AtenRa
Lifer
- Feb 2, 2009
- 14,003
- 3,362
- 136
- Status
Isn't it telling that AMD has to give away $TEXAS worth of games and still the 670 outsold the 7950 and 7970 combined while being on the market for a smaller amount of time?
AMD always tried to offer a better value propositions in GPU, no? The only time they didn't do that, 7000 series, they lost market share big time. So instead of lowering prices, they are offering games for free. Different tactics, same results.
Obviously, AMD marketing isn't getting the job done. Personally, even though I've stuck to NV for a while an want a Maxwell card. But if NV keeps its prices sky high and fails to keep up to date with new versions of OpenCL, my next will likely be from AMD. My current GPU would have been a 7970, but I was working on a cuda project.
AMD gained discrete graphics market share last quarter, they didn't lose market share.
If you bothered to check last quarter numbers, why didn't you bother to check last year numbers?
Why not check since they started Never Settle Reloaded? AMD has gained market share on desktop since they started bundling the newest games.
That's the obvious thing to happen, no? They improve their value proposition of their products, they sell more units.
I just pointed out that they had to correct their initial value assessment of their 7000 series. Nobody wanted to pay Nvidia prices for AMD products, so they first reduced prices in order to stay aligned with Kepler prices, and instead of going further down in the price ladder they started to give games for free.
Just think of Bulldozer. It's a dog, but at the right price, it becomes your dog. And the 7000 series is no dog, it has a nice prospect at the right price.
NTMBK
Lifer
- Nov 14, 2011
- 10,423
- 5,727
- 136
frozentundra123456
Lifer
- Aug 11, 2008
- 10,451
- 642
- 126
The end product is the whole point isnt it?? I still cannot figure how AMD managed to leave so much performance and efficiency on the table, as shown by the improvements made in Vishera, while still bringing the BD to market late. What ifs are really irrelevant in the business world. What counts is the final product brought to market in the same time frame as the competition.
It's that old story of a new architecture on a new process. It doesn't work and Bulldozer was yet more proof of it.
Maybe AMD but certainly not ATi. I didn't see 8500,9700/9800,X800XT PE/X850XT, 1900XT, 1950XTX etc. compete on price. Sometimes it was NV that had to compete on price, at the time when ATI released 1900XT I bought 7800GTO(much cheaper) which was really 7800GTX with lower memory clocks, but the RAM was the same and ram clocked easily to the level of 7800GTX. All of those cards were faster then what NV got at the time. The only card that they had that really had to compete on price was 2900XT, It was quite a big chip but still not as big as G80 but 512-bit and very expensive PCB probably meant it was more expensive to produce then 8800GTX. Also 3800/3800X2 but that was when ATi belonged to AMD.
But only by 4 months and that was due to numerous BD delays. Anyway the point being 32nm was in bad shape at the start, and BD itself was bad (not to mention the wrong part). All of it conspired to make it an awful chip.
You are looking at the timeline from the retail availability perspective - the consumer perspective - which is not why IC designers and manufacturers stagger their development timelines.
They stagger them because of what all has to happen on their side of the fence, within the black box. Which is all the iterative back-and-forth design rule negotiations that go on between designers and the process development team. As well as all the work that goes into validation and respins.
You can look at the back-end of that pipe and say "yeah but it is only four months apart", but you are completely missing what transpired the four years of work leading up to that four month delta between retail releases.
You can convince me that bulldozer should have never seen the light of day in the retail channel, that just like Larrabee it should have been cancelled and bagged, and that AMD should have expedited Piledriver's release timeline instead.
But Piledriver is not the product it is because 32nm was improved or refined, that isn't the case. Piledriver is what it is solely due to what the IC architects did at the drawing board. So you aren't going to convince me that any part of bulldozer's flaws (in comparison to Piledriver) are attributable to it being released 4 months after Llano.
(and I'm not saying 32nm is not problematic, it has its Achilles heel to be sure, am just saying 32nm was never really improved over time and as such bulldozer's release date has no relevance on the node's maturity...but bulldozer's design was not mature, as piledriver shows by being produced on the exact same node)
That's what I did, my old operon 165 lasted 5 years, and could hold a 50+% OC, i never pushed it on my old WC loop, i should have gone for 3 GHZ.
Yeah, I hope they rebound. The more competition between intel and AMD the better it is for consumers. Lower prices and more innovation.
Its a shame they dropped the ball so much after intel's c2d was released. They essentially made the exact same mistakes intel made with netburst.
- Status
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter