Didn't Asetek announce that they scored a major deal that was due to ship H1 2015? Is Asetek leasing the design to Cooler Master, who is selling the parts to AMD?
http://asetek.com/press-room/news/2014/asetek-announces-largest-ever-design-win.aspx
Actually, it looks like that might not have been for AMD after all, if this is what the design win was...
http://www.datacenterdynamics.com/focus/archive/2015/02/asetek-sells-35m-liquid-cooling-california
Ya, I guess I'll just continue to follow hardware like I always do. Like you said, I don't need to upgrade (I am not gaming enough to even justify it.
AIO cooling for super computers and HPC servers? But I hear its not reliable... leaky and all that water around is dangerous!
Also, I can't imagine what Sapphire and EVGA get from AMD/NV for them to be exclusive partners of those brands. I would bet both would make some decent profits if they were brand agnostic. This is going a bit off-topic but I think if AIO CLC have a shot of becoming more mainstream with future generations, a longer warranty would give gamers who fear AIO CLC a better peace of mind.
I have to say HBM + 550mm2 rumoured die
nah, more like they are wining as a brand right now D: AMD needs to step up their PR game.
I sure hope so, and about time AMD starts making some big dies again for the GPUs!
It would be interesting to see how well 120mm AIO CLC and its VRMs compare when it comes to overclocking against the best air coolers for 390X.
You need a product out there that helps with the PR game first. The only news that people keep seeing is delay this, delay that, can't build 22nm, etc., etc., etc.
Yet in the meantime, Nvidia (who is suffering the same problem with 22nm), made lemonade from the lemons and released a 28nm part that cut power and improved performance. They saw the market and looked at the manufacturing problems and pulled a rabbit out of the hat to fill the void, ...
Makes me wonder why can't they do full cover block AIO? any reason other than cost that it couldn't or shouldn't be done?
R9 390X should be the 1st step of AMD realizing they need large die chips to compete with the best from NV. For next iterations starting 14nm, AMD can start focusing a lot more on drastic rework of the GCN architecture for perf/watt. This round they might bring 30-40% increase in perf/watt but it's not going to be 2X the perf/watt increase because it takes 3-4 years, if not 5, to design a brand new architecture from the ground up. GCN 1.3 won't be a built from the ground-up architecture like Maxwell. Because NV has more financial resources, they can do full architectural redesigns like that but AMD can't. This is why it's way harder for them to match NV in perf/watt and performance at the same time. Also explains why AMD needed to take a lot more risks with AIO CLC and HBM because they can't just match NV's 2x perf/watt with a 3.5 year old GCN architecture that's basically has been enhanced but not fully redesigned from the ground-up.
AMD could have charged a higher price for their GPUs (and hence higher margins), maintained ATI's "High end" image instead of looking like the bargain bin value brand, and wouldn't have to face the inevitable "spoiled consumer" backlash that they are facing now with a truly competitive high end GPU.
......
I can't help but look back and think of what a different position AMD might be in today if they charged on ahead with monolithic dies and let Nvidia take on water with their GPGPU strategy. Instead they gave Nvidia the room they needed to create then entrench themselves in the GPGPU market while being competitive in gaming performance.
Your thoughts?
small die strategy worked well when amd was fighting fermi...
now that nvidia catched up in perf/mm², the strategy just don't work anymore...
we will see the >500mm² again, from amd and nvidia...
So Nvidia played smart this round. They stripped the top of the line chip of its double precision capabilities (it's not completely neutered in the compute department though), which in turn probably saved a good amount of die space. That smaller die size translates into [theoretically] lower production costs, and overall a significant reduction in power consumption.
It's not semantics when it's $500 out of my pocket for next gen "mid-ranged" performance.
Or $600 for the 7970 prior to that goofy 680.
What it is is mid range cards occupying high end price brackets, what is it is a generation worth skipping, what it isn't is sematics as calling a fiero a high end car just because you can't get a lambo doesn't make it worth 2 million or high end.
I believe we would instead see longer times between process tech improvements. Oh, there will certainly be 1000$ [GPU] processors just after a factory starts production, but we will see longer concurrent production on older process tech for more years as moors law finally ends due to the limitations of physics.
Contrary to what the denizens of the microcosm on this board (and others, to be fair) believe, the rest of the world couldn't care less about benchmarks or any reviews. Most profits come from the OEM sectors and there it's what gets the job done for the least amount possible. AMD's small die strategy is probably the only thing that kept them afloat for this long. What's sinking them is their chronic mismanagement, lack of diversification, and poor support/PR/customer relations (which stems from chronic mismanagement I suppose).
That entire post is a gross misinterpretation of history, even though it seemed to be in good faith. Reread the RV770 and RV870 stories on Anandtech to refresh your perspective.
And a near 500mm², 250-300$ R9 290(X)?The end result was $500 GK104, $550 HD7970 and now $550 GM204.
As usual RS you are just hyping Maxwell and simply not telling the truth.
Perf/watt/sq mm or perf/watt normalized to die size has gone up 1.65/1.35 = 1.22 or just 22% higher. So this is why you should stop hyping Maxwell.
AMD R9 390X (550 sq mm) is likely to match or beat a GM200 (650 sq mm). AMD needs to keep making significant improvements on a yearly cadence (12 - 15 months) and return to what made them successful during the HD 4870 - HD 7970 days. They need to stop having large gaps of 18-21 months like they did between HD 7970 and R9 290X and now R9 290X and R9 390X.
Ironically AMD did this kind of refreshes better than Nvidia during the HD 4870 - HD 7970 days but forgot it completely after that.
AMD should also focus on notebook market as thats where they have fallen behind badly in market share. AMD should focus on perf/watt as thats the key to winning in notebook market apart from robust software (Optimus is better than Enduro and AMD need to correct that).
The current R9 3xx series presents a golden opportunity to AMD to gain back notebook market share and even leapfrog Nvidia over the next 12 - 18 months. Nvidia is not going to have HBM till late 2016 or early 2017. AMD should utilize this time to market advantage with HBM to gain market share in notebooks over the next 18 months. They should not get complacent and keep working on delivering a steady yearly improvement in perf/watt and perf/ sqmm.
But the biggest way in which AMD can and will gain market share in notebooks is APUs with HBM and competitive CPU performance (both single and multi thread). .....If they do that AMD can finally deliver a product which both Nvidia and Intel cannot deliver. Here is hoping for some Jim Keller magic in Zen :thumbsup:
I am not sure what will matter more in 2015-2017 games though -- the huge memory bandwidth of 390X or the 6GB of VRAM of the GM200. Even if they are 10% apart at launch, those 2 factors could have huge consequences later. Look at 680 2GB now, it's basically a write-off in modern games compared to an HD7970Ghz 3GB. In this case it won't be as clear cut since AMD would have a huge memory bandwidth advantage but less VRAM.
AMD will push for a 300 - 350 sq mm at Samsung 14LPP with atleast 25 - 35% higher performance than R9 390X and 8GB or 16 GB 2nd gen HBM. AMD need to follow up in a 2017 with a real tock. They need a significant microarchitectural improvement and look to hit 450 sq mm to be able to compete against Nvidia Pascal which is likely to ship the big die > 500 sq mm flagship GPU with 1 TB/s and 16 GB capacity.I am not sure how they are going to do it with 390X though. If they push 390X to near limits to try and beat GM200, I doubt it will be that easy to refresh it with 10-15% faster card in 2016 without relying on 14nm/16nm.
yeah I think GF could not deliver 2.5D packaging and Hynix was not ready with HBM any earlier than early Q1 2015. So now we will see products in Q2 2015.I am not sure if they forgot. I think some huge mismanagement happened where either a refresh of 290X failed in testing to show any significant improvements, or AMD simply took the risk of saving their limited financial resources/engineering resources to go all in with a 30-50% increase with HBM+WC. Perhaps they didn't anticipate HBM to be delayed for that long? That could be 1 factor why 390X is significantly behind 290X on the timeline.
I am confident of a top to bottom R9 3xx GPU stack with completely new ASICs built at GF 28SHP. I expect R9 390X, R9 390, R9 380X and R9 380 to be HBM based. I expect a 4096 sp Bermuda XT flagship GPU with total VRAM > 4GB ( 6GB HBM or 4GB HBM and 4GB GDDR5) and a 3072 sp Fiji XT with 4GB HBM.That's why I am having a really hard time believe that AMD will just do a mild-refresh of Hawaii as Grenada for 380/380X, and have flagship 390/390X but nothing much more power efficient. I am having a hard time believing AMD has just given up completely on the laptop discrete GPU market. If they haven't, we are ought to see something much better than R9 290M/295M.
AMD has been hit the most in enthusiast notebook GPUs where they are now non existent. HBM will help AMD gain market share at the high end where they need it the most. Fiji with 3072 sp and 4 GB HBM will easily beat GTX 980 both in desktop and notebook form factors. You think AMD will go with HBM on desktop and not use that advantage in notebooks. Wow you sure don't have an idea. AMD will leverage HBM in enthusiast notebook GPUs where they stand to gain the most.I don't think it's humanly possible. I bet the market share split is 80/20% in laptops, if not worse. If AMD managed to gain just 10% in the mobile dGPU sector in 18 months, that already would be remarkable. I also don't think HBM has sufficient volume to go into millions of low-end and mid-range mobile chips. If AMD actually manages to go HBM with laptops in 2015, I will be very impressed.
AMD APUs are a completely different ball game once HBM is incorporated. With DX12 and Mantle reducing the CPU bottleneck and AMD aiming to build a competitive CPU core with Zen its not required for AMD to match Intel IPC as long as they are close (perf gap of <= 10%). On the other hand GCN with HBM will murder Intel HD graphicsIsn't Zen slated for 2H of 2016? That's still a long way to go. Intel has an undisputed manufacturing node advantage for high-performance parts over anyone in the world today. Even if AMD could match Intel's IPC, I don't see them matching Intel in perf/watt or absolute performance. Also, as you remember even when Athlon XP+ and A64 whipped Intel's P3/P4s and P-D, OEMs and consumers still bought Intel CPUs primarily.
Bermuda - 4096sp (6 GB HBM or 4GB HBM and 4 GB GDDR5)
Fiji - 3072sp ( 4 GB HBM)
Treasure Island - 1536 sp (2GB or 3 GB GDDR5)
That entire post is a gross misinterpretation of history, even though it seemed to be in good faith. Reread the RV770 and RV870 stories on Anandtech to refresh your perspective.
First of all, AMD's small die strategy was a resounding success, the RV770 performing between the GTX 2x0 flagships for 100$ less than the slower variant! They priced NVIDIA out of the market at the time, who had a panic knee-jerk reaction with the GTX260-216. Read the launch article of the 4870 again.
The strategy was NOT to beat NVIDIA's big die, but to focus on time-to-market with the very best high-margin, higher-volume midrange chip they could build.
And for almost 3 generations it worked perfectly. AMD had the market cornered during the HD 4x00, 5x00 and part of the 6x00 generations.
NVIDIA had the performance crown yes, but you had to divvy up several hundred dollars for a mere 10-15% improvement and you got a big and hot chip.
In fact, the positioning today is more like a role inversal - the result of AMD failing to continue executing on this strategy and making the mistakes of NVIDIA of the past, and NVIDIA incorporating the strategy AMD pioneered.
* Since GTX-6x0, NVIDIA started releasing their mid & small die *before* their big die (historically this was never the case!).
The only lesson AMD had to learn, is that 2 smaller dies are not a replacement for one big die. Beyond that, I just hope they return to their former strategy.
You think AMD will fall behind in VRAM after leading for 3 generations ( GTX 580 1.5 GB vs HD 6970 2GB, GTX 680 2GB vs HD 7970 3 GB, GTX 780 Ti 3GB vs R9 290X 4GB). I don't think so.
You think AMD will go with HBM on desktop and not use that advantage in notebooks. Wow you sure don't have an idea. AMD will leverage HBM in enthusiast notebook GPUs where they stand to gain the most.
You are only looking at a short-term perspective of the consumer winning on price/performance temporarily. If the firm cannot sustain this strategy financially or gain substantial market share using it, it's a failure long-term.
This strategy ONLY works until your competitor catches up or beats you in perf/watt at which point their 550-600mm2 die chips and 400mm2 chips pummel your 325-350mm2 chips into the ground. Secondly, selling small- to mid-die chips against superior large die competing products means lower average selling prices, lower margins, and worse brand equity.
I am well aware of all the NV and AMD launches in the last 15 years. It was only a resounding success for the consumers but for the GPU market as a whole overall, it was a horrible outcome because it meant once NV caught up and exceeded AMD in perf/mm2 and perf/watt, they could charge flagship prices for mid-range chips = this all started with Kepler. Kepler proved AMD's small-die strategy was not sustainable long-term.
There is no way selling HD4870 for $299, HD5870/6970 for $369-379 produced high margins for AMD. NV sold 256-bit small die chips for $500 with GTX680, and $550 for GTX980. Those are high margins mid-range chips. Also, the small die strategy has no contingency plan. If you are late with a mid-range die, you lose the ENTIRE market until next generation. With the large die strategy, even if your large die chip is late, you still have solid low-end and mid-range chips.
And you seem to not get this point at all - the main reason AMD hang on to market share in the HD4000-6000 generations was because NV had much more inefficient perf/watt and perf/mm2 architectures. Once NV fixed that, it's game over for that AMD strategy.
-> If your large die chip is 500-600mm2+, then you can build mid-range 350-400mm2 chips. If AMD's flagship chip is not even 400mm2, how do they expect to even compete in the highest segments? They won't.
HD4000 vs. GTX280 generation
HD4870 = 256mm2
HD4890 = 282mm2
vs.
GTX280 = 576mm2
HD4870 was only as fast as NV's 3rd top card - GTX260 216.
AMD's small die strategy didn't work in terms of profits or large market share gains. They started off with 38% market share vs. 61% for NV, but market share collapsed as low as 31% and went up as high as 43.8%. By the beginning of HD5800 generation, AMD was at about 40% to 60% for AMD, or barely a 2% increase after releasing a $199 4850 and $299 4870. Ouch. :hmm:
HD5000 vs. GTX400 generation
HD5870 = 334mm2
vs.
GTX480 = 526mm2
HD5870 did beat GTX470 initially but with time, lack of geometry performance meant a GTX470 OC would crush a 5870. AMD started off that generation at about 40% and by the time Barts (6800) and Cayman (HD6900) came out, AMD had 38.8% market share. OK so going from HD4870 with 38% market share all the way until the end of HD5870 generation, AMD was still at just 38.8% market share. Are you starting to see a trend now? And all this time AMD is not making a lot of $ on GPUs.
During this generation, HD5850 significantly undercut NV's 470 and 5870 the GTX480. It took 9 months for GTX460 to come out to match a 5850 but AMD still lost market share. :thumbsdown: AMD basically beat NV by 6 months with GTX470/480 and 9 months with GTX460 and below. This huge head-start, and amazing price/performance didn't even make a dent in the market share by the end of the Fermi vs. HD5000/6000 generation; and neither did this make AMD a lot of money!
HD6000 vs. GTX500 generation
HD6970 = 389mm2
vs.
GTX580 = 520mm2
This is where AMD is starting to realize you can't have your cake and eat it too. You can't magically perform next generation DX11 effects like geometry/tessellation and global illumination for "free". You need to grow your die size with more functional units of you'll get run over.
Shockingly, despite HD6950 selling for $299 and unlocking to an HD6970, basically undercutting GTX570 and having the option of effectively almost buying dual 6950s for the price of a single 580 1.5GB, AMD doesn't gain market share.
At the start of that generation AMD was at 38.8% and by HD7970 launch, AMD is at 37.8%, another 1% loss of market share. :thumbsdown:
All this time a gamer could use HD4000/5000/6000 for mining which made those upgrades all free.
HD7000 vs. GTX600 generation
HD7970/7970Ghz = 352mm2
vs.
GTX680 = 294mm2
Now, the turning point. Finally what many have said would become true did. Once the competitor catches up or beat AMD on perf/watt and perf/mm2, AMD's small die strategy would come crashing down. AMD even beat GTX680 to launch by 2.5 months, and even after GTX680 came out, AMD recaptured the performance crown as of June 2012 and held on to it until the Titan. HD7770/7850/7870/7950 all beat every single Fermi GTX500 chip in perf/watt and VRAM and performance but still failed. NV easily coasted until GTX660Ti and below launched 6+ months later.
AMD started this generation with a market share of 37.8%, and it went down as low at 33% (!). By the time R9 290/290X launched in Q4 2013, AMD's market share fell to just 35%, or another 2.8% fall in market share from the end of the HD6970 generation. :thumbsdown:
Recall the green market share number of 38% - that's where AMD started off with HD4870, and by the end of HD7970Ghz generation, they ended with just 35%. Therefore, the small die strategy absolutely failed what it set out today -- it failed to make $ and it failed to gain market share. :\
You pretty much know today the market share is close to 70%/30%, which means AMD lost even more since R9 290/290X.
Data:
This is not reality. It only had the market cornered from a price/performance point of view but since most of the market didn't respond to price/performance, the strategy didn't work. I compiled market share numbers for you but I won't spend time compiling the financial because you know AMD's GPU division performed horribly for the last 4 years.
A lot of people must have paid extra for NV because AMD's price/performance & small die strategy did not result in market share conversion long-term. Those are the facts. NV's large monolithic die strategy prevailed, despite NV being late with both Fermi and Kepler.
NV's strategy is nothing like AMD's. AMD's focus is on GPGPU compute, OpenCL and small to mid-size die chips, using dual-chips for uber flagships. HD5870, 6970 and 7970 all had a massive amount of double precision compute, not just gaming performance. NV has tried all it can to make efficient gaming cards and has continued to neuter compute performance to maximize gaming efficiency. Maxwell just barely caught up to GCN in some areas of compute and that's only because it's 1 full generation ahead of GCN, but NV's DP ratio is now usually 16:1, 24:1 or even 32:1 vs. SP. Even their Kepler products had a fixed function compute unit which used way less power.
NV only continues to grow the die size of their flagship chips, and it's now able to use a mid-range chip as "flagship" for 2 consecutive generations because they have surpassed AMD in perf/mm2 and perf/watt. GK104 and GM204 are priced at $500 and $550 because gamers keep paying those prices for next-gen mid-range chips and because NV is now so far ahead of AMD in perf/watt and perf/mm2.
This was not the original AMD small die strategy. NV simply releases the smaller die chips first as flagships because of how much more ahead they have gotten in terms of efficiency. However, they have their high-end GK110/GM200 series of cards on top of the mid-range. AMD's mid-range strategy never had any large die chip sitting above; instead they used a dual-chip card to fill-in that niche.
These 2 sentences are almost contradictory. The whole point AMD should go back to large monolith dies (if financially possible) is exactly because you can't use 2 smaller dies to try and fight the uber-high-end market segment. Not only will AMD improve its brand equity, it would be able to raise prices, increase profit margins, and provide competition so that we are not being sold $550 mid-range next gen chips. That's why a 500mm2+ die is the best thing to happen to the GPU industry if true. It would mean having the old ATI-style mentality of wanting the single chip performance crown back.
Also, another consequence of having large die chips other than strong pricing competition, is we as gamers get a lot more performance. Imaging taking a perf/watt efficient Maxwell, Volta, Pascal architecture and extrapolating that performance to a 550-600mm2 chip. If both AMD and NV are onboard for superior perf/watt and perf/mm2, we will get huge performance increases in the next 5 years and if AMD executes on time, hopefully next gen mid-range chips won't be sold at $500-550. Although I think it's very likely that both AMD and NV will bifurcate a generation from now on since yields on newer nodes are going to get worse and well financially it's more beneficial for both of those firms to have 2 flagships every 2 years since it makes them more $.
Maybe it's going to become this new trend:
1st revision on a new node = 350-400mm2
2nd revision on that node = 500mm2
3rd revision of that node = 550-600mm2
NV is more likely to be gunning for a 500mm2 earlier on since they have a proven track record of great execution with large monolith die chips and AMD is only now supposedly making their first such chip!
Based on all the information I've read so far, I am inclined to say yes. I've never read any rumour of R9 390X card being anything other than a 4GB HBM product.
Right now we don't have any info to say if AMD will use HBM for laptops or not this round. I am going to say I just don't know than guess. My point was that I didn't believe AMD would having nothing better than R9 290M/295M Tonga but I don't know if they will continue with GDDR5 or HBM for 2015 high-end laptop GPUs. I don't have any info to support either case.
Fiji is coming with cooler master heatsink.
Anyone knows of a good cooler master heatsinks for GPUs?
Does heat-sink mean air (non-water) cooling?
65-86% depending on how you look at it. I don't think R9 390X will match GM200 in perf/watt but I personally don't care about 50-100W of extra power when discussing 400W+ overclocked i5/i7 rigs. Some people on our forum seem to care though.
It should matter to enthusiasts to some extent, if for nothing else other than HEADROOM. Hawaii OCs like crap and GK110 overclocks like a champ. If Fiji comes along with water cooling standard and uses 265 watts under load, neato you got a 60 C quiet card taking up 4 spots that can't fit into many SFX cases and I can almost guarantee it won't have much headroom left. AMD will already be squeezing out as much as they feasibly can with regards to yield.