TrulyUncouth
Senior member
Revenue. How about profit? Seems like every time we hear about Tegra it is losing money for Nvidia.
-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
AMD back into the red, post 20M$ loss.
Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Revenue. How about profit? Seems like every time we hear about Tegra it is losing money for Nvidia.
nVidia would have never won both console designs.
BTW: With GM107 and GM108 they have secured the low- to midrange notebook market. That is around 50% of their Geforce business - ~$250 million with a gross margins of ~40%.
But yeah $30 million look so much better than $100 million in operative income. D:
Last edited:
NTMBK
Lifer
Why would they not win both consoles? Tbh it feels like Nvidia got unlucky with timing- if the consoles were coming out in 2015 instead of 2013, they could offer Maxwell and Denver.
As for GPUs- yes, very good, but why not do both? A console would mostly reuse Tegra IP, just scaled up, and would be a damn sight more profitable than Tegra and Shield.
Last edited:
witeken
Diamond Member
No one is waiting for a 10 year update cycle for consoles.
ShintaiDK
Lifer
Would it? People focus so much on consoles, yet we already see the real world result. And the road only goes downhill from here.
And if you thought Jaguar cores was slow that limited the consoles to 30FPS in alot of titles. ARM would give what? 10FPS? 15FPS?
Roland00Address
Platinum Member
Considering AMD outlook, this is honestly really good numbers it could have been a lot worse.
AMD problems are problems they can't really control, they need a better factory process, they need faster memory bandwidth, and they need a more efficient architecture. Only the last one is something they can control, and the last one is very hard to do and takes years to develop.
It could have been a lot worse for AMD.
- They do not have competitive cpu speed in the desktop range.
- They do not have the right power consumption in the laptop and tablet range
- They do have great graphics, but the graphics are memory bandwidth limited. Furthermore due to cost savings many oems want to go with a single channel 4gb stick further increasing the memory bandwidth problem.
- It decreases AOSP and profits
- AMD has higher die sizes and thus increased the cpus are more expensive to produce
- AMD has to now pay a middleman to make their cpus, so some of the AOSP goes to a third party making it harder to compete on price.
AMD problems are problems they can't really control, they need a better factory process, they need faster memory bandwidth, and they need a more efficient architecture. Only the last one is something they can control, and the last one is very hard to do and takes years to develop.
It could have been a lot worse for AMD.
NTMBK
Lifer
The real world result being that the GPU division turned a profit 6 times larger than last year? :S
ShintaiDK
Lifer
Its is alot worse. Look beyond consoles, and AMD lost in all PC related segments. And consoles revenue and income will only keep declining. The best days are over there.
ShintaiDK
Lifer
Profit and revenue is already dropping in that segment. And it will only continue so in the future.
Consoles is not paying the bills. R&D is constantly being reduced. CPU division keeps shrinking, dGPUs struggles with both profitability and revenue.
We can already see AMD dropping one segment after the other for its CPUs. Soon they only sell netbook style CPUs.
NTMBK
Lifer
It fell QoQ because of seasonal difference. People buy lots of consumer electronics at Christmas. :eyeroll: Console R&D is done, barring shrinks, and will provide a useful revenue stream for years to come.
AMD has plenty of problems, but the consoles aren't one of them.
blackened23
Diamond Member
ARM SOCs have their place in mobile devices, but I don't really see any ARM SOC as having been viable for the consoles. On the flip side of the coin, i'm not trying to be obnoxious or anything about the console CPUs but....they're not great CPUs. Objectively, they're bad CPUs by any metric measured for 1080p gaming. I'm thinking that MS and Sony really wanted a 1 chip solution. Had they gone another route their costs would have been substantially higher with an NV/AMD graphics chip and intel CPU, certainly that would have given both a massive performance boost but at what cost. I don't know.
I really do think the situation with the consoles isn't ideal as a consumer. The CPUs are rather weak. Again, not trying to be facetious but it is basically true - these are the same Jaguar cores used in the recently released AM1 APUs right?
I think both MS and Sony went overboard in trying to shave costs, they weren't willing to take losses for years as they did with PS3 and the 360. I mean they could have used an AMD or NV graphics chip with an intel low power many core CPU. Would the end result be better? Yes. Heck, there are ULV core CPUs that are pretty darn powerful. Certainly more powerful than the Jaguar cores in the consoles. And more expensive of course. I don't see expensive as an issue, though, I mean the PS3 was obscenely expensive at launch. But still sold really well.
Really a no win situation. The consoles are relatively underpowered with especially the XB1 struggling to even do 1080p. And the PS4 isn't miles better. Oh well though. It is what it is. I would have preferred a different CPU regardless of what graphics were used. Clearly there are some issues with the consoles that prevent them from meeting their full potential. But, ARM SOCs weren't even in contention. Regardless of the AMD SOCs, they're still more powerful than the ARM SOCs of 2012/2013 or so. But it would have been so much better to get a 2 chip graphics and CPU solution for these consoles, even if they retained AMD graphics. They would be so much more powerful had sony and MS done that.
I'm thinking something like a GK104 mobile Kepler, or perhaps an AMD 7870 level GPU in addition to a ULV core i5 intel CPU would have made these next gen consoles much more beastly. But we have what we have. A weaker 1 chip SOC solution. And yeah ARM SOCs couldn't possibly be considered for these machines during design. I'm sure that will change in the future as ARM SOCs are really getting pretty powerful now. But not during the design of the consoles they weren't.
I really do think the situation with the consoles isn't ideal as a consumer. The CPUs are rather weak. Again, not trying to be facetious but it is basically true - these are the same Jaguar cores used in the recently released AM1 APUs right?
I think both MS and Sony went overboard in trying to shave costs, they weren't willing to take losses for years as they did with PS3 and the 360. I mean they could have used an AMD or NV graphics chip with an intel low power many core CPU. Would the end result be better? Yes. Heck, there are ULV core CPUs that are pretty darn powerful. Certainly more powerful than the Jaguar cores in the consoles. And more expensive of course. I don't see expensive as an issue, though, I mean the PS3 was obscenely expensive at launch. But still sold really well.
Really a no win situation. The consoles are relatively underpowered with especially the XB1 struggling to even do 1080p. And the PS4 isn't miles better. Oh well though. It is what it is. I would have preferred a different CPU regardless of what graphics were used. Clearly there are some issues with the consoles that prevent them from meeting their full potential. But, ARM SOCs weren't even in contention. Regardless of the AMD SOCs, they're still more powerful than the ARM SOCs of 2012/2013 or so. But it would have been so much better to get a 2 chip graphics and CPU solution for these consoles, even if they retained AMD graphics. They would be so much more powerful had sony and MS done that.
I'm thinking something like a GK104 mobile Kepler, or perhaps an AMD 7870 level GPU in addition to a ULV core i5 intel CPU would have made these next gen consoles much more beastly. But we have what we have. A weaker 1 chip SOC solution. And yeah ARM SOCs couldn't possibly be considered for these machines during design. I'm sure that will change in the future as ARM SOCs are really getting pretty powerful now. But not during the design of the consoles they weren't.
Last edited:
ShintaiDK
Lifer
It also fell YoY. CPU division dropped from 751M$ to 653M$. Cash reduced from 1183M$ to 982M$. Debt increased from 2044M$ to 2138M$. R&D budget lowered from 312M$ to 279M$.
Console R&D may be payed. But the money spend on it that was to be used elsewhere will punish the company for years ahead. Because the consoles got a very low margin and it will take several years to recover. And the best days of the console sales are over.
AtenRa
Lifer
R&D for the consoles was paid by Sony and MS. Thats why AMD doesnt bother that much about the low Console margins. They just manufacture APUs using a low cost Fab process.
monstercameron
Diamond Member
it will take a while before AMD properly stabilizes but thare on the up and up, your post seems overly negative and doesnt seem to take in the big picture, IMHO that is. Think about it, how much money were they losing in 2012-2013 and how much are they losing now?the losses, they too shall pass and then there is only profitability left.
Seems that most dont know basic accounting otherwise they would have noticed that :
So if they keep their quartely revenues constant they ll finish this year with about 550m gross profit, Rory Read management is quite efficient given the dire times for the PC industry.
AMD has stabilized in the short term. What we can expect from the company is small losses or profits, none of them meaningful. Intel won't put more competitive pressure on them until the arrival of 14nm chips in q115, which means they should have a stable share for the rest of the year. What will dictate the extent of these small losses or profits is console demand, if demand is high they should gain a little, if demands gets weak they will take losses.
But while the company is not in the same tailspin that they were from 2011 until mid-2013, medium and long term, the company has the same viability problem they had in the past, they must face increasingly bigger competitors with a decreasing R&D budget. Can AMD fight against Intel in the bottom x86 market and the leader ARM designers on the server market?
Yes, all those facts are real, however you're spinning them in usual negative way.
In the same quarter last year AMD have lost 146 mln dollars having also a 22 percent lower income - could you explain logically how it was better for them?
Last edited:
monstercameron
Diamond Member
I dont think anyone disputes that AMD will have a hard time going forward, it always has and they mostly deliver [plz stop bringing up bulldozer guys], they are innovative and still have possible avenues for growth.
/rant
I don't get why they just dont go ham on the apu idea, the a10-7850k is an awesome part, very performant but could be better. I mean forget the cpu AMD, double down on gpu and bandwidth ffs AMD!
I would like to see 300mm^2 apus with quad channel ddr3/4 or even gddr5. Like the original leaks, maybe on an enthusiast platform not unlike intel extreme stuff. The secrecy is kinda lame but I guess it is a symptom from before my time -over promising and under delivering.
What do you think about AMD avenues for growth?
They face heavy direct competition in the ARM server market, and indirect competition from Intel. Their cat line is checked, and their big core line is shrinking.
The avenue for growth would be... more 10% margin embedded chips? If you look at the cash flow profile, the numbers won't add up, you just can't fund bleeding edge products R&D with lagging edge products cash flow. That's not a question of the quality of AMD engineering (Brazos is proof of that), but a question of giving the engineers the bare minimum conditions for them to get the job done.
IMO, the make or break product is the ARM server line up. If they can make money out of that line, then they should grow. If that line fails, this will leave them with a heavily pressured cat line.
Just for the curiosity, how much would you be willing to pay for such processor?
frozentundra123456
Lifer
This seems like a pretty accurate analysis to me. The problem I see is that no matter how great one thinks the console wins are now, what happens when that revenue stream runs out. Will AMD win the next console design? This time around they were pretty much the only game in town, but by the time the next consoles come out, I see intel as having very competitive igps and ARM wil be much more competitive as well. If they dont get the next console, what will they do then? Certainly they are losing ground in basically every cpu area, and I dont see that changing unless they make a big jump in performance and performance per watt, considering eventually they will have to compete with much better intel processors.
Those are some good questions. On the low end x86 market, what does AMD have in it's pipeline for 14/16nm follow-ons to its 28nm Cat family and when will they be in production? I would think vendors would need to have some sense of this before committing to AMD centric product lines (why carry the overhead of AMD support for only one year).
I think AMD has done allot of upfront work to be a leader in the ARM server market (plus they have the advantage of their prior experience in x86 servers). AMD 'just' needs to deliver a good product. I don't tend to think that market will be very large or have the legs that a low power x86 market will likely have.
monstercameron
Diamond Member
its hard to say, maybe there professional gpu range, micro servers and/or servers, embeded and maybe dip their toes back in ultra mobile/tablets. All it takes is the right product at the right time.
lets say the theoretical fusion of a fx6300 and radeon r7-265... I would pay $250USD assuming it also has: hsa, huma, hq, 4k video and gddr5 or quad channel ddr3
NTMBK
Lifer
Console APUs are counted under the GPU division, not CPU, and is massively up YoY.
Do you know that you'd end up with cheaper platform costs if you added a dGPU for your processor, right?
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
-
-
