Navi discussion thread

So what you are saying is that AMD is missing a competitior to the GeForce GTX 1660 Ti.
That is correct, maybe even 1660 as that directly compete with RX 590, there is a big gap between RX 590 and the impending release of 5700. You can pick up a Vega 56 for under $300, but those cards are being replace by 5700 and I don't see AMD keep on producing Vega series due to high component cost to compete the GTX 1660Ti.
 
Which works quite well, is in games now, and is implemented in hardware now. This is the kind of think that AMD is known for not pulling off, thus the basis for caution.
It looks and works worse than upres in every game apart from a single easily optimized tunnel shooter half a year later. Ai shits the bed in open world.
BIQWWmlCQAEN7hO.jpg small.jpg
 
Last edited:
That is correct, maybe even 1660 as that directly compete with RX 590, there is a big gap between RX 590 and the impending release of 5700. You can pick up a Vega 56 for under $300, but those cards are being replace by 5700 and I don't see AMD keep on producing Vega series due to high component cost to compete the GTX 1660Ti.

Well, the Radeon RX 590 is the competitor to the GeForce GTX 1660.

AMD just needs a competitor to the GeForce GTX 1660 Ti.
 
They're way behind in performance and specs. Years behind.

Sadly true. And the lack of real competitiveness from AMD is keeping GPU prices in the stratosphere. With poor Nvidia pricing, and poor AMD performance, I'll have this 1080Ti for a looooong time, it seems.
 
Ref cards are usually better quality than aftermarket and they are not binned to fit the 55 different models

No matter what i get it will have to get water cooled at some point, not really to OC the hell out of it, but more to keep things nice and quiet with a lot of fans but spinning at a substantial lower speed.
TBH i have never overclocked a GFX card as i can recall, CPU i have since 700 MHZ was it and up until a couple of years ago.

Okay my 12 core TR are kicked up to 4 ghz and my ram do run at 3600 MHZ, but really that was just a few toggles thrown in bios and not what i call OC.
 
Sadly true. And the lack of real competitiveness from AMD is keeping GPU prices in the stratosphere. With poor Nvidia pricing, and poor AMD performance, I'll have this 1080Ti for a looooong time, it seems.

Competitive doesn't mean price war.

Navi 5700XT appears quite competitive with 2070 and likely costs a little less to build.

But AMD is likely only going to offer a mild perf/$ advantage over NVidia, not a huge one that some people keep dreaming about.
 
Competitive doesn't mean price war.

Navi 5700XT appears quite competitive with 2070 and likely costs a little less to build.

But AMD is likely only going to offer a mild perf/$ advantage over NVidia, not a huge one that some people keep dreaming about.

What the heck? I agreed with Snowdog? Now what? ;) :D Personally, I think Nvidia did nothing more than rebadge their previous generations card and added RTX features to the mix, meaning TitanXp = 2080Ti. The 5700 series prices are fine and in fact, the 5700XT is faster than the Vega 64 and costs less than the Vega 64 at launch.
 
What the heck? I agreed with Snowdog? Now what? ;) :D Personally, I think Nvidia did nothing more than rebadge their previous generations card and added RTX features to the mix, meaning TitanXp = 2080Ti. The 5700 series prices are fine and in fact, the 5700XT is faster than the Vega 64 and costs less than the Vega 64 at launch.
You clearly do not know what the term "rebadge" means. If you actually take the time and look at any in depth review then you will know that Turing has many architectural changes and is far from just Pascal with RTX thrown in. Heck it has just as good as an IPC gain over Pascal that Navi has over Polaris and Vega. A true example of a rebadge would be the RX 580 which is essentially nothing more than an RX 480.
 
Last edited:
What the heck? I agreed with Snowdog? Now what? ;) :D Personally, I think Nvidia did nothing more than rebadge their previous generations card and added RTX features to the mix, meaning TitanXp = 2080Ti. The 5700 series prices are fine and in fact, the 5700XT is faster than the Vega 64 and costs less than the Vega 64 at launch.

wut
 
Sadly true. And the lack of real competitiveness from AMD is keeping GPU prices in the stratosphere. With poor Nvidia pricing, and poor AMD performance, I'll have this 1080Ti for a looooong time, it seems.
Do not understand the poor AMD performance. Navi if AMD is accurate is faster than a 2070 overall (minus RTX gaming which is rather insignificant for games and for performance the 2070 gives on top of that).

Vega Vii competes with the 2080, pushing both to max quality settings, newer games and 4K it is Nvidia 2080 that seems to fall behind. 8gb maybe just sufficient for today games with some hiccups but for the price it is kinda a ripoff which is compounded even further when RTX is used, look at video between the two. Vega Vii is kinda a weird card but it does do well compared to the 2080 in games and for other stuff way faster. Navi I see can only improve when drivers mature more with the new architecture and game developers get a handle on it. Unless Nvidia has some surprises with Super, AMD will do fine.



In other words AMD does not really have poor performance per the price except in the $250-$300 range except Vega's are still available covering those price points.
 
  • Like
Reactions: N4CR
like this
Competitive doesn't mean price war.

Navi 5700XT appears quite competitive with 2070 and likely costs a little less to build.

But AMD is likely only going to offer a mild perf/$ advantage over NVidia, not a huge one that some people keep dreaming about.

GPU prices have been stupid for a while. We all remember when the top end consumer cards weren't pushing $1000, and midrange products weren't pushing $500. Now, AMD has ceded the high-end to Nvidia. By itself, I guess that's fine - though I miss the competition - but where AMD could make a big dent in the market is on midrange pricing. And they don't. I presume some beancounters somewhere in the company think they can make more profit this way, but I disagree. I think AMD could make some serious volume sales if they pushed prices down a bit. But hey, beancounters and marketing moonies disagree, I guess.
 
I presume some beancounters somewhere in the company think they can make more profit this way, but I disagree. I think AMD could make some serious volume sales if they pushed prices down a bit. But hey, beancounters and marketing moonies disagree, I guess.

This is the knee-jerk reaction of so many internet armchair "experts".

But this is a superficial reading IMO. It's also they same reason many here argued that AMD would have drastically lower prices on Navi, before release, and now when proven wrong, they will argue that AMD doesn't' know it's own business. :rolleyes:

The company has ALL the numbers. The upfront R&D they need to recovery, the production costs, the profit margins, and decades of sales history.

Not only that, but just an outsiders common sense to realize that if AMD lowers prices enough to eat NVidias market share a lot, NVidia can just lower prices to return to the status quo on market share, expect now with less money for each of them.

There is nothing to be gained by AMD to start a price war that will only result in both AMD and NVidia losing profit margin.

But I get that wishful thinking is a powerful force.
 
GPU prices have been stupid for a while. We all remember when the top end consumer cards weren't pushing $1000, and midrange products weren't pushing $500. Now, AMD has ceded the high-end to Nvidia. By itself, I guess that's fine - though I miss the competition - but where AMD could make a big dent in the market is on midrange pricing. And they don't. I presume some beancounters somewhere in the company think they can make more profit this way, but I disagree. I think AMD could make some serious volume sales if they pushed prices down a bit. But hey, beancounters and marketing moonies disagree, I guess.

Not defending $1000+ consumer GPUs. Still have to say... GPUs are now big business. At one time there where 20 players, and pretty much any little group of smart folks with a good idea could spin up a GPU company.

GPUs we are using today have 100s of millions in R&D behind them... and are fabbed on the same process high end CPUs are. Then there is the more and more expensive memory.

The worst part about AMD giving up on the high end is how it effects mid range costs for them. NV has the advantage of making one super complicated bit of silicon. Then selling the cast offs as mid range parts (and to be honest their high end consumer parts are also cast offs). I do expect navi+ costs down the line might be better when AMD is using the fully functioning versions in their server cards.
 
This is the knee-jerk reaction of so many internet armchair "experts".

I don't claim to be an expert, mang. Do you?

But this is a superficial reading IMO. It's also they same reason many here argued that AMD would have drastically lower prices on Navi, before release, and now when proven wrong, they will argue that AMD doesn't' know it's own business. :rolleyes:

Apparently "when proven wrong" means something different to you. We won't know who is right until sales data comes in.

The company has ALL the numbers. The upfront R&D they need to recovery, the production costs, the profit margins, and decades of sales history.

Because companies are never wrong and don't make bad calls despite having all the data. Got it.

Not only that, but just an outsiders common sense to realize that if AMD lowers prices enough to eat NVidias market share a lot, NVidia can just lower prices to return to the status quo on market share, expect now with less money for each of them.

This depends greatly on how much Nvidia cares. Nvidia owns the high-end, high-margin market. And it's not like they'd go broke if AMD clawed a bigger chunk out of the midrange and low-end markets.

There is nothing to be gained by AMD to start a price war that will only result in both AMD and NVidia losing profit margin.

Market share is no small thing. AMD's GPU brand is suffering.

But I get that wishful thinking is a powerful force.

What wishful thinking? Under current conditions, I'll just buy Nvidia. No skin off my back.
 
I don't claim to be an expert, mang. Do you?

You disagree with AMD business decisions, therefore the implication is that you know the GPU business better than AMD. So you are setting yourself up as the expert, not me.

Apparently "when proven wrong" means something different to you. We won't know who is right until sales data comes in.
.

Proven wrong on Navi pricing. Many were using this argument as the basis for prediction of low Navi pricing. Obviously that pricing prediction was wrong.
 
This is the knee-jerk reaction of so many internet armchair "experts".

But this is a superficial reading IMO. It's also they same reason many here argued that AMD would have drastically lower prices on Navi, before release, and now when proven wrong, they will argue that AMD doesn't' know it's own business. :rolleyes:

The company has ALL the numbers. The upfront R&D they need to recovery, the production costs, the profit margins, and decades of sales history.

Not only that, but just an outsiders common sense to realize that if AMD lowers prices enough to eat NVidias market share a lot, NVidia can just lower prices to return to the status quo on market share, expect now with less money for each of them.

There is nothing to be gained by AMD to start a price war that will only result in both AMD and NVidia losing profit margin.

But I get that wishful thinking is a powerful force.
Since that R&D budget is not depending on just pc sales I would say that you are way of the mark again.
Sales and tactics (let me put this bluntly since you decided to call someone else armchair tactics and superficial) do not differ between sales of a previous product when they did make money on and the current situation.

To presume that they did it wrong before and now AMD is doing it right without substance (read somewhere where people posted numbers that would validate the change of tactics preferably from AMD) is launching yourself in the same role as you put someone else in.

Polaris was fine there was no indication it was not bringing in money so why would they change all of that .

I know what AMD wishes they wish they could beat Nvidia in sales. And most of us share that one wish :) (not that I care for the well being of the financial situation of AMD those dire times have passed).
 
Not defending $1000+ consumer GPUs. Still have to say... GPUs are now big business. At one time there where 20 players, and pretty much any little group of smart folks with a good idea could spin up a GPU company.

GPUs we are using today have 100s of millions in R&D behind them... and are fabbed on the same process high end CPUs are. Then there is the more and more expensive memory.

Agreed. However, there are two corollaries. Architectures have longer shelf lives, and there's less risk of competition from said 20 person startups. The players are Nvidia, AMD, and MAYBE Intel. Nvidia charges what they do because the can get away with it. In Nvidia's place I'd do the same. The question is, in AMD's place, is this the right move? I'm skeptical. AMD isn't just losing market share, they are losing mindshare.

The worst part about AMD giving up on the high end is how it effects mid range costs for them. NV has the advantage of making one super complicated bit of silicon. Then selling the cast offs as mid range parts (and to be honest their high end consumer parts are also cast offs). I do expect navi+ costs down the line might be better when AMD is using the fully functioning versions in their server cards.

Navi would be an excellent value, IMHO, at around $100 cheaper. The RX580 was kind of a dog - moar Polaris - but I recommended it to a friend who was building a budget rig some time ago, because the things were so dirt cheap, it just made sense for the kind of rig he was building. 1060 6GB prices were significantly higher. Just made sense. But even he - who wasn't well up on the market - was like "not an Nvidia card?" Mindshare, my man.
 
You disagree with AMD business decisions, therefore the implication is that you know the GPU business better than AMD. So you are setting yourself up as the expert, not me.

Everybody who disagrees with somebody's decision is claiming to be an expert. Lolwut? "I wouldn't drive that fast, man." "What, are you a racing expert or something?"

The consumer (hey, us) gets a vote.


Proven wrong on Navi pricing. Many were using this argument as the basis for prediction of low Navi pricing. Obviously that pricing prediction was wrong.

Well since I didn't make that claim... don't see why you're attributing it to me. My argument is "I think the pricing should be lower." It has nothing to do with what other people were predicting.
 
GPU prices have been stupid for a while. We all remember when the top end consumer cards weren't pushing $1000, and midrange products weren't pushing $500. Now, AMD has ceded the high-end to Nvidia. By itself, I guess that's fine - though I miss the competition - but where AMD could make a big dent in the market is on midrange pricing. And they don't. I presume some beancounters somewhere in the company think they can make more profit this way, but I disagree. I think AMD could make some serious volume sales if they pushed prices down a bit. But hey, beancounters and marketing moonies disagree, I guess.

You can disagree all you want but you would be wrong. AMD tried that for years, making cards and selling them much cheaper. You know what, it didn't work, they didn't improve their market share, they didn't make money and to top it all off they became known as a budget brand. It would only be madness to keep following that strategy.

So now days AMD price their cards differently. Remember Polaris, everyone expected them to be dirt cheap and undercut Nvidia and people were shocked when they weren't. But, the thing is Polaris was the first line of GPUs that actually made money for AMD. So while they aren't setting the world on fire, they aren't losing money on the GPU side either.

Lisa Su is a very shrewd operator. She has turned AMD around. They aren't hemorrhaging money on the GPU side trying to compete with Nvidia on the high end and cutting prices in the mid range.
 
But, the thing is Polaris was the first line of GPUs that actually made money for AMD. So while they aren't setting the world on fire, they aren't losing money on the GPU side either.

Polaris was priced perfectly. If they followed the same philosophy with Navi, that'd be great!
 
Since that R&D budget is not depending on just pc sales I would say that you are way of the mark again.

The ability to spend on R&D depends on how much money the whole company makes, and NVidia as a whole make MUCH more money than AMD does as a whole. So can afford to spend much more on R&D.

So as usual you are off the mark.
 
The ability to spend on R&D depends on how much money the whole company makes, and NVidia as a whole make MUCH more money than AMD does as a whole. So can afford to spend much more on R&D.

So as usual you are off the mark.

AMD managed to to do quite well for itself lately in the CPU space despite this handicap. And has, in the past, done well on the GPU side despite the handicap, too.

Though as Dan pointed out in one of the CPU threads... never really at the same time.
 
AMD managed to to do quite well for itself lately in the CPU space despite this handicap. And has, in the past, done well on the GPU side despite the handicap, too.

Though as Dan pointed out in one of the CPU threads... never really at the same time.

As I pointed out in another thread AMD kicked Ass twice in CPUs, both when Intel shot itself in the foot (Pentium 4 and 10nm failures).
 
Radeon 480 was priced very similar to GTX 1060.

5700 XT/Pro are priced very similar to RTX 2070/2060, so it look like the same philosophy to me.

480/580 undercut the 1060. Launch price of the 1060 was $249-$299. Launch price for the 480 was $199-$239. Launch price for the 580 was $199-$229. Pricing in the wild often showed a greater difference, like when I recommended a 580 for my friend - the difference was ~$50. Polaris was a mediocre GPU that was nonetheless priced well, sold well, and I was comfortable recommending - and would have purchased for myself if my builds were in that price range.

Vega? Meh. Navi? Looking like a "meh" too.

Part of the difference is also that in general, pricing went up a tier with Nvidia's Turing lineup. AMD probably thinks the answer is to follow Nvidia up the pricing chain. But the ~$200-$250 GPU buyers didn't see much of a gain from any of this, and those were strong buyers of AMD products.
 
As I pointed out in another thread AMD kicked Ass twice in CPUs, both when Intel shot itself in the foot (Pentium 4 and 10nm failures).

That's an oversimplification. The Athlon did well against the Pentium 3 variants, too. The P4 was a mistake in some ways, it is true. But Athlon did well against the whole Intel lineup, not just the red-headed stepchild. It wasn't like the previous K6 lineup, which was a sort of "eh good enough" attempt. It was an outright push to design something excellent on its own. Don't sell K7 short. Nobody saw that coming.

Similarly, Intel has stepped up their competitive game of late - bolting on more cores, and pushing stock clocks, and AMD is hanging with them anyway. Intel's 14nm has been refined to a level of perfection that means AMD's 7nm process advantage is barely there, in some respects. Intel may be having problems with 10nm, but they are pushing 14nm products at the top of their game. AMD's chiplet idea was brilliant, given the limitations they had to work with.

The frustrating thing with AMD is that they are not a crappy company, despite what some folks like to imply. They have occasional bursts of brilliance and excellence, and we're all excited to buy their products when they do this. Everybody likes to see the underdog win. And they have historically been, in general, much more reasonable with pricing, especially when they don't have the performance crown.
 
That's an oversimplification. The Athlon did well against the Pentium 3 variants, too. The P4 was a mistake in some ways, it is true. But Athlon did well against the whole Intel lineup, not just the red-headed stepchild. It wasn't like the previous K6 lineup, which was a sort of "eh good enough" attempt. It was an outright push to design something excellent on its own. Don't sell K7 short. Nobody saw that coming.

That is a nonsense argument. It's like arguing that RX 480 did well against the GTX 960. What matters is the current supposedly better generation. P4 was a big failure
 
Navi 5700xt beats the 2070 (hope I don't eat my words) but AMD does not compete on the high end. Vega Vii maxed out to 2080 maxed out (which is faster than the 1080Ti) at high resolutions, Vega Vii wins on the newer API titles but AMD does not compete on the high end. What exactly is the high end? The 2080 Ti only? So the 2080/2070/Navi/Vega Vii are midrange??? AMD is competing on the high end. As for the Ultra high end where the 2080 Ti/Titan Turing sits AMD does not compete. As for pricing, most 2080's available and 2070's are above the $699 and $499 price point - you can get a barebone 2070 or 2080 for MSRP but they are not pretty. The pricing of Navi initially will be way lower than the average 2070 price available today with better performance (most likely some very fine wine to go with it).

In my prospective AMD has the better deals, price/performance with Navi and with Vega Vii (especially if you do more than just game). AMD is competing but the mindset is so skewed in my view it is ridicules. From my prospective, RTX has not brought much to the gaming experience, while a technical marvel the benefits do not outweigh the vises at this time.
 
That is a nonsense argument. It's like arguing that RX 480 did well against the GTX 960. What matters is the current supposedly better generation. P4 was a big failure

Except that's not how it went down, and you should probably know that. The P6 uarch continued to be developed because, in some ways, it was superior to Netburst (especially on power efficiency, where the mobile CPUs were usually P6-derived), and eventually it evolved into the early Core lineup which replaced Netburst. So no, it's nothing like your analogy. The only thing correct there is that the P4 was a failure, in many respects (but not all).

The K6 was designed to be competitive-ish. The K7 was designed to win. At the time it took a lot of people by surprise. It wasn't just a new CPU, it was a new philosophy from AMD. Those were some good days for the PC market, for sure.

Edit: quote from your post: "Both happened while Intel was shooting itself in the foot for quite a while(Pentium 4 and 10nm failures)."

This is incorrect. Athlon predated the Pentium 4 by a year and a half or so.
 
Last edited:
Agreed. However, there are two corollaries. Architectures have longer shelf lives, and there's less risk of competition from said 20 person startups. The players are Nvidia, AMD, and MAYBE Intel. Nvidia charges what they do because the can get away with it. In Nvidia's place I'd do the same. The question is, in AMD's place, is this the right move? I'm skeptical. AMD isn't just losing market share, they are losing mindshare.

Can't disagree in terms of discrete cards. However I am not sure AMD really cards primary about discrete card business. They have the major consoles locked down, they have their major cloud gaming play. Sounds like they are even moving into the mobile space in some form with Samsung.

I'm not suggesting they are right to abandon segments of the consumer market. Just say I believe they might not see the consumer discrete GPU market as a growing one. NV put their R&D money into AI... AMD seems to have put there R&D money into scalability. Being able to use their tech in everything from phones thru servers.
 
Can't disagree in terms of discrete cards. However I am not sure AMD really cards primary about discrete card business. They have the major consoles locked down, they have their major cloud gaming play. Sounds like they are even moving into the mobile space in some form with Samsung.

I'm not suggesting they are right to abandon segments of the consumer market. Just say I believe they might not see the consumer discrete GPU market as a growing one. NV put their R&D money into AI... AMD seems to have put there R&D money into scalability. Being able to use their tech in everything from phones thru servers.

Good point, albeit somewhat depressing. We don't know how important this market really is to them, internally. Could be the discrete GPU division is where they send the marketing benchwarmers in the company, because they're focused elsewhere with respect to GPU tech.

The console wins are huge, for sure.
 
  • Like
Reactions: ChadD
like this
Good point, albeit somewhat depressing. We don't know how important this market really is to them, internally. Could be the discrete GPU division is where they send the marketing benchwarmers in the company, because they're focused elsewhere with respect to GPU tech.

The console wins are huge, for sure.

It is a bit sad no dobut. lol

I am starting to think Lisa sees a future where 95% of all gaming is happening on a console or in the cloud. Which sort of makes sense... her first real big project was the Cell processor after all.

With that in mind it really is interesting looking at what they have done the last number of years. More cores... chiplets. Scalable graphics pipes. Its almost like she in a lot of ways has just taken everything they wanted to do with Cell all those years ago and find ways to turn AMD in to a big Cell compute platform. I have a feeling the chip they revel for the PS5 and next Xbox will look a lot more like a modern Cell then some would imagine. It will have a CPU chiplet, a GPU chiplet... with a controller chip housing the IO and sound ect. Throw in a potential curve ball ASIC (perhaps for ray generation or something) and you really do have Cell 2.0.
 
Except that's not how it went down, and you should probably know that. The P6 uarch continued to be developed because, in some ways, it was superior to Netburst (especially on power efficiency, where the mobile CPUs were usually P6-derived), and eventually it evolved into the early Core lineup which replaced Netburst. So no, it's nothing like your analogy. The only thing correct there is that the P4 was a failure, in many respects (but not all).

The K6 was designed to be competitive-ish. The K7 was designed to win. At the time it took a lot of people by surprise. It wasn't just a new CPU, it was a new philosophy from AMD. Those were some good days for the PC market, for sure.

Edit: quote from your post: "Both happened while Intel was shooting itself in the foot for quite a while(Pentium 4 and 10nm failures)."

This is incorrect. Athlon predated the Pentium 4 by a year and a half or so.

I should have said K8.

K7 was competitive, it was trading blows with Intel, until K8.

K8 was AMDs clear winner, and it was competing against P4 that was running out of steam. Sure they continued PIII into mobile because P4 sucked at that from the beginning, and it was later evolved to replace P4 because again P4 was dead end. It was a low IPC design that was supposed to scale to 10GHz, but instead maxed at 3.8GHz before they abandoned it and retreated to evolving P3.
 
Polaris was priced perfectly. If they followed the same philosophy with Navi, that'd be great!

People complained that it was priced too high. People on forums were expecting the price of Polaris, based on the performance and power use, to be $189/199 for the 8GB version and $169/179 for the 4GB version. When the 4GB Version came out and was $199 there were a lot of angry posts on forums about AMD's stupidity, about bean counters and marketing moonies not knowing what they were doing, that they should lower price to help sell more stock.

But, they aren't doing that anymore. They aren't following that losing strategy of pricing cheap to gain market share. It didn't work.

Part of the difference is also that in general, pricing went up a tier with Nvidia's Turing lineup. AMD probably thinks the answer is to follow Nvidia up the pricing chain. But the ~$200-$250 GPU buyers didn't see much of a gain from any of this, and those were strong buyers of AMD products.

AMD will still have products in the $200/250 range. The trouble is that people think AMD should be selling a 2070 competitor at $250 and that's the problem. Getting back to your AMD losing mindshare as well as marketshare. Well, that's brilliant, because all that 10 years of lowering prices has done is to make AMD look like a budget brand. And trying to win market share with that Strategy hasn't really worked. They have made no significant gains but lost money.

A company can't keep cutting into it's margins to gain market share. It's not sustainable. And AMD weren't even winning market share, they have been in a steady decline for the last 10-12 years.

Their success now depends on one thing, the quality of their GPUs. What they need is a few generations of solid cards. They have their driver department working pretty well, now follow that up with a few GPU releases that are without any glaring issues and then you gain back both mindshare and market share. The right kind of mindshare.
 
People complained that it was priced too high. People on forums were expecting the price of Polaris, based on the performance and power use, to be $189/199 for the 8GB version and $169/179 for the 4GB version. When the 4GB Version came out and was $199 there were a lot of angry posts on forums about AMD's stupidity, about bean counters and marketing moonies not knowing what they were doing, that they should lower price to help sell more stock.

But, they aren't doing that anymore. They aren't following that losing strategy of pricing cheap to gain market share. It didn't work.

Ah, yes... The "Streets are paved with gold" Polaris thread. I went back to browse that thread a while back and holy cow the over-optimism was off the charts. There were a few people who had thrown out all rational sense and expected the heavens to open up and AMD to bestow them with a 50% price/performance increase.

AMD will still have products in the $200/250 range. The trouble is that people think AMD should be selling a 2070 competitor at $250 and that's the problem. Getting back to your AMD losing mindshare as well as marketshare. Well, that's brilliant, because all that 10 years of lowering prices has done is to make AMD look like a budget brand. And trying to win market share with that Strategy hasn't really worked. They have made no significant gains but lost money.

A company can't keep cutting into it's margins to gain market share. It's not sustainable. And AMD weren't even winning market share, they have been in a steady decline for the last 10-12 years.

Their success now depends on one thing, the quality of their GPUs. What they need is a few generations of solid cards. They have their driver department working pretty well, now follow that up with a few GPU releases that are without any glaring issues and then you gain back both mindshare and market share. The right kind of mindshare.

I somewhat disagree with this point. Sure, AMDs R&D and production is on a separate path and they need to recoup funds spent (and bring in revenue for forthcoming projects, shareholder value etc.). However, they do not exist in a vacuum where the market is concerned. The need market share, they need to move units -- and a $450-500 RTX 1070 competitor is not going to do them any favors, especially when nVidia will be lowering prices and releasing revised cards in short order. $400 would have moved the needle.

I do however think there's another reason AMD wants to stay close on pricing with nV: they don't want to create any impression that nV has features that set it apart -- namely raytracing. Lowering their prices too far relative to competition essentially validates that nV's new features have value above and beyond what AMD cards offer.
 
Last edited:
Ah, yes... The "Streets are paved with gold" Polaris thread. I went back to browse that thread a while back and holy cow the over-optimism was off the charts. There were a few people who had thrown out all rational sense and expected the heavens to open up and AMD to bestow them with a 50% price/performance increase.

There seems to be a constant core belief, that the Saints at AMD will bless users with massive perf/$ increases, followed by the constant grumbling when it doesn't happen.

You think eventually that this belief would be challenged by the reality of it not happening repeatedly.
 
I somewhat disagree with this point. Sure, AMDs R&D and production is on a separate path and they need to recoup funds spent (and bring in revenue for forthcoming projects, shareholder value etc.). However, they do not exist in a vacuum where the market is concerned. The need market share, they need to move units -- and a $450-500 RTX 1070 competitor is not going to do them any favors, especially when nVidia will be lowering prices and releasing revised cards in short order. $400 would have moved the needle.

All the Evidence points to this strategy not working though. It's ok to lower the prices once or twice to shake up the market or whatever, but, if you aren't making money, you can't keep doing it, especially when you are not actually increasing your market share from it.
 
Back
Top