AMD Q2 2015 Financials Fall Short

[expunged]

im holding out for this next cpu release and unless I see meaningful differences.

I am in the process of building a new system now going to use a haswell chip, due to skylake not having enough pci lanes. Come on Intel 20 lanes are you kidding..

You have it backwards. Skylake will have support for 40+ PCI-E 4.0 lanes. You are thinking of the Intel Haswell-E 5920K CPU which only has support for 20 PCI-E 3.0 lanes.

skylake specs
14 nm manufacturing process
LGA 1151 socket
Z170/H170 chipset (Sunrise Point)
Thermal design power (TDP) up to 95 W (LGA 1151)
Support for both DDR3 SDRAM and DDR4 SDRAM in mainstream variants, using custom UniDIMM SO-DIMM form factor with up to 64 GB of RAM on LGA 1151 variants.
Support for 20 PCI Express 3.0 lanes (LGA 1151)
Support for Thunderbolt 3.0 (Alpine Ridge)
64 to 128 MB L4 eDRAM cache on certain SKUs
Up to four cores as the default mainstream configuration
Support for SATA Express
AVX-512 F, CDI, VL, BW, and DQ for the Xeon variants
Intel MPX (Memory Protection Extensions)
Intel ADX (Multi-Precision Add-Carry Instruction Extensions)
Skylake's integrated Iris Graphics GPU supports Direct3D 12 at the feature level 12.0

From what i have read Skylake will not have 4.0 lanes until the "-E" (extreme) version (for which the release is expected in 2017)

 

[madman83]

Amd will be fine. Zen will be FX-53 and arctic islands will be 9800xt. Intels broadwell-e will compete along with nvidias pascal. Amd may not hold the crown for long as intel and nvidia products will follow 1-3 months after amd launches them. But really people crying about amd not being competitive? 1-7 fps losses with fury, grow up thats competitive. People crying about overclocking? Grow up there is a reason cards come at said clock rates. If you factor overclocking into your purchase decision sure, but over 80% of the market DOES NOT. I've purchased many amd, intel, nvidia and ati products. I always purchase what is best or 2nd best in the market; that being said amd/ati accounts for roughly 50% of my purchases ranging back to the K7 when 3dfx was getting my income every year. Folks forget Nvidia was never competitive until it released its riva tnt2 and ati was not until it released its radeon 7000 series. Nvidias geforce 3, 4 and fx5k series were not competitive at all in terms of price per performance , settling for 2nd best the whole generations. Amd always performed better then intel until the core 2 duo launched. Amd wrote the x64 instructions set, first cpu to 1ghz, first dual core cpu. The company has great ideas. Its a shame that the management has been so awful since the 90's that the company suffers year in and out. This new management will take amd to a plateau it hasnt seen since the 64x2. Intels IGP dominates the market, amd will change that with zen. Even if amds chip performs 5% less than intel its IGP will still perform 40% better. Thats where amd wants to make it money. Not to mention all the work amd has put into multithreaded workload performance? Once that becomes the norm amd will wear the crown for quite some time. I am a fan of all camps and hold no bias toward any one of them, but people saying amd isnt competitive? the 390 kicks but and for the price there is nothing better, it beats the 290x and we know its not the slight clock increase that does it. The market is in a strange place at the moment im looking to replace 670 sli and the choices are tough. I could get a furyx or 980ti but waiting until next year is sure to yield a 30% gain over current gen. Amd needs better marketing, clear and realistic goals and proper implementation and all will be well. I have high hopes because without amd intel and nvidia have no reason to improve.

 

[InvalidError]

with VR and other techs coming along with in the next year or two I want to upgrade. up until now I haven't had a good reason to.

VR has been "coming soon" for over a dozen years since the first LCD shutter glasses from ATI and Nvidia or Nintendo's VirtualBoy. While the barriers to entry are getting lower over time, enthusiasm about it has not been rising quite as much. If you look at 3DTV flop, most people do not appear to be particularly interested in stereoscopic 3D entertainment, so VR may very well end up being a perpetual niche market.

 

[geraldfryjr]

For one I think that they screwed up big time in a Big chunk of sales when they shorted us by reneging on producing the Steamroller and Excavator core for the AM3+ platform.

I for one Love AMD and I had the chance to build a great machine for cheap when I got my FX-6300.
Don't get me wrong because I love my machine!
But, I held out building it for the longest time waiting for the 8 core chip and the promise that a better version would be out the next year.

Had I known that they were not going to come out with the new super improved version instead of just a clocked up version of the 8 core I would have went with the Intel platform!!!

Instead they introduced some new lame FM2 platform with that core that was a waste and a pain to try expand and build nice super desktop computer out of.
They lost out on a huge chunk of sales and recognition from the diehard AMD fans and followers like myself.

Ya, I got my FX-6300 up to 5.0Ghz and even booted at 5.1Ghz on air and cheesy power supply, and, I have ran it 24/7 at 4.85Ghz for a whole year rarely ever shutting it off and rebooting!!

So what exactly is so special about the FX-9590 ?!!!!
Clock?.....Phooey!!!!

Maybe one day I will still get a FX-9590 when the price comes down a bit more IF it is still available!!!! :/

They did the same thing with the 939 socket!!
Just weeks after I had bought my Opty185....Poof they were discontinued!!
And the new Hex core Phenom's were basically the same in clock for clock performance!!

I am glad I didn't fall for that one even though it still stands as one of the best processor's they have produce, until the Pildriver's came out!!

jer

 

[hcslorg]

The interesting thing about AMD is that they compete on two fronts--GPUs and CPUs--with companies that each (for the most part, though not completely) focus on just one of those arenas: Nvidia in the graphics world and Intel in the CPU world. It takes enormous resources to double up like that, especially if a company expects to come out on top of the industry. Basically, they're spread too thin to win.

I use an AMD processor and and AMD graphics card and I'm happy with them. I like that they offer such cost effective options. The fact is, "bang for the buck/jack-of-all-trades" is their business strategy of choice and they will continue to lose ground until they make a true effort to be the best in the field, an effort that would require some big changes. Until then, people need to change their expectations. If you're looking for someone to challenge Nvidia or Intel, you'll have to look somewhere else.

I know this post is a few days old, but I thought I should add that Intel has a cross-license agreement with Nvidia. It is mentioned in Intel's Form 10K. A summary of the report can be read here. I speculate that Intel HD graphics (which happened to come out around the same time as this agreement) was brought about in part by this agreement. Also, Nvidia came out with the Tegra 3 in 2011, which again may have come about in part by this agreement. I am no expert, but it seems to me that Intel and Nvidia have had an agreement that greatly benefits both organizations and have left AMD as the odd man out.

 

[synphul]

I'd like to see amd come out with something new finally. Seems they've been stuck in a perpetual mode of rehashing old products, whether the fx series or their gpu's. I was kind of sad to see ati swallowed up by amd but maybe ati wouldn't have survived on their own. Looking at the r9 290 which has been out for awhile, they add $100 and double the vram with a few insignificant tweaks and call it a 390. In most reviews they've resorted to nitpicking details to try and show it's not just a rebadge. If you have to work that hard to convince people, it's a rebadge. Just call it what it is.

I don't think intel's prices will go up, as it is the price of current chips and cost of upgrading vs the slight performance increases have seen people hanging onto their chips longer. To raise prices would just make upgrade paths slower and hurt overall sales. What good is raising the price $50 and end up having a mean time between upgrades go from 4 to 6yrs. The higher cost won't make up for those losses in sale frequency.

On a somewhat unrelated note, people are complaining about the high price of pc's. Intel is 'too expensive'. Imagine a $1500 build today, quality components from psu to motherboard, an intel i5 or i7 paired with a gtx 980 and 8gb of ram, a complete tower. Now compare it to the gateway 2000 with a p2 266, 128mb of ram, 8mb video card and a 6.4gb hdd (for "all" your files lol). It was considered a 'budget' system at $1499 in the day (albeit 20yrs ago).

The point being a 'budget' system was $1500, not a nicer upper end system like today with the i7 build. I can recall many pc sales flyers and systems were regularly $2-4,000. Factor in inflation since a us dollar from 20yrs ago had a lot more value than it does today. All in all, the price of pc's today is cheap by comparison. You didn't have a pc for every family member, you had A pc - 1 - for the family to use if you were lucky. It certainly didn't play gta v or crysis. It barely got onto the internet to check email.

 

[Ryangrant]

The interesting thing about AMD is that they compete on two fronts--GPUs and CPUs--with companies that each (for the most part, though not completely) focus on just one of those arenas: Nvidia in the graphics world and Intel in the CPU world. It takes enormous resources to double up like that, especially if a company expects to come out on top of the industry. Basically, they're spread too thin to win.

I use an AMD processor and and AMD graphics card and I'm happy with them. I like that they offer such cost effective options. The fact is, "bang for the buck/jack-of-all-trades" is their business strategy of choice and they will continue to lose ground until they make a true effort to be the best in the field, an effort that would require some big changes. Until then, people need to change their expectations. If you're looking for someone to challenge Nvidia or Intel, you'll have to look somewhere else.

I agree they could be spread thin, but lets not fool ourselves on their sales over the past few years, we know they cleaned house with the mining community backing AMD for at least 2 years now. That was basically two years of funding money they collected on and could use for R&D/expansion. Other companies haven't had the mining community backing them as strong, and mining is a large industry. It should have been expected that the Q2 results would be falling short, they won't be having a re-occurrence of prosperous sales now that the mining community has no use for video cards really.. Sales will be down and now they'll have to put the money they made from the mining community to good use on R&D and come out ahead of nvidia/intel and hope to win over more gamers/consumers.

 

[expunged]

It will be interesting to see what happens in the CPU market with Intel being behind in the 7nm process. This will force both AMD and Intel to step up their games.

 

[AJ5000]

Can you even remember the last time AMD had the best video card or processor on the market for high end devices. It was probably back when Matrox was a player... They might have had a good low end card, but for the upper end performance products they have not been the top dog for over 20 years. Now that they are using HBM and have the HBM2 deal going, they should partner up with IBM on the 7nm process, then they would have a good shot. True competition in the market place is good for us end users.

The last time they had the best graphics card on the market? That would be when the Radeon R9 290X launched and before it, the Radeon 7970 when it launched. Before that, the Radeon 5870 when it launched. For CPUs, it's been a while. I think the last itme they had the best for CPUs was before Core 2 launched when it was netburst against Athlon 64.

Of course, that depends on what you mean by best... I'm assuming you mean highest performance with a single GPU consumer card or a single consumer CPU.

Benchmarks show that the Radeon R9 290X was below the gtx 690 as shown here, while it did win in some categories it fell behind in most of them. Don't get me wrong i am not a fanboy, If AMD was to bring out a card that would out perform it's rivals i would buy it. Currently I am waiting on windows 10 with DX12 so I can see the results of the fury x. then i will make the decision on either the Fury x or 980 ti. As it stands now they are neck and neck.

http://www.tomshardware.com/reviews/radeon-r9-290x-hawaii-review,3650-10.html

GTX 690 is a dual GPU solution. 680 is their single GPU solution. To compare apples to apples, 690 should be compared to R9 295x2, which is the fastest GPU solution currently 2 GPU on single card, and excluding SLI and CF.

 

[Haravikk]

The biggest problem I see for AMD is that they're still failing to really capitalise on some of the great technologies they have, particularly in the gaming space.

What I'd love to see is someone manufacturing a system (a Steam Machine could be a good candidate) combining an AMD APU with an AMD discrete GPU. If they could get an engine onboard, such as Unreal Engine 4, they could have a capable but inexpensive machine if they were able to combine OpenCL physics on the APU with Mantle (or other hardware-focused) graphics, HSA to maximise memory and so-on.

But so far we haven't seen any systems showcasing what these technologies combined are capable of, and that's the biggest problem here, as it means that instead of AMD products being sold together we're mostly seeing AMD processors with the lower end Nvidia cards for budget systems; while that is still money for AMD, it's not quite the result they need.

I'm also a bit disappointed that we aren't seeing more chips like those featured in the Xbox One and PS4; I know the chips they use are relatively slow (1.8ghz or so) but low-power eight core chips could really work in AMD's favour, especially if they can get their turbo boost equivalent to fill in the gap on raw speed for less threaded applications.


Of course business wise they also need a big win in some other market; I still think the server space is the way to go, but I'm not sure what kind of hook they can give that would let them claw out more space for themselves.

 

[Adam Shades]

Where are the Intel Vs AMD Linux compiled benchmarks?? All I see is windows and a lot of googling round has shown me that AMD CPU/APU chips perform a lot closer to Intel when running under an AMD compiled OS using AMD compiled executables.
A lot of this has to do with everything using Intel compilers I have read.
I'd like to really see some real bench marks for AMD under environments that don't use Intel compilers throughout.
Not that I like to post other competing web site links so, I won't however, it seems there is some room for AMD yet. Intel are still faster per core however AMD don't suck so badly. To me it was obvious, if you want the best performance out of your AMD CPU, a Linux environment is going to be the OS of choice. Google AMD Vs Intel and a swag of stuff starts to come up.
http://forums.anandtech.com/showthread.php?t=2394740

 

[ateam131]

I seriously don't know how they can be down financially unless its either due to R&D costs for new tech, or something sketchy is going on inside there. Their CPU/GPU combos are in every XBOX ONE and PS4 in the world, they must get some royalty for every single of those units sold not to mention their GPUs being about half of the market, there is no way Nvidia has that completely locked down, I personally have been using Nvidia GPUs since GTX280 era but I have often swapped between the two flavors... AMD makes good cards.. just not very good drivers, which I hear have improved a lot since I last used and ATi / AMD card. And they sell massive amounts of CPUs.... I have used AMD CPUs for over 10 years now, being a gamer I don't have the need for Intel CPU. So what is happening here... lol I hope they get their ducks in a row because better competition between these manufactures will just mean better products and price points for all of us... the CONSUMER!

 

[InvalidError]

I seriously don't know how they can be down financially unless its either due to R&D costs for new tech, or something sketchy is going on inside there. Their CPU/GPU combos are in every XBOX ONE and PS4 in the world, they must get some royalty for every single of those units sold not to mention their GPUs being about half of the market, there is no way Nvidia has that completely locked down

AMD gets about $10 per PS4/XBO SoC they ship and the two combined represent only 10-15M units per year. A 150M$ cash injection from consoles does not help much against their much larger losses.

On GPUs, AMD does not have half of the market. Based on the Steam Survey, only 28% of people use AMD/ATI GPUs as their primary GPU in their PC vs 52% for Nvidia and 20% for Intel.

CPU-wise, AMD simply does not have the performance and power efficiency to command decent margins on their chips, so they are stuck selling their chips dangerously close to costs just to shift inventory. A "massive amount of CPUs" with razor-thin margins does not help much with covering other costs like R&D for future products.

 

[ransu]

AMDs problem is the lack of competition on the high end market. For me using heavy software like 3D rendering, video editing, game creation (which files take 90 minutes to pull together "zip" on an i5 2500K computer), one really wants the best for the next machine.
At the moment there isn´t even anything to discuss from AMD. Even the coming Intel Skylake is "only" about 30% faster than this old i5 hardware from 2011. AMDs ZEN is far away in the horizon, and one doesn´t know if it can compete or if it´s just talk, I fear it´s much talk but no real competition when it comes out.
On the GPU side again AMD is at least up to the fight with FURY X, it´s a bit slower than NVIDIA 980 Ti, but enough good to give a hard competition. Also the R9 300 series are a good option for speed.
I think AMD will get some incomes from the R9 series, and some small stuff from the 6th gen APU, but the real deal for the future is if ZEN can deliver, and be released in time for that.

 

[sirstinky]

AMD answered questions about how it has managed its cash and its debt, with company executives pointing to a long runway where its financial maneuvers could continue to withstand performance weakness.

AMD Q2 2015 Financials Fall Short : Read more

AMD is in trouble and this isn't anything new. After the Core CPU's from Intel came out, AMD has been doomed. They have been unable to compete with Intel in the CPU business for quite some time, and haven't been able to introduce any new CPU's in over 2 years (with the exception of the APU's) whereas Intel has rolled out it's 5th generation Core models. Yes, they have the new APU chips being rolled out, but as of now, they still can't compete with the comparable Intel chips. The integrated graphics are nice if you're building a general purpose office/home rig, but most everyone who does a gaming machine just buys a separate card anyways, and the processor side is still no match for the low-end i5.

What does AMD have going for it? The 3xx GPU's are nice, and have a lot of promise for later improvements, but still fall short to NVIDIA's similar cards. Yes, the 295 is still the fastest card you can get, but it's not practical for most users. It's a tough call when you're selling a $600 card to a small group of enthusiasts that demand the best performance. AMD has cashed in on the console market, supplying APU's for the Xbox and Playstations, but it's not enough to save them without new CPU's to go with the GPU's. Intel owns that market segment, and are able to put more money back into their R&D.

AMD needs creative thinking and aggressive strategy if it wants to save itself. Let's see if ZEN does that. CPU tech has just about hit a wall, and if AMD can deliver something that nudges that wall, I think it would go a long way in long-term success. However, all that takes time and money. Intel was able to pull it off back in 2005 with the Pentium D after cobbling together a couple of P4 dies on a single PCB with very high clock rate (and the addition of true enthusiast chips) to beat the Athlon X2, but times have changed. You can't win customers with an insane CPU that draws 220W, needs a special motherboard and ridiculous cooling to run-and still not beat the i7 that can run on an $80 board and a stock cooler just fine.

 

[expunged]

AMD answered questions about how it has managed its cash and its debt, with company executives pointing to a long runway where its financial maneuvers could continue to withstand performance weakness.

AMD Q2 2015 Financials Fall Short : Read more

AMD is in trouble and this isn't anything new. After the Core CPU's from Intel came out, AMD has been doomed. They have been unable to compete with Intel in the CPU business for quite some time, and haven't been able to introduce any new CPU's in over 2 years (with the exception of the APU's) whereas Intel has rolled out it's 5th generation Core models. Yes, they have the new APU chips being rolled out, but as of now, they still can't compete with the comparable Intel chips. The integrated graphics are nice if you're building a general purpose office/home rig, but most everyone who does a gaming machine just buys a separate card anyways, and the processor side is still no match for the low-end i5.

What does AMD have going for it? The 3xx GPU's are nice, and have a lot of promise for later improvements, but still fall short to NVIDIA's similar cards. Yes, the 295 is still the fastest card you can get, but it's not practical for most users. It's a tough call when you're selling a $600 card to a small group of enthusiasts that demand the best performance. AMD has cashed in on the console market, supplying APU's for the Xbox and Playstations, but it's not enough to save them without new CPU's to go with the GPU's. Intel owns that market segment, and are able to put more money back into their R&D.

AMD needs creative thinking and aggressive strategy if it wants to save itself. Let's see if ZEN does that. CPU tech has just about hit a wall, and if AMD can deliver something that nudges that wall, I think it would go a long way in long-term success. However, all that takes time and money. Intel was able to pull it off back in 2005 with the Pentium D after cobbling together a couple of P4 dies on a single PCB with very high clock rate (and the addition of true enthusiast chips) to beat the Athlon X2, but times have changed. You can't win customers with an insane CPU that draws 220W, needs a special motherboard and ridiculous cooling to run-and still not beat the i7 that can run on an $80 board and a stock cooler just fine.

To me it seems obvious what they need to do. Nvidia partnered up with intel. AMD needs to do something similar. Either get out of the CPU business or cut down to the low end market for budget CPUs. They can't afford to compete on 2 fronts right now. Go all in on the GPU market. Make strategic partners to help with the R&D (like IBM and Samsung) get in on their 7nm process, couple that with HBM2. Spend some time at the top and make money. After they recoup some of the money they have lost slowly get back into the high end CPU market. Right now AMD reminds me of Amiga. The have a great product that only appeals to a niche market. But instead of focusing on what they do best, they are trying to please everyone and driving the company into the ground. May 5 2000 their stock was worth $92.25, today its $1.76. if they are not careful and drop below $1 its all over.

 

[TruthX]

Microsoft should buy AMD and build it a couple of new 10 -12 nm fabs and maybe acquire a couple of 20 nm fabs MS gets exclusive use of 10 nm APU' for use in surface and Xbox pro tabs and a full x86, and with unique ARM Licensing this could prove useful to hardware and differentiating 1st party products from Microsoft XBox and Lumina arms it will also give MS a new portfolio of products that are customized from the ground up solely from the innovation labs at MS

 

[sirstinky]

I'd like to see amd come out with something new finally. Seems they've been stuck in a perpetual mode of rehashing old products, whether the fx series or their gpu's. I was kind of sad to see ati swallowed up by amd but maybe ati wouldn't have survived on their own. Looking at the r9 290 which has been out for awhile, they add $100 and double the vram with a few insignificant tweaks and call it a 390. In most reviews they've resorted to nitpicking details to try and show it's not just a rebadge. If you have to work that hard to convince people, it's a rebadge. Just call it what it is.

I don't think intel's prices will go up, as it is the price of current chips and cost of upgrading vs the slight performance increases have seen people hanging onto their chips longer. To raise prices would just make upgrade paths slower and hurt overall sales. What good is raising the price $50 and end up having a mean time between upgrades go from 4 to 6yrs. The higher cost won't make up for those losses in sale frequency.

On a somewhat unrelated note, people are complaining about the high price of pc's. Intel is 'too expensive'. Imagine a $1500 build today, quality components from psu to motherboard, an intel i5 or i7 paired with a gtx 980 and 8gb of ram, a complete tower. Now compare it to the gateway 2000 with a p2 266, 128mb of ram, 8mb video card and a 6.4gb hdd (for "all" your files lol). It was considered a 'budget' system at $1499 in the day (albeit 20yrs ago).

The point being a 'budget' system was $1500, not a nicer upper end system like today with the i7 build. I can recall many pc sales flyers and systems were regularly $2-4,000. Factor in inflation since a us dollar from 20yrs ago had a lot more value than it does today. All in all, the price of pc's today is cheap by comparison. You didn't have a pc for every family member, you had A pc - 1 - for the family to use if you were lucky. It certainly didn't play gta v or crysis. It barely got onto the internet to check email.

AMD answered questions about how it has managed its cash and its debt, with company executives pointing to a long runway where its financial maneuvers could continue to withstand performance weakness.

AMD Q2 2015 Financials Fall Short : Read more

AMD is in trouble and this isn't anything new. After the Core CPU's from Intel came out, AMD has been doomed. They have been unable to compete with Intel in the CPU business for quite some time, and haven't been able to introduce any new CPU's in over 2 years (with the exception of the APU's) whereas Intel has rolled out it's 5th generation Core models. Yes, they have the new APU chips being rolled out, but as of now, they still can't compete with the comparable Intel chips. The integrated graphics are nice if you're building a general purpose office/home rig, but most everyone who does a gaming machine just buys a separate card anyways, and the processor side is still no match for the low-end i5.

What does AMD have going for it? The 3xx GPU's are nice, and have a lot of promise for later improvements, but still fall short to NVIDIA's similar cards. Yes, the 295 is still the fastest card you can get, but it's not practical for most users. It's a tough call when you're selling a $600 card to a small group of enthusiasts that demand the best performance. AMD has cashed in on the console market, supplying APU's for the Xbox and Playstations, but it's not enough to save them without new CPU's to go with the GPU's. Intel owns that market segment, and are able to put more money back into their R&D.

AMD needs creative thinking and aggressive strategy if it wants to save itself. Let's see if ZEN does that. CPU tech has just about hit a wall, and if AMD can deliver something that nudges that wall, I think it would go a long way in long-term success. However, all that takes time and money. Intel was able to pull it off back in 2005 with the Pentium D after cobbling together a couple of P4 dies on a single PCB with very high clock rate (and the addition of true enthusiast chips) to beat the Athlon X2, but times have changed. You can't win customers with an insane CPU that draws 220W, needs a special motherboard and ridiculous cooling to run-and still not beat the i7 that can run on an $80 board and a stock cooler just fine.

To me it seems obvious what they need to do. Nvidia partnered up with intel. AMD needs to do something similar. Either get out of the CPU business or cut down to the low end market for budget CPUs. They can't afford to compete on 2 fronts right now. Go all in on the GPU market. Make strategic partners to help with the R&D (like IBM and Samsung) get in on their 7nm process, couple that with HBM2. Spend some time at the top and make money. After they recoup some of the money they have lost slowly get back into the high end CPU market. Right now AMD reminds me of Amiga. The have a great product that only appeals to a niche market. But instead of focusing on what they do best, they are trying to please everyone and driving the company into the ground. May 5 2000 their stock was worth $92.25, today its $1.76. if they are not careful and drop below $1 its all over.

AMD's glory days are over. Yes, they bought ATi years ago and have done well with it, but so far that's been the extent of their strategic expansion. They did well with the APU's and that's borne fruit for them, but that only goes so far. Yes, they need to take advice from their competitors and stop trying to please everyone. That is a sure way to go broke if you don't do it right. From the beginning AMD was known for one thing: CPU's. They revolutionized the CPU market with their dual core chips and owned Intel for a short while, until Intel stepped up with their army of engineers and rolled out Pentium D, then Core Solo, Core Duo, Core 2 Duo, ix-xxx line, then their Sandy, Ivy, Haswell, Broadwell, and now SkyLake, while AMD is stuck on old 28nm modular dies with high clock rates and power consumption (sound familiar? Pentium D960/70?). Not to mention Intel's thriving enterprise branch with Xeon's found in probably 90% of servers and workstations. AMD failed to do that with Opteron.

Now, they are afloat because of their pricing for performance metrics, GPU's and loyal fan base. The price of the FX-8xxx chips has come down, and although it still takes a 8350/70 to challenge an i5-4440, some of us can't drop $200 on a chip, so AMD is the only option. The 290x cards a year and a half ago sold for almost $500. Now you can get one for a little over $300 (or less) because of the 970 and 980. That hurts profitability, and it's showing. Failing to innovate and make it scalable in this market and you won't last long. You can't survive by launching products that barely compete with the competition at a similar price point. I want to see AMD succeed, but you can't do it with lackluster products and emaciated balance sheets.

 

[sirstinky]

Microsoft should buy AMD and build it a couple of new 10 -12 nm fabs and maybe acquire a couple of 20 nm fabs MS gets exclusive use of 10 nm APU' for use in surface and Xbox pro tabs and a full x86, and with unique ARM Licensing this could prove useful to hardware and differentiating 1st party products from Microsoft XBox and Lumina arms it will also give MS a new portfolio of products that are customized from the ground up solely from the innovation labs at MS

Either MS buys AMD or partners with them. If it's a buyout, shareholders won't be pleased, but Microsoft has the cash and the know-how and the infrastructure (fabs, labs, and products) necessary for product development and deployment. Unfortunately, MS already has a processor contract for Surface, and that's Intel. It would take some persuasion to get their APU or a SoC into a Surface tab. However, with MS in it's back pocket, I think AMD would be able to put their hands to the grindstone and turn out some great products. But that takes time-time for infrastructure (fabs) R&D, that would only work if they had help. It's possible to turn out a revolutionary product with the right help in a year and a half, but the key is "help."

 

[sirstinky]

Can you even remember the last time AMD had the best video card or processor on the market for high end devices. It was probably back when Matrox was a player... They might have had a good low end card, but for the upper end performance products they have not been the top dog for over 20 years. Now that they are using HBM and have the HBM2 deal going, they should partner up with IBM on the 7nm process, then they would have a good shot. True competition in the market place is good for us end users.

The last time they had the best graphics card on the market? That would be when the Radeon R9 290X launched and before it, the Radeon 7970 when it launched. Before that, the Radeon 5870 when it launched. For CPUs, it's been a while. I think the last itme they had the best for CPUs was before Core 2 launched when it was netburst against Athlon 64.

Of course, that depends on what you mean by best... I'm assuming you mean highest performance with a single GPU consumer card or a single consumer CPU.

AMD has a storied past. GPU's: with 290X and 295 X2 (still has it-for now). Before that there was the 7970 (single), then the malarky 7990 (for a short time). Before those it was the 5970, 6970, then the abominable 6990 (we don't talk about that around here). For CPU's, how many of us remember Athlon 5000? Then the 64 X2 on 939 came out. Intel wasn't pleased and stuck two P4's (still Netburst) on a single PCB, clocked it to the moon, and beat the Athlon. It's been that way ever since. Core 2 Duo put the nail in AMD's coffin. Maxwell is doing something similar. Until 390 and Fury are optimized, and ZEN is delayed any more (or if it's scrapped), it's on to Chapter 11 for AMD.

 

[sirstinky]

I seriously don't know how they can be down financially unless its either due to R&D costs for new tech, or something sketchy is going on inside there. Their CPU/GPU combos are in every XBOX ONE and PS4 in the world, they must get some royalty for every single of those units sold not to mention their GPUs being about half of the market, there is no way Nvidia has that completely locked down

AMD gets about $10 per PS4/XBO SoC they ship and the two combined represent only 10-15M units per year. A 150M$ cash injection from consoles does not help much against their much larger losses.

On GPUs, AMD does not have half of the market. Based on the Steam Survey, only 28% of people use AMD/ATI GPUs as their primary GPU in their PC vs 52% for Nvidia and 20% for Intel.

CPU-wise, AMD simply does not have the performance and power efficiency to command decent margins on their chips, so they are stuck selling their chips dangerously close to costs just to shift inventory. A "massive amount of CPUs" with razor-thin margins does not help much with covering other costs like R&D for future products.

I worked for a company that was forced to sell at MAP (the lowest advertised price- usually 85% of retail-varies from vendor to vendor- meant to keep money in the vendor's pockets) and in some cases, take a loss just to get money to pay for product that resulted in massive overhead and lost profits. You can't survive selling 20 items bought for $15 at vendor, sold for $30 retail at a $10 profit after shipping. That company closed. Let's hope this doesn't be the new normal for AMD or it's only a matter of time. I can't imagine having millions of unsold units in a warehouse that you could only get maybe 40% off cost for them-some probably less). It's got to be a big burn on the balance sheets.

 

[mdsiu]

As for AMD, 2015 is more of a stop gap than anything. They said basically that at the beginning of the year. Their embedded and GPU sales will keep them treading water for this year as they set themselves up for 2016. Dibs on HBM 2 as well as experience gained using HBM 1 will help them on the GPU front, and Jim Keller will help them everywhere else. Remember, along with the Zen CPU it is highly likely he is re designing the way the whole system will operate. Hyper Transport will likely be tweaked or maybe changed completely to handle faster RAM and PCI-e speeds which will make the whole system faster of course. I am confident we will see another Atlon 64 with Zen. On top of design, Lisa Su seems to know very well what she is doing and I think she can take the company into the future given the proper hardware to work with. They are headed in a good direction overall.

If Lisa Su would know what she is doing they would not be in red for years now, she has no real strategy except trying to stretch AMD on all fronts and pray for a score somewhere.

Zen is like a unicorn at the moment, we don't know when it will really come and Intel is so advanced now that Zen literally needs to be alien technology at this point to truly compete, my guess it will be another stillborn CPU that will be decent only because of aggressive pricing. Same story with Fury, yes it's good, but when you need liquid cooling, HBM and a gigantic super-expensive die to compete with the green team on air - you know there is a major technological gap here between the two.

Let's not even talk about the fact that Fury won't bring AMD any money - the true money getters are the likes of 950TI or even lower, but with Intel's discreet graphics becoming better and better with every passing year this market closes rapidly.

What Lisa Su needs to do is to finally pick up priorities - AMD just can't try and do everything while having a mountain of debt and laughable R&D budget - they need to either focus on CPU or GPU or semi-custom/custom designs and in my opinion the last one is the most feasible, because their CPU and GPU divisions are hopeless cases really with healthy Intel and Nvidia onslaught.

LOL you blame Lisa Su?? Are you kidding? She's been CEO for less than a year. Good grief.

 

[JackNaylorPE]

One thing that has been bothering me with the latest releases is it seemed that AMD was competing w/ itself and sending mixed messages.

.... Oh 4 GB is all that's needed 4k, so the Fury is just fine
.... Oh get the 390x, it has more memory than the 980 / 980 Ti

Was having trouble getting it straight in my head let alone verbalizing but this article sums up a lot of what's been troubling me

http://arstechnica.com/gadgets/2015/07/the-r9-fury-is-amds-best-card-in-years-but-just-who-is-it-for/

LOL you blame Lisa Su?? Are you kidding? She's been CEO for less than a year. Good grief.

There was widespread speculation when she was named that she was being set up to take the fall.

 

[InvalidError]

One thing that has been bothering me with the latest releases is it seemed that AMD was competing w/ itself and sending mixed messages.

.... Oh 4 GB is all that's needed 4k, so the Fury is just fine
.... Oh get the 390x, it has more memory than the 980 / 980 Ti

Was having trouble getting it straight in my head let alone verbalizing but this article sums up a lot of what's been troubling me

The 390X does not quite compete with the 980Ti for performance, so AMD is promoting the extra memory despite the fact that the 390X lacks the memory bandwidth to make any meaningful use of the extra RAM. On the other hand, the Fury has a fair amount of extra memory bandwidth allowing it to make better use of the 4GB it has.

I'm guessing there are still tons of people who buy GPUs based primarily on marketing hype rather than benchmarks and reviews. Shout the benefit(s) of a single aspect of the product loud enough and it might just be enough to bag a few extra sales even if it has negligible real-world use in most circumstances.

 

[JackNaylorPE]

The 390X does not quite compete with the 980Ti for performance, so AMD is promoting the extra memory despite the fact that the 390X lacks the memory bandwidth to make any meaningful use of the extra RAM. On the other hand, the Fury has a fair amount of extra memory bandwidth allowing it to make better use of the 4GB it has.

I'm guessing there are still tons of people who buy GPUs based primarily on marketing hype rather than benchmarks and reviews. Shout the benefit(s) of a single aspect of the product loud enough and it might just be enough to bag a few extra sales even if it has negligible real-world use in most circumstances.

It doesn't quite compete with the 980 either when overclocked....the 390 doesn't compete with the 970 overclocked.

When the "tons of people" were arguing for 4 GB 680s over the 2 GB models, actual published testing could find no significant difference in performance

https://www.pugetsystems.com/labs/articles/Video-Card-Performance-2GB-vs-4GB-Memory-154/

When the "tons of people" were arguing for 4 GB 770s over the 2 GB models, actual published testing could find no significant difference in performance

http://alienbabeltech.com/main/gtx-770-4gb-vs-2gb-tested/3/

When the "tons of people" were arguing for 4 GB 960s over the 2 GB models, actual published testing could find a significant difference in performance

http://www.guru3d.com/articles_pages/gigabyte_geforce_gtx_960_g1_gaming_4gb_review,12.html

Even today, at 1080p and 1440p, we're not exactly seeing a significant hit, except in poorly done console ports.

I am not commenting on what is, I am pointing out the discrepancy in the message. Talking about 4k is hardly a relative topic as just 7 / 10,000 gamers have 4k monitors and AMD's financial position will depend on the other 99.93% of the market not the 0.07 using 4k.

As far as benchmarks and reviews, I don't feel that 99% of them put the user in a position to make any decision. How do you do a review that compares cards, w/o comparing the cards overclocked ? As web sites depend upon "hits" for their fortunes, the overriding editorial policy is to provide the "appearance of non bias".

What do most forum users do when they arrive at a breaking GFX card review .... skip to the last page (or the test pages) and if it doesn't compete well with the current leader in it's price category, you move on. No real reason to read each page which diminishes ad revenue.

One could argue that this is the product that's on the market so testing head to head at "stock" settings ... I can appreciate that position for some audiences. But if you consider yourself an enthusiast site, then ignoring the overclocking component is an unforgivable omission. Worse, overclocking the card in the title and comparing it against a list of competing cards at stock settings is inappropriate. In short, most reviews provide only partial information