment that AMD is horrible at marketing and they should have released slides about R9 300 series, chances are the same people who are easily swayed by marketing for their GPU purchase, are the types who would buy NV anyway, or they are the type who don't follow these tech sites closely. The rest of us have been reading about specs of R9 390X and even have seen AMD's leaked slides a long time ago. The average joe, well he wouldn't even know what HBM is and is happily buying a GTX960 2GB for $200 or GTX960 4GB for $240, while being proud of its awesome perf/watt.
