AMD is missing a big opportunity

AMD is making some fairly dumb moves right now. They seem to be investing a lot into the idea that people will want APUs, at least in the casual market. If you look at it objectively, it would make some sense. Casual games make up the majority of what people are actually playing on pc. LoL, Dota, CS, etc are all relatively easy to run games. The 7850k, which is now $140, can play all of those really well. In fact, it can max out LoL at 1080p, getting 60fps from the numbers I saw. So then there must be a huge market for those products, right? Wrong. Those casual gamers don't know much about the market, so they don't know that what would best suit their needs is an APU.

It is also the cheapest way to go about building a decent gaming rig for games like that, but all that casual gamers tend to know is "Intel and nVidia are the best". It wasn't until recently that I had the epiphany that AMD NEEDS NEEDS NEEDS to be marketing to those people. They need an ad campaign centered around making the cheapest gaming pc capable of running whatever game they are advertizing on. They are at a Dota championship? Then tell them how awesomely cheap it is to build a capable Dota box with an APU. A LoL championship or Twitch stream? Show some numbers of frame rates between the their APUs and equivalently priced competition in LoL. But they aren't doing that. People know Intel. They want to be able to say that they have an i7 or whatever, or that they have a dedicated nVidia gpu, even if their system is a piece of junk overall and generally wasted money on it. Casuals need to be pandered to, and AMD just isn't pandering. That seems to me to be why the only people who are into AMD are those that are more knowledgeable (and those who want to get away from the "mainstream" support of Intel and nVidia).

What do you guys think? Should AMD market towards the F2P crowd more or something similar?

its probably because intel is dominating the Gaming market since 2012.
AMD is not getting a feet on the ground in this market.
But they can only blame their selfs in my opinnion.

Intel is miles ahead and im realy doubting that AMD is going to make a big comeback, even with the ZEN rumors, im still doubting that they going to make it in the gaming enthusiast world realy.

I think they should improve their marketing statergy firstly, but next to that, they should simply speed up.
How long are we waiting for the new Radeon cards allready?
It just takes way to long, and people are sick of waiting and hoping i guess.
And that is understandable.

How AMD can compete:

Ditch "high end" APUs
Make a Haswell E competitor for half the cost
Improve single core performance
Put more focus into the GPU market, and design programs that take the load off the CPU and onto the GPU (IE a version of mantle that accelerates all apps)
Release the entire 300 series with HBM
Ditch FM2+ in favor of AM4
Implement DDR4
Use PCIe Gen 3.1 (or is it 4)


AMD simply does not have the funds to advertise right now. Poor management got them into this situation, but read and now su has seemed to do a pretty good job of managing the books. I think that zen will compete with intel, but not win by a small margin. There is almost no way that AMD will bead Intel unless they pull off the greatest turnaround of all time. However I think that as long as zen is close, they can hopefully make more progress with subsequent architecture upgrades.

I am not talking about Zen or the high end or the enthusiast markets. What I am talking about is right here, right now, casual gaming. Going with a 7850k seems to be the cheapest way to get a decent F2P capable pc. And no one really knows that, and that is AMD's fault. This is a very huge market just waiting for them to market. I also think that the reviewers need to have realized this themselves. They are testing products aimed at casual gaming by playing games like Crysis and Tomb Raider. Come on, guys. Why can't I find a good comparison of low end gpus and apus in games like LoL?

I didn't entirely agree with your post but you hit it on the head now - people don't even know they exist!
The real question is though, do OEMs? Most casuals don't build their own PCs

The problem there is that OEMs have no incentive to contradict the marketing that Intel and nVidia have already done. They know Intel, so OEMs sell them Intel. No reason to try to push AMD if they are willing to shell out the cash for the more expensive options.

I think AMD is simply going too slow. When the last batch of 8 cores was about to come out I was thinking that it would maybe be a turning point for them, and was considering building a system with one, but when I found out they were just 8350 re-badges, I lost all hope and interest in AMD. They stayed on 32nm for WAY too long. Also, I think they missed out big on competing with the 970. They did drop prices on the 290x, but the availability and marketing was not there to back it up.


Granted not everyone has an idea of what the magic box does and how to make the magic box faster. (Literally some customers [worked at geek squad]). I don't think this is an advertising issue with less money in the coffers. AMD for the past few years has been focusing on R&D and burning money left and right. Advertising is not exactly a priority. You have to also remember, the casual market is not the market that matters most. How do other businesses view AMD atm. A company falling off a cliff with no parachute. OEMs think they can get away with pushing AMD into really crap low end laptop markets. This era is over. We might see a price hike with AMD laptops in the future, and less plastic junk.

I think AMD is trying to go in the right direction. If the leak is true from the report on Ars Technica. Focus on high end releases first, then move down to the APUs.

Maybe I am missing the point of your post. Casuals are idiots and AMD has not put out enough advertising and marketing is what I am reading in summary. Those casuals are getting their information from us whom are more informed which then filters down to sales associates at Best Buy, Tiger Direct or Microcenter all saying screw AMD Intel is better because words and the experts say so. I would hazard to guess word of mouth and the straight performance reviews that have come out, have influenced poor advertising. Even if AMD did put more money into advertising.

Article I refer to.

I agree that AMD doesn't need to waste money in advertising. PR and advertising are expensive ventures, and AMD needs to sink that money into the future, not the present. They can't make their current lineup of mid-range chips more appealing because one look at the benchmarks tells everyone otherwise. They need to be sure that NEXT year they can make those benchmarks look better.

1 Like

Whatever AMD lacks in the gaming market, it's still a good bet for virtualization, especially if you compile the source for QEMU & KVM on Linux. As this seems to be their remaining strong point, I think this is why they are looking at developing Zen and the K12.

1 Like

Sorry mods, don't mean to necro this thread, but I just read this article and they are advertising the new apu's to be good for eSports.

HSA , This is part of where they are going with APUs.

But wouldn't a CPU with HSA work all the same? Why not offer APUs with HSA to the laptop market rather than the dektop market? I think HSA and APUs would play a vital rolein the budget market but not the enthusiast market.

I saw that and am very happy that they are taking these steps. Considering that the godavari is just a refresh, nothing is really different between now and a year ago on the hardware side of things. It is like they woke up to the market. So that is awesome. In my arrogance, I like to think that I had something to do with that.

As far as high end APUs and HSA, if HSA works well, then we could see a complete replacement of CPUs with APUs, at least, that seems to be AMD's hope. If they can use the horsepower of GPUs in the APUs for general processing, then it could really accelerate things and completely displace the high end CPU market, which would be a really interesting thing to see.

As far as HSA in laptop's, it seems that AMD is neglecting the laptop market by and large. However, I think that is on purpose. I think that they realize that APUs are a really immature product at the moment and seeing as how HSA support is very limited, there isn't really much for them to take advantage of, limiting them to low end gaming laptops right now which is really not great. I think that once they get HBM (which lest we not forget they have been working on for about as long as APUs in general) up and rolling, the potential for APUs in both gaming and HSA applications will increase significantly. I think that they are waiting for HBM implementation with the whole Zen thing to see how much power they can conceivably get out of them. With those APUs, I expect that HSA support will increase, and we might even see Adobe support it (considering that they are moving more and more towards OpenCL acceleration and that isn't too far away from HSA, from what I understand, or at least they are moving in that direction), which would be a real game changer. If APUs were faster than Intel's stuff in things like After Effects and Premier and Photoshop and whatnot, then the added sales would be very, very impressive. The professional market is a very important space (though so are the server and mobile markets, all of which are being neglected at the moment by AMD).

Yeah, I think the post under yours summed it up better than I could have.

I have literally been speculating on AMD's intentions for years now, and so far, everything that they have done has pointed towards me being right (in general at least). I think that their goal, from the time they started thinking about acquiring ATI was to get APUs to become mainstream because of the potential increase in the compute performance. That was around the time that they started looking into HBM, which was going to be necessary for something like an APU to work well. So yeah, I hope it works out well for them.

Huh. Thats pretty interesting.

Not sure if this interests you, but Linux kernel 4.0 added full HSA support...

Like for general compute? I would be interested in seeing some benchmarks from that.