[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vip /vp / vr / w / wg / wsg / wsr / x / y ] [Search | Home]
4Archive logo
AMD continues to struggle financially: For...
If images are not shown try to refresh the page. If you like this website, please disable any AdBlock software!

You are currently reading a thread in /g/ - Technology

Thread replies: 58
Thread images: 4
File: fuck me.jpg (208 KB, 1087x773) Image search: [iqdb] [SauceNao] [Google]
fuck me.jpg
208 KB, 1087x773
http://www.anandtech.com/show/9976/amd-reports-q4-2015-results-and-fy-2015-earnings

https://www.techpowerup.com/219376/amd-reports-2015-fourth-quarter-and-annual-results.html

http://www.guru3d.com/news-story/amd-2015-fourth-quarter-and-annual-results.html

''Also they can't pay their old debts''
>>
Not bad considering net loss was almost $200 million Q1 2015. Obviously they're doing something right.

Hang in there you cunts, you still have to deliver Zen and Polaris.
>>
AMD can't fail or the market becomes monopolized.
>>
>>52531639
LOL, an x86 monopoly would be illegal.

The reason AMD can't fail is because intels balls would be ripped off. AMD owns x86_64 and if they go bankrupt intel won't be able to legally produce 64-bit x86 processors.

At worst intel will panic and lend AMD a couple hunded.
>>
>>52531615
But the debts of Q1 15, Q2 15, Q3 15, 2014...
>>
>>52531759
welp hopefully Zen will be as cash money as people and AMD themselves claim and bring AMD out of shit creek
>>
So AMD bankrupted Lehmen Brothers in 2008 then went cap in hand to the Abu Dhabi government less than a month later has still not sorted out it's finances? Shock horror.
>>
>>52531719
>LOL, an x86 monopoly would be illegal.
Via still make x86 chips, but they only sell them embedded now. Still about as competitive.

>AMD owns x86_64 and if they go bankrupt intel won't be able to legally produce 64-bit x86 processors.

Intel would just buy the patents behind x86_64 from the receivers. It wouldn't be an issue for them.
>>
>>52531472
I suppose that makes sense? R&D is very expensive; the console contracts would be handy but the margins are probably thin; their current line-up is workable but lacklustre (although reasonable and competitive at the low end); and their new stuff they've been working on that everyone's waiting to see to see whether it's really good or not, isn't out yet.

I'd expect times would be tough until then.

Just as long as they don't die and Intel goes full cackling monopoly.
>>
>>52532086
>Intel would just buy the patents behind x86_64 from the receivers. It wouldn't be an issue for them.
AMD could deny intel that chance. AMD literally invented x86_64.

Also AMD and intel have a cross licensing agreement where if AMD dies means intel wouldn't provide AMD their x86 license anymore so the x86_64 license would no longer have to be honored by AMD. Which means AMD can sell their x86_64 license to someone else or die with it as a final "fuck you" to intel.

Simply put AMD cannot and must not die for the sake of intel and everyone else.
>>
>>52531472
AMD has been in survival mode for 2-3 years leading up to the release of Zen bringing with it new chips for every segment. They were ALWAYS going to be bad years, always. The question was HOW bad, how little they could lose and if they could actually stay alive till Zen was out. As the company was when this decision was made, since then their performance has been IN THIS SITUATION, excellent. They reduced losses, they put R&D where it was needed, they brought in the right people to take charge of Zen design and GPU moving forward. They've won huge contracts with consoles which has kept their revenue going and have more semi custom wins that begin production this year.

People also need to remember that a huge proportion of their losses are 'book' losses. The company is valued at lets say 4billion one quarter and deemed to be worth 3.5billion a year later, that goes on the books as a half billion in losses but isn't a 'real' loss. They didn't spend 500billion more than they brought in. If after Zen the company value goes up to 4billion again then they gain that half billion back. They have made real losses, but again the majority of their losses listed aren't a big problem.

If Zen sucks this is a different matter, but any company who decides to effectively shut down the CPU line for 2-3 years before a huge different architecture is going to have a bad time. You can have a really bad time, losing 30% a quarter, or lose 5% a quarter. Losses weren't ever in question, how big they were was, they've been kept relatively small. 28% revenue loss, think about that, with no real new GPU lines and absolutely no new CPUs for years that is genuinely a good job.
>>
>>52532173
>AMD could deny intel that chance. AMD literally invented x86_64.

No they couldn't if they go bankrupt then the patents are in the hands of the recovers not AMD, AMD have no control over there assets at this point and a receiver has a legal duty to get the best financial deal for the company's debtors, Intel would easily be able to offer more than anyone else interested in the IP and the receivers would have to take that offer or face substantial litigation from both intel and AMD's debtors.
>>
>>52532260
It would never get to this point though. There will be a ton of larger companies and funds fighting it out for AMD if they ever do fall so far as to basically be finished. What they own is too valuable as a package.
>>
>>52532260
Not him but the cross licensing agreement does complicate things. intel is court ordered to provide AMD their x86 license and in return intel can use AMD's amd64 license.

IF AMD goes bankrupt, intel immediately loses their right to manufacture amd64 processors.

You can't just "buy" the amd64 patent, it's not that simple.
>>
>>52532237
The problem is they still have masses of debt, around $4 billion after the Abu Dhabi bailout, with continued losses quarter on quarter sooner or later one of the big investors will cut there losses and that will be that.
>>
>>52532329

It actually is just that simple, go look up similar cases it has happened before.
>>
who cares if they fail, intel can just invent a new instruction set.
>>
>>52531719
nah, Intel would probably create 128-bit architecture
>>
>>52532237
>If Zen sucks

Of course it will. AMD can't into CPU's anymore.

I bought a Phenom II 965 back into 2010; back in those days, AMD's CPUs were pretty awesome. The computer ran pretty much every game of that era like a champ (with a 5870 ATI card)

I've been using that CPU until just recently when I bought an I5 6500. I wanted to buy an AMD CPU, but the last one they released was back in 2012/2013 and even then it was aimed at multitasking more so than gaming.

From a business standpoint, they should probably give up on gaming CPUs ; of course, that will give Intel a monopoly, but so be it. They basically have a monopoly already.
>>
>>52532388
It would flop HARD. Just look at Itanium
>>
>>52532387
I care, I don't want pay my next i3 1000$ and my GTX 1260 1700$
>>
File: jkljklkj.jpg (71 KB, 415x502) Image search: [iqdb] [SauceNao] [Google]
jkljklkj.jpg
71 KB, 415x502
seeing how zen will offer 2013 haswell performance, it won't be the big release amd needs to gain decent market share.

intel already offers zen performance, since 2013, with haswell. intel already offers thread and core count that zen will offer, called haswell-e. intel will also be coming out with broadwell-e for the x99 platform this year. moar cores with stronger ipc.

intel already has zen beat with single threaded performance. its called broadwell and now skylake. intel will also be coming out with skylake refresh, kaby lake, this year.

all zen will do is keep amd fans happy and loyal. some intel users who still rock nehalem or locked sandy / ivy's might switch over, but most will hop back to intel.
>>
>>52531639
True :(
>>
>>52532399
>Of course it will. AMD can't into CPU's anymore.

Anything before Jim Keller came back is kind of irrelevant really. He has a proven track record of creating great architectures and worked on Zen for two years. It would be a genuine surprise if it's not at least good, even if it's unlikely that it's going to blow Intel the fuck out.
>>
>>52532760
Nice speculation based on your preferences, but if AMD can offer an octa-core chip with Haswell IPC or better at a competitive price, it will sell. Broadwell and Skylake are practically irrelevant in terms of single-threaded performance over Haswell. A 4790K at 4.6GHz has identical single-threaded performance to a 6700K at 4.4GHz.

It's a tiny, tiny jump covering two architecture changes, with Broadwell sitting in the middle (although the Broadwell chips currently on the market are such shit and limited so much by their TDP that they really can't compete even with a 4790K).
>>
>>52532760
nice memes faggot
>>
>>52531472
>Also they can't pay their old debts
lol

gonna start asking amd fags to pay debts now
>>
>>52531719
Intel would just stop making 64 bit processors. They would have monopoly, they can just churn out the cheapest processors they can make. Making quality is a net loss for a monopoly
>>
>>52532866
its not speculation

amd has admitted zen will be 40% faster than excavator, and if they are correct, that will match haswell.

so what if skylake is only slight faster than haswell? intel is faster. intel already has zens performance and they've had it since 2013.

if you don't care about core count, or just want a decent gaming rig, go skylake right now. it will be faster than zen. if you want core count now, go haswell-e. if you want faster performance than haswell, and MOAR CORES, wait for broadwell-e.

why wait for zen at all when intel already offers it, since 2013.

i don't see amd being hugely competitive because they won't be offering anything that's strongly competitive against intel outside of finally releasing a chip that matches intel's performance from 2013.

amd shills are shilling today.
>>
the stock is undervalued as fuck, I bought 1000 shares this morning
>>
>>52533080
I'm sure you'll make some massive profits anon
>>
>>52533108
I'm expecting returns to come in ~Q2 2017. if zen and Polaris are priced right it won't matter that intel/Nvidia have a marginally better product. plus amd has already secured a number of big contracts and the 32 core zen opteron could very well get them more server market share. I have high hopes for the next few years.
>>
>>52533053
just want to add:

why switch to zen if you're already on a 6 core haswell-e? you have the core count, you have the thread count. you have the ipc performance. why? you're on a platform that can accept broadwell-e once it releases. higher core count than haswell-e and higher ipc than haswell*/zen.

only reason for intel users to jump ship to zen in mass is if they want a change of face. yes some will switch because they're stuck on i3's, but all those users on unlocked quads and more so unlocked quads with ht? all those x99 users? haswell-e to zen is a side upgrade. same performance for same performance. haswell 4790k? why? you can have zen core count now with the same identical ipc of zen that you have now with haswell by switching to x99 with a 5830k.

unless amd does MASSIVE undercutting with zen, it won't be the athlon64 days. or even the athlon xp days in terms of market share.
>>
>>52533405
Because I'm still on a Phenom 2
>>
>>52533539
and yet right now if you switched to intel you will have your zen. but i guess you don't support the jewish intel?
>>
>>52532810
>>52533405
>everyone is shilling zen while i wait for the comfy K12
>>
>>52533657
maybe cuz the zen will be cheaper than the jewtel equiv????
>>
>there are people who want AMD to fail and create a horrible monopoly that would hurt everyone.
>>
There are so many people that think that just because Intel have "faster" processors than Hasswell no one will switch to Zen. What they fail to understand is that AMD are not targeting mainstream PC gamers with Zen, they are targeting data centers.

Large data centers want out of the Intel monopoly. Yest they have the performance required, but the prices of Xeons are too high. Bring in some Zen based Opterons with all of the power efficiency sauce that AMD has cooked up to deal with Bulldozers power levels and you have something potentially more power efficient with similar performance to what Intel is offering. So even if AMD only price their Opterons at 10-15% off of equiliven Xeons, AMD wins.
>>
>>52533854
Price adjusting can be done with a flick of a finger. (Intel side)
Performance adjusting with a flick of a finger is not possible. (AMD side)
>>
zen must be extremely price competitive to be successful against intel.

a 8 core with smt (hyperthreading) for $400, 6 core with smt for $250, and 4 core with smt for $180, followed by an arsenal of dual and quad core apu's in the low $100 to mid $200 range would severely hurt intel.

a 6 core zen with smt would literally be a amd branded 5830k in all respects, but for $250 instead of $370. a 6700k would be faster in all tasks up to 4 core usage, but once that 5 core, and higher usage is achieved, the zen and 5830k would come out winning.

and the 8 core zen with smt for $400 would mop the floor. intel's only viable solution would be a $1,000 chip, which would be identical in performance.

zen based apu's should take the crown in the mainstream market. dual cores and quads with smt and extremely powerful gpus would destroy intel in the budget gaming segment. and pair that apu up with a discrete radeon gpu and you can easily run them together in crossfire.

zen's release at that price point would really devastate intel's mainstream platform. a zen based "5830k" for $250 would hurt both the "i7 unlocked" and "i5 unlocked" processors. the ipc will be good enough, the core count well makes up for it by causing it to be drastically faster in 5+ threaded software, and smt tossed on is the cherry on top. only mainstream intel quad core processors with smt is the i7 line, and the only true overclockable based ones are in the mid $300 range. $350 for a processor thats only faster than zens $180 quad.
>>
File: jer01nov14BAZ.jpg (145 KB, 1500x1054) Image search: [iqdb] [SauceNao] [Google]
jer01nov14BAZ.jpg
145 KB, 1500x1054
>>52534437
where intel will be safe is their "-e" line. they already offer zen's core count and same exact ipc performance. intel can easily justify the price increase with the increased ipc performance with broadwell-e. haswell-e not so much outside of the "now" where you want zen "now" without waiting for zen.

personally i don't see amd being that flexible in pricing. i see their 8 core being more in the $450 - $500 range, the 6 core in the $300 - $350 range, and the quad being in the $200 - $250 range. it would still hurt intel's mainstream line, but not to the extent. amd wants to be seen as a "first class, leading innovator in semi conducting" to keep investors happy. higher prices, with a decent product, would do just that.

intel would hurt in the $300 range, but would give them a lot more wiggle room. they can easily drop the i7 unlocked with smt down to the $250 range, rename it to the i5, causing it to be the superior product compared to amd's quad based zen, and change all i3's to quads without smt starting with kaby lake. i7 will be exclusive to the enthusiast socket "2011" platform. having it take on amd's $350+ offerings.

it will be interesting to see how well broadwell-e overclocks. we know it will offer stronger ipc, but if it can also offer stronger overclocking capabilities than haswell-e, it can really hurt amd's zen platform since broadwell-e will be priced competitively against it.


zen success is going to come down to price point.
>>
>>52533854
Not to mention their APUs.

A Zen Based, HBM enabled APU running HSA complaint software will be a meaty thing.

I believe AMD already have plans for 16-core Zen based APUs for data centres do they not?

No doubt someone has the relevant images and slides to what I roughly describe, would you care to post them if you do?
>>
>>52534461
>dat pic
http://en.wikipedia.org/wiki/List_of_multinationals_with_research_and_development_centres_in_Israel

>List of multinationals with research and development centres in Israel
>Intel
>AMD
>Apple
>Samsung
>Cisco
>...etc

At least, NVIDIA aren't in Israel
>>
>>52534538
>16-core Zen based APUs for data centres
They will be rolling in cash if they pull that off. Server chips with more than 8 cores is where the real money is. Check out the pricing for Haswell-based Xeons. The release price for the highest end 18 core CPUs was over $7000.
>>
>>52531615
>Zen and Polaris
You mean disappointment and rebrands?
>>
>>52532930
>a new amd meme

kek
>>
>Advanced Micro Devices
>advanced
>>
>>52536907
They should change their name to
RND
Regressed Nano Devices.
Ironic because their RND budget is laughable.
>>
>>52534685
AMD only pulls Southbridge design off in Israel

Intel's actual CPU's are partially designed there.
>>
>>52533405
>why switch to zen if you're already on a 6 core haswell-e
>6 core haswell-e
>$400

Can't wait to build my my AMD/AMD build later this year, getting tired of this jew machine
>>
>>52534911
These chips are also very performance/W sensitive as they run continuously and in large numbers, AMD tend to struggle with that.
>>
>>52534928
>implying the relevant cards in the 400series will be rebrands
>implying fucking Zen will be a rebrand

If you buy a poorfag card you deserve a rebrand.
And Zen is all new. You don't tell everyone that all cards are on the table, and Fuck it up with old tech.
>>
>>52538545
Why do you even care so much that they have a factory in Israel or not? It's just R&D. Fuck off to /pol/ and give actual arguments next time
>>
>>52533854
And you seriously think that intel doesn't have the money to take in whatever loss they wanted to beath that? If it comes to that, they're just gonna do what nvidia does every so often when amd poses a slight threat, which is lower the price of their products. Intel CPUs are probably overpriced to hell and beyond because there's no competition. They can definitely lower it a lot before incurring any losses. If you seriously believe that at this point AMD can play this kind of game you're seriously delusional.
>>
what happened with bulldozer again
Wasn't it something about using automation and not human counted transistor or something
>>
>>52531719
>AMD owns x86_64 and if they go bankrupt intel won't be able to legally produce 64-bit x86 processors.

If AMD goes into receivership, Intel will buy all their patents at bargain basement prices so AMD can repay their debtors before Chapter 11.
>>
>>52539887
b-b-b-but we don't need human for the design. Automatic vhdl will create the architecture himself :3
Thread replies: 58
Thread images: 4
Thread DB ID: 448827



[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vip /vp / vr / w / wg / wsg / wsr / x / y] [Search | Home]

[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vip /vp / vr / w / wg / wsg / wsr / x / y] [Search | Home]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
This is a 4chan archive - all of the shown content originated from that site. This means that 4Archive shows their content, archived. If you need information for a Poster - contact them.
If a post contains personal/copyrighted/illegal content, then use the post's [Report] link! If a post is not removed within 24h contact me at [email protected] with the post's information.