Nah, that's for AMD. Talking about op margin for the underlying markets isn't really useful as this number would be unique per company.
I thought that slide was talking about the markets as a whole, not AMD's slice of them? (Didn't listen to the talks, so could be wrong)
They arent just bowing out of ultramobile. They are pretty much abandoning the entire cpu market to intel except for HEDT and servers, which is the only place I see for gpu-less Zen. Does anybody seriously think amd can keep refreshing the same APUs that are getting slaughtered already by intel, while intel brings out Broadwell, Skylake, and probably Cannonlake before any Zen apus come out, and still maintain any semblance of competitiveness. Honestly this roadmap is bad, bad news, almost worst possible case bad news, and it is still being spun into a positive.
I thought that slide was talking about the markets as a whole, not AMD's slice of them? (Didn't listen to the talks, so could be wrong)
Lisa Su was heavily pushing the "become a design house" strategy before being appointed CEO. While it may reduce their exposure to being compared directly to Intel by OEMs it won't magically fix the management and structural issues AMD has. Even then it looks like Intel is eyeing that space as conventional PC and laptop market appears to have reached a peak. The AMD slides show this retrenching into custom designs is still their main survival plan.
Really ?? so an 85W TDP Opteron will cost more to cool than a 95W TDP Intel XEON ???
Didnt know that
Nope, it's quite the opposite. The semi-custom area is now rated as a mild growth area. Big growth comes from GPUs and servers. Servers they come from a very low base, but I'm wondering whether they are serious about taking on Nvidia. They already proved that they can't build a healthy CPU business, but there is no reason to not be able to follow Nvidia's lead and build a healthy GPU business.
My understanding was AMD and others expect a good chunk of the server market growth to be in semi-custom designs. I don't see how their estimates of semi-custom business replacing lost revenue from their traditional sources would pencil out without including the server/cloud/datacenter market.
If those 85w Opterons are so good, then why we can't find them in data centers? Put it simple, Intel has amazing perf/watt and great perf density. So that is whu AMD has currently zero presence in the minds of decision makers.
If those 85w Opterons are so good, then why we can't find them in data centers? Put it simple, Intel has amazing perf/watt and great perf density. So that is whu AMD has currently zero presence in the minds of decision makers.
Putting Opteron in a datacenter is the quickest way to get yourself fired.
The answer is both, we are using the Intel Xeon class processors and a bunch of our infrastructure is using the Intel. Recently we switched to AMD Opterons, and as you wonder why, one of the things we are looking for is increasing the memory per host and the Opterons allow us in a single socket configuration to address more memory, and so all of the new deployments we are putting out there are carrying 64GB per host and the 8-core Opteron processors.
Probably because they're still at 32nm or do any of you deny that? Also Intel's process lead will shrink to zero sometime in 2016 when Zen launches, if 14nm is anything to go by then Cannonlake will not be the home run that everyone is expecting it to be & that gives AMD a good 6~12 months to try & rebuild some bridges in the server & semi-custom markets.Verizon is your only argument. And they did it solely for memory.
There is a reason why AMDs server share is essentially 0. And AMD will try to "re-enter" the segment with Zen.
Even AMD says your favourite CPU of today is a complete failure.
Probably because they're still at 32nm or do any of you deny that? Also Intel's process lead will shrink to zero sometime in 2016 when Zen launches, if 14nm is anything to go by then Cannonlake will not be the home run that everyone is expecting it to be & that gives AMD a good 6~12 months to try & rebuild some bridges in the server & semi-custom markets.
This is all water under the bridge, you're basically equating the worst phase in AMD's (recent) history to that of the start of Intel's domination of the server market. Hell even an Excavator based 28nm Opteron would give a good fight to the SB Xeons at the low end, perf/$ mostly since I can't extrapolate power usage based on pure conjecture, but that would require a tonne of (R&D) money for very little ROI.SB is also 32nm and killed of any hope for AMD they had left after Nehalem/Westmere.
Probably because they're still at 32nm or do any of you deny that? Also Intel's process lead will shrink to zero sometime in 2016 when Zen launches, if 14nm is anything to go by then Cannonlake will not be the home run that everyone is expecting it to be & that gives AMD a good 6~12 months to try & rebuild some bridges in the server & semi-custom markets.
Two things here, see my post above for reference, the perf/$ wasn't competitive when BD launched & perf/watt was abysmal. The fact that it was a delayed launch made it a losing battle for AMD, one that they'd never win.32mm certainly makes things worse for AMD, but it was SNB-EP that wiped out AMD server business, and AMD was on the same node back then. The node is just part of the issue, the gap in performance/watt is what matters, and even small differences on the server market can have devastating effects on the NPV of a given processor, and that's why SNB-EP mopped the floor with Bulldozer servers, even if the advantage at the time wasn't nowhere near of what it is today.
No it doesn't, especially not when Intel will be on 10nm (likely delayed as well) as perf/$ also matters, quite a lot in fact. If they can deliver ~80% of Xeon E7's perf at a similar TDP, probably ~90% as efficient, whilst costing only ~60% as much I believe they will have a winner with Zen. The Intel server margins are egregious & small & medium sized firms would like to avoid that price premium upfront as & when they upgrade their systems.
You think it's easy to maintain profitability when you're essentially paying double for hardware that could cost far less when bought from another competitor, AMD in this case ? The likes of FB, Google, Twitter, MS, Amazon don't care much about the cost of their hardware because they have an even higher margin (being software/services firm) than Intel & with the truckload of server farms they have efficiency is paramount to their cause, as for the rest they can certainly do with something ~10% less efficient but costing only two thirds
Really ?? so an 85W TDP Opteron will cost more to cool than a 95W TDP Intel XEON ???
Didnt know that
Without a performance and/or power context, perf/watt is only a marketing metric.
Example, We need a server that can produce 100fps
Server A performance = 100fps
Server A power = 200W
Server A perf/watt = 0.5
----------------
Server B performance = 80 fps
Server B power = 80W
Server B perf/watt = 1
Server B has double the perf/watt than Server A but server B doesnt fulfill the performance guidelines.
This is all water under the bridge, you're basically equating the worst phase in AMD's (recent) history to that of the start of Intel's domination of the server market. Hell even an Excavator based 28nm Opteron would give a good fight to the SB Xeons at the low end, perf/$ mostly since I can't extrapolate power usage based on pure conjecture, but that would require a tonne of (R&D) money for very little ROI.
If BD launched circa 2009/10 & then was improved upon till Excavator, say in 2013, it would've given AMD something to work with. The lack of (relative) performance as compared to the competition, SB at that time, is what killed it in the server space & the perf/watt goof up was the cherry on top.
So paying ~70% more for ~25% more perf, btw only ~12% more efficient, is good economics for you? Glad you don't advise small businesses cause that is an unmitigated disaster of epic proportions. Also I mentioned performance relative to a single core Xeon (obviously) & we don't know how many cores the top Zen SKU will be equipped with. Even after disregarding my other worldly predictions for the Zen there's no way any small or medium business is going to dole out even 50% more for a chip that performs only 25% better, that doesn't work & at least not in my part of the world.Actually that looks like a looser. The Xeon performs 25% better meaning you can get away with 25% less servers completely erasing any cost advantage on the CPU. The infrastructure cost (floor space, RAM, mobo, storage, software) compounded by the 11% higher power usage means you would be pretty dumb to go with the opteron unless there was some other reason.
Is that irrelevant for everyone that buys a Xeon or only those firms where the cooling costs exceed the hardware (server) costs? If that's a market AMD can exploit then that is where they can succeed & surely that must equal something like ~20% of the current market where Xeon operates.perf/$ is irrelevant, at least when the $ part is restricted to acquisition costs. That's why AMD didn't even bother to launch excavator on the server market. The CMT family was a real unmitigated failure on this one.
Is that irrelevant for everyone that buys a Xeon or only those firms where the cooling costs exceed the hardware (server) costs? If that's a market AMD can exploit then that is where they can succeed & surely that must equal something like ~20% of the current market where Xeon operates.
You better stop now while you can still walk away,
http://www.moorinsightsstrategy.com/whos-processors-are-used-in-the-giant-verizon-cloud/
So paying ~70% more for ~25% more perf, btw only ~12% more efficient, is good economics for you? Glad you don't advise small businesses cause that is an unmitigated disaster of epic proportions. Also I mentioned performance relative to a single core Xeon (obviously) & we don't know how many cores the top Zen SKU will be equipped with. Even after disregarding my other worldly predictions for the Zen there's no way any small or medium business is going to dole out even 50% more for a chip that performs only 25% better, that doesn't work & at least not in my part of the world.Is that irrelevant for everyone that buys a Xeon or only those firms where the cooling costs exceed the hardware (server) costs? If that's a market AMD can exploit then that is where they can succeed & surely that must equal something like ~20% of the current market where Xeon operates.
Maybe you should follow your own advice. That deal is dead, Seamicro no longer exists.