Slashdot Mirror
AMD Withdraws From High-Density Server Business
An anonymous reader sends word that AMD has pulled out of the market for high-density servers. "AMD has pulled out of the market for high-density servers, reversing a strategy it embarked on three years ago with its acquisition of SeaMicro. AMD delivered the news Thursday as it announced financial results for the quarter. Its revenue slumped 26 percent from this time last year to $1.03 billion, and its net loss increased to $180 million, the company said. AMD paid $334 million to buy SeaMicro, which developed a new type of high-density server aimed at large-scale cloud and Internet service providers."
Looks like they're focusing on ARM chips:
"AMD still sees growth potential in the server market, but not from selling complete systems. It's returned its focus to x86 chips and to the development of its first ARM server processor, code-named Seattle."
8 core 64 bit ARM chips with GPU built in are fairly common and 10 core chips already announced (Mediatek), with 16-48 core vaguely hinted at for servers by other vendors. So if AMD plan on entering the ARM processor market they'd better get something special out and fast, and be prepared to stick at it and upgrade it and take the initial losses. Because they're unlikely to win companies over first time till they're confident AMD are in it for the long run and won't leave them hanging without a supplier.
On the other hand they could focus on x86 chips where Intel is already deep discounting at the low end, and likely will have to do that all the way up the range to compete.
AMD face a tough time either way.
It looks like HP will have a more compelling product offering in the area of their high density product Moonshot. They are adding cartridges from a lot of different CPU vendors including Intel, AMD, TI (ARM CPU), and Applied Micro (ARM CPU). Their strategy of different servers for different work loads seems more well thought out. They also have the network experience and blade server experience when it comes to the chassis build.
So in closing... I don't think SeaMicro failed because of the concept, but simply the execution and product that was brought to market.
FTFY... in the future.
Sadly, I don't see an "out" for AMD. Their x86/amd64 chips don't perform as well as Intel's. The ARM market is saturated. They don't have their own foundry.
What does modern day AMD bring to the table that anyone wants? Even at cut-rate pricing, they've saturated their channels with chips and can't even manufacture and ship new inventory until the backlog clears.
It's a shame, but I think they're on their last legs. :(
I do not fail; I succeed at finding out what does not work.
AMD has not positioned itself even in a close second in any area of processing. ARM has flooded the market and its rather cheap and competitive so even if AMD was to garner some ARM business its not going to be lucrative. Desktops are not dead, but the ones surviving are low end stuff running minimal hardware. Not expensive gaming machines. Even Intel admits the low end spectrum of devices is increasing and that premium experiences are being required less. So many devices now are low powered, passive cooled and ridiculously chip to satisfy the device makers. I always wished Apple would have bought AMD/ATI and actually did some good chips for Mac's and even done some work on custom ARM CPU's. Alas, it never happened.
Actually AMD defined the 64 bit extensions to the 32 bit x86 architecture, and Intel had to follow and is letting the Itanic sink.
Of course Intel does not even remotely admit it (and even Linus ranted on this fact in a mail a decade or so ago), but they have still not come over the NIH syndrome that it caused them.
This said, amd64 (as Debian calls it) could and should have been better designed.
The trouble is that Intel really needs some competition, this will take years, but it may come now that they put MBA at their head. Their lead on process technology won't last forever once we hit fundamental physical limits. Also they have a tendency to forget that they are where they are thanks to IBM: if IBM had chosen another architecture 34 years ago, Intel would be an also ran. Unfortunately the rumor is that IBM selected the 8086 crapitecture (segemented addressing) because it was so weird that it would never come to compete with their own high-end proprietary products.
j/k, 'course I know what that is lol.
AMD has played a losing strategy for as long as I have can remember. It is sad, but I remember my first few PCs were all AMD machines. I bought AMD on principle, and because they were price/performance leaders. They were even outright leaders for a while, but failed to capitalise on that. I think, however, that the whole Sledgehammer/Clawhammer phase has ultimately ruined them. Obviously, those processors were streets ahead of the Intel offerings at the time, but it was always a long term losing strategy, in particular if they were depending on selling CPUs to make money. Their obsession with OEM deals also hurt them.
AMD could have done one of a few things, in my opinion, to reinvent themselves.
- They could have become a whole-hog PC builder, using their own chips and pricing their laptops and desktops accordingly.
- When Android happened, AMD, without as much baggage as Intel, could have produced an Android phone and Android tablets, and gone to market with that, using their chip making expertise to develop offerings that would have been more competitive than Qualcomm, Samsung etc.
AMD was obsessed with being a mini Intel, which was never going to work out for them.
AMD should have taken a page out of Apple's playbook. At best, they might be taken over by a Chinese company, otherwise they are doomed to irrelevance.
...we need AMD. Because if AMD goes away, Intel has zero competitors in the x86/64 market. Most people here probably aren't old enough to remember that CPU's used to cost an arm-and-a-leg in margins, and then when a bunch of hot shots like the 6502 came out, prices dropped literally over night. How could they drop so much? Because it was nothing but margin to begin with.
If AMD goes the way of the dodo bird, so do our cheap processors. Moreover, we'll likely lose a great deal of software freedom as what Intel says becomes law across the whole board. UEFI and TPM? Disneyland to what Intel can demand under the guise of "security" from every future computer.
They were even outright leaders for a while, but failed to capitalise on that.
Wow, that is the understatement of the century. AMD at one point did decide not to be a "mini Intel" and become a technology leader. Do you realize that while AMD had a far superior product for several years, Intel threw money (and threats - as was proved) to every retailer/integrator/etc out there to not carry AMD (and did other "interesting" things such as rig their industry standard compilers etc). Intel was allowed to use strong-arm tactics that "scream" anti-trust and after many years an almost bankrupt AMD was allowed to accept a small payment and Intel went scot-free.
If you have a product that is far ahead of the competition, you should be allowed to capitalize on that. If you are illegally not allowed by thepowerful players, there should be some sort of protection for that, before it is too late. But I guess the DoJ was sleeping at the wheel...
You have to remember, the Athlon was getting a firm lead on the P3 and Intel got out the P4 as a "response". The P4, the processor now universally known as the biggest "dog" by virtually everyone (even in its final and much, much improved incarnations), eventually abandoned even by intel to go back to a saner P3-derived architecture, was actually welcomed with laurels, both by (most of) the press and the integrators. AMD put all this R&D effort and they got nothing out of it, instead the were bleeding money for years, while Intel was making money with the current situation being a very weak AMD next to a behemoth. It is too bad for us, because the sole reason Intel CPUs are affordable is AMD - I won't remind you how much Intel charged per-CPU before there was competition. The sole reason Intel CPUs are this fast (or even that their consumer products are 64bit) is AMD. I only hope in some miracle for AMD to survive and get some competition going, otherwise there will be no-one left to keep Intel in check and consumers will pay for it...
So, yeah, the greatest industrial robbery of all time has been largely forgotten. AMD just "failed to capitalize", they were "obsessed with being a mini Intel"...
Violence is the last refuge of the incompetent. Polar Scope Align for iOS
Segmented addressing is actually a clever idea for 8-bit (8080/Z80) software compatibility, it's definitely easier that bank switching on 8-bit CPUs (2600 carts, mappers on NES carts, C64 and CoCo RAM/ROM swap, etc.) IBM probably designed the PC as a generic, but superior machine to run 8-bit CP/M software, and compete against Apple ][. Too bad they did not consider improving their '70s 5100 portable computer.
There were lots of practical reasons for IBM to use the 8086 in the form of the 8088. It had compatibility with the existing base of 8080 CP/M software, the 8 bit external bus could use 8080 peripherals and halved the memory granularity, and Intel was willing to allow alternate sources. The prime alternative was the 68000 which lacked an 8 bit external bus and was more expensive to produce.
uh there is a 68008.... yes an 8 bit external 68000
The IBM PC was released in 1981. An 8-bit bus 68K was not an option for the PC team.
They were even outright leaders for a while, but failed to capitalise on that.
Because if Intel's illegal business practices, for which they didn't receive any criminal sanctions. All they had to do was pay $1e9 to AMD, which is far less than they've profited by it.
SJW n. One who posts facts.
We've got Intel and AMD based servers running VMware in our datacenter and honestly - in the real world - I can't tell them apart.
The limiting factor on most of our boxes is memory and storage. Add enough to either chip company and you get similar performance.
The basic 1u 4S machines with 64 cores and 512G of RAM were denser than anything seamicro ever made.
The "dense server" companies were working on the myth that servers were still 1CPU in a 4U box. That stopped being the case years ago. Commodity stuff is already really dense.
SJW n. One who posts facts.
I am writing my own (multi-threaded) software and recently I had a chance to do a test run on an intel i7 processor (8-core, 2.67GHz) to compare it with my old Athlon II X4 (3GHz). Both programs compiled with the same version of GCC (4.6.1), both compiled with -O3 optimization. Running 8 threads on the Intel machine was only marginally faster than running 4 threads on the old Athlon. The threads were independent, so no threads were inactive while waiting for something else to finish.
Where Intel have the lead is in the compiler business. Back in 2003 or so they released their ICC 8.0 for free for Linux users. I was writing only single-threaded software at the time, and simply re-compiling it with ICC made it run about 5 times faster than the version compiled with GCC 2.96. And that was on a 2GHz Athlon XP.
What AMD have done right is the integration of the CPU and GPU allowing them to gobble up the console market. However, their bet that all developers will jump on the heterogeneous computing bandwagon did not pan out. But with HSA 1.0 coming up their lead will be too large and neither Nvidia not Intel will have a competitor ready for the next console refresh. All that Nvidia will do is to continue to pay game developers to optimize their engines for GeForce cards, and refuse to optimize for Radeon. AMD's resources are so limited that they will be forced to have a desktop version of their console processor, and maybe an ARM core for good measure.
Exiting the "dense server" are makes perfect sense, as the market is very limited. Running across many small cores is hard and developers will avoid it. It is the same story as taking advantage of the GPU, which also provides many simple cores.
So no, they are not dead, they are simply adapting to market realities and accept that they made a mistake when they jumped in the dense server bandwagon. Unlike Intel, who even now refuse to let go of the Itanium.
AMD could never capitalise on their lead for long enough because Intel ultimately had more money, could spend more on R&D and would eventually catch up to and surpass. They were also leaders in process technology - AMD never caught up with them in that department, and were able to squeeze out more performance from what was a worse architecture. Ultimately, the likes of Dell, although they might have, of their own volition, used AMD, were always going to be Intel shops. AMD was always one step away from disaster, and their obsession with getting OEM deals was a huge blind spot.
As I said, AMD were trying to be like Intel, only smaller. That strategy was never likely to work. If AMD had packaged their own kit, they could have developed a strong brand as the maker of reasonably powerful PCs, and at times the outright most powerful PCs. They could have achieved much better brand recognition that way, may have been able to produce bigger profits, and had more money to invest.
I don't disagree that Intel had illegal business practices. But as you also point out, it was better for Intel to strong arm its "partners" to not deal with AMD in the long run, and they bet on AMD continuing to take them on in a game they could not win.
If AMD, by making its own computers, had been able to get an additional (completely made up) $25 per PC sold, they might have been making a billion or so dollars extra a year, which would have been a big deal for them, and might have given them the revenue to compete with Intel.
It might even have helped them rationalise their product line. At some point ,they were manufacturing half a dozen or more processor varieties (Duron (or later Sempron) Athlon XP, Athlon 64, Opteron etc). They were also trying to manufacture for every price point that Intel was manufacturing for, which really allowed Intel to take advantage of their size. Intel had products for every price point that its partners wanted to sell product (Dell, HP, Acer, Gateway, IBM/Lenovo) and AMD had to try and compete with that. it was never going to work out.
Look at what Steve Jobs did at Apple. He got rid of the mentality of making products for every conceivable market segment and concentrated on the segments where he could make most money and that is what saved Apple.
AMD lacked a CEO with such a vision, and kept trying to catch up with Intel. Every decision they have made has meant they fell further behind. Some of the decisions were reasonable e.g. selling their foundry business (they were never going to compete with Intel on foundries) so they could control their losses, but they lacked the big decision that would affect how much money they actually could make.
Say it again!
It's curious they're having money problems since as I understand it they provide CPUs to both the XBOX One and the PS4. So that combined with the PC market is still not enough, huh?
While I agree in part, they have a few outs.
One thing I don't understand about the post is the "High-Density" bit. I am not sure if that is some techno babble for some super duper specialized server construct, however sever chips have been one of the few places AMD has excelled in the last number of years. Another place they do well is the budget segment where cost is more of an feature than actual processing speed. The difficulty with that segment is that the margins are likely very low, so you have to make up with a lot of volume, and you have newer technologies such as smartphones, tablets, etc... cutting into the market (i.e. why do I need a low end PC when I have other devices already that can do that role adequately).
Lastly, one of their saving graces was the purchase of ATI, and having a video arm of business. While I don't think it has worked out exactly as planned so far, the integration of CPU and GPU hasn't really given them the competitive advantage they were trying to get, there are still gains there to be made down the road should they stay the course and get some actual R&D done. Still they maintain pretty much a duopoly with nVIDIA as the only game in town for video technology.
That said I've never bought a AMD CPU. As an enthusiast it hasn't really made any sense since the Althon64 days (their last good chip). However, I have always gone with ATI/AMD video cards.
They should build to their strengths, and not try to be everything to everyone. They make the mistake to try and emulate and compete with Intel on everything. Intel can do R&D on a lot of stuff that ends up being complete flops and not worry as they have the market and the capitol to do it. The Atom for example. Not low enough power for embedded designs like ARM, and not powerful enough for anything worthwhile. A non-existent market so far as I am concerned. AMD did their own version of that... why? AMD dabbled in MB design, bet never really kept at it to really make anything of it (nforce chipset for example).
Anyway there aren't many companies that have not only one but TWO duopolies (CPU & GPU), so it is really hard to say they are out of options. Strategically they were positioned a bit less well than Intel to weather the emergence of newer technologies supported by embedded chips. However on the flip side of that coin, while it may have hurt their low end/low margin business, all those embedded devices now more numerous than ever before, all need to connect to infrastructure than is invariably run by servers, to which they do supply the lion's share of more profitable chips, so on balance shouldn't hurt as much as you might think... long term as those services tend to have upgrade cycles less than that of consumer products, but eventually they all need support.
So if they focus on server technology (which apparently they tried, failed, and are apparently giving up which seems short sighted), and expanding on their video technology, while pursuing R&D projects into core technologies (no pun intended! :) they should be fine. If they keep hemorrhaging capitol however they may eventually need to find some money somehow in the shorter term,.. However again with their position, investors should be not that hard to find, particularly if they show they have a solid strategic plan, and are not just flailing about in all directions trying to copy whatever Intel does... You're not going to beat Intel in being like Intel in short.
The Athalon64 days where their peak oil.
Not only did they have chips that were faster, lower powered, but also had 64bit support well before Intel. (granted way before any useful purpose of 64bit really)
Then Intel came out with Core 2 Duo, which beat them in every category, to which AMD had no answer, then Intel refined the design even better, to which AMD again had really nothing, and it has been that way ever since.
They should not have become a whole hog builder. Margins suck. They would have been a Chinese company by now should they have done that.
They could not have done the Android thing, they lack embedded designs.
I agree, their attempted emulation of Intel was doomed.
As for Apple, they sell software and a brand, neither of which AMD has. The only reason Apple has done as well as they have, is that they built their branding to be expensive (more less). They were also positioned well for innovations with the iPod/iPhone/iPad because their closed source software makes integration easy, and by locking more and more in, you expand you market by force (literately).
They should have done a lot of things. However if they wanted to be really underhanded and sneaky, they should have focused efforts server side, which all the afore mentioned devices require to be connected to, gain significant market share, then create some technology that locks say certain devices using said technology built by AMD into their supporting server structure which is now dominated everywhere. It could even be something say less underhanded, where it just works better, or faster, or whatever (i.e. not true lock in like apple) to promote their other products, and even licencing the tech for others to use, and make money on every other device using their licensing fees etc...
Then there is their whole video division. Another strategic decision could be to really take a stab at destroying nVIDIA, and once gone when you are the only game in town making video solutions, then repeat the above but leveraging your video technology which is now the dominate market factor, and again licencing etc...
Anyway they have some outs. But yes, they have to get out of the be like Intel mindset. You're not going to out Intel, Intel. At least not until they get some sort of advantage which they currently do not.
Remember that AMD sold GPU expertise to Qualcomm and has never really had low-power CPU expertise. The old mobile GPU org been doing very well since developing Adreno cores. AMD had little expertise left that could reasonably target Android power levels.
While Intel did pull a whole bunch of shady crap, AMD's downfall was not (entirely) Intel's doing.
AMD has been serially mismanaged for almost a decade now, and the blame falls squarely on the shoulders of "Business minded" execs plundering the company for bonuses and golden parachutes.
AMD's big fuck-ups:
Spinning off their fabs - This was done purely to jiggle and manipulate stocks so a few connected organizations could make a lot of money. Becoming abstracted from your process tech is a stupid idea when making the fastest processors demands being on the cutting edge of fabrication technology.
Saving money by firing their best processor engineers, and moving away from hand layout to save money - All those brilliant minds they brought in to make the athlon, and x86-64? Fired. Gone.
Acquisition of ATI - This has never worked well. Sure, it eventually led to design wins for the Xbone and PS4 but they aren't making a whole lot of money there.
AMD's greatest contribution in recent years has been a kick in the ass to Intel. Intel screwed up bad with the P4, and then again with Itanium. AMD was poised to eat their lunch and Intel was forced to respond with what was the one of the best processors ever made. The core2duo, and it's Xeon counterparts. The core2 was introduced in 2006. - Almost a decade ago and core2 based computers are still quite damn fast today. This message is being typed on a quad core variant and I don't see needing to replace it for another three or four years. This machine will run windows 10 without the slightest hitch (Granted I've updated the GPU because there are no modern drivers for the original, but the motherboard/processor are going strong!)
Intel has been /killing/ it since. Each iteration is faster, lower power, loaded with more features. Sandy bridge was a huge leap, and systems from that family will likely also be usable for a decade. I have an intel based /tablet/ that runs windows 8.1 and cost 150 bucks on sale.
I'm sorry to see AMD floundering but Intel offers faster processors for a reasonable price in /ALL/ market segments. AMD can't even compete on value, especially when you factor in power costs. (That 120 watt CPU doesn't run on good feelings or brand loyalty)
AMD chips run all three latest-gen gaming consoles...
"[Regarding the 'cloud,'] ownership was what made America different than Russia." -- Woz
Look up "Shadowplay" by nVidia. That is their software that uses the "nvenc" feature of their new GPUs. It has near zero CPU and GPU load, just load on the disk. All encoding is done by a special dedicated encoder on the chip. It's a fast encoder too, it can do 2560x1600@60fps.
The downside is it is not as good looking per bit as some of the software encoders (particularly X264) so if the target is something low bitrate you may wish to capture high bitrate and then reencode to a lower bitrate with other software later.
Bandicam also claims to support the hardware encoders of all the platform (Intel calls their QuickSync, AMD calls there's AMD APP).
What is your proposal, people should purchase AMD chips as a charity?
Nobody other than Intel zealots wants to see AMD go away. However if AMD's products are not competitive for what they want, why should they buy them? Trying to argue charity buying is a non-starter and a very bad strategy.
AMD has been really screwing up on their processors as of late. Their performance is not that good in most things and their performance per watt is even worse. So for a great many tasks, they are not a great choice. Their "APU" concept is an interesting one, but one who's time seems to be up as Intel's integrated graphics have been very good lately and getting better with each generation so "a CPU with good graphics" is likely to just be what we think of as a CPU.
If AMD wants more sales they have to make a product that is compelling in some way. As it stands, it isn't compelling in that many markets.
Sorry but you are having some selective memory. AMD actually was only a performance leader for a very brief period of time, that being the P4 days. That was also not because of anything great they did, but rather because the P4 ended up being a bad design because it did not scale as Intel thought it would. Outside of that they were competitive during the P3 days, but behind other than that.
They also had serious problems outside of any business practices from Intel. The three big ones that really screwed them today:
1) Their disastrous chipset situation. When the Athlons came out, their chipsets were garbage. The AMD made chipsets lacked any advanced features. The VIA chipsets were full featured, but poorly implemented. I bought an Athlon, excited at the performance upgrade I'd get from my P2 and drawn in by the price. I spent two weeks fighting and fighting to make it work, before finally finding out that GeForce graphics card were just incompatible with the boards because of VIA's out-of-spec AGP implementation. I sent it all back, got a P3 on an Intel chipset, and it all worked from the word go. Experiences like that really put many people and vendors off of AMD (combined with things like lacking a thermal halt on the chip so if a heatsink fell off the chip would bur out).
2) Their utter lack of innovation/resting on laurels. AMD took FOREVER to get out any kind of real new architecture, that being the Bulldozer, and it was poor when it happened. For too long they kept rehashing their same CPU architecture, while Intel kept moving theirs forward. This became particularly acute when the Sandy Bridge came out, which was a really good architecture improvement. Having nothing new and just trying to glom more cores on the server products was not a winning strategy long term.
3) Ignoring the software side of things. One of the things that makes Intel chips perform so well is their excellent compiler. It generates faster code than any other compiler, in every single test I've ever seen. That matters in the real world since people aren't going to waste time hand-optimizing assembly. Only recently did AMD get a compiler out (I haven't seen benchmarks on how good it is), for most of their life they just relied on other compilers and whined that the Intel compiler was mean to their chips. That has been a problem, particularly in research settings where people need high performance but are not primarily programmers and need something good at automatic code optimization.
AMD has done a lot to screw themselves over long periods and it has built up to a situation now where they are struggling in a big way. If you think Intel is all to blame you've your head in the sand.
The 68008 became available years (3?) after the 8088. I do not remember it even being planned or announced at the time.
I suspect Intel will find a way to keep AMD alive using pricing games to avoid both anti-trust accusations and the appearance that X86 cpu's are a dying biz. Intel and AMD need each other whether they like it or not.
Table-ized A.I.
"The core2 was introduced in 2006. - Almost a decade ago and core2 based computers are still quite damn fast today."
Agreed. My hands-me down i3-2100 is faster than the 3.2 C2D it replaced, yes, but I didn't fall off my chair. Anything more recent than a P4 will be usable for everyday tasks for most people.
As for AMD going down, I really don't want to go back to paying a thousand dollars for a CPU...
I've got better things to do tonight than die.
Intel will not get competition in x86 until Fabrication processes cannot be shrunk further due to laws physics.
right now they Intel dominates x86 market due to two reasons:
-they best chip Fab processes in the world; allowing for best power/performance rates.
-they have the fastest operating chips due to their R&D for the chip layouts
The barriers of entry are staggering to enter a Fab process; and for efficient Chip layout.
Also, its neigh impossible to get patent license for the x86/AMD64 instruction-sets either. I believe even if AMD were bough out; the buyer is not licensed to then produce x86 chips from what i recall.
I see WHY AMD went into ARM market, as it COULD potentially one day usurp x86 in the same way Intel replaced IBM's.... but that's some time off still. They likely were trying to move into a market without monopoly+antitrust abuse competition.
I remember when in the 90's you could pick an X86 Intel compatible CPU from not only AMD but NEC, CYRIX or IBM among some others.
Opening a "PC" without powering it on first was surprise back in the of clones.
All PS4 and XBones combined = less than 40 million units total, since fall of 2013.
Compare that to 130 million desktops, 200 million laptops annually. Not quite, but the latest gen consoles over 18 months represent about 10% of 12 months' worth of PC + Laptop sales.
Add to that, tablets, of which 15% are Intel powered (and this will trend upwards over the next three years) which sell about 200 million a year. I wouldn't imagine the console contracts are particularly valuable, as they had to under-bid Intel to get that.
moox. for a new generation.
Actually AMD defined the 64 bit extensions to the 32 bit x86 architecture, and Intel had to follow and is letting the Itanic sink.
Of course Intel does not even remotely admit it (and even Linus ranted on this fact in a mail a decade or so ago), but they have still not come over the NIH syndrome that it caused them.
That was the brief window that AMD had the upper hand. On the processor architecture side, Intel was splitting resources between hoping non-backward compatible Itanium was the future, and the disaster known as Netburst in the Pentium 4. Netburst had terrible performance, too high power consumption, but they were able to claim higher Ghz. All Intel had was name recognition, but technologically AMD was leading with the K7 and K8, and on this they were able to launch 64 bit. Even though it'd be the better part of 6 years before 64 bit really caught on, it was backwards compatible so it didn't cost anything to have but not use.
Intel only got ahead when a small group in Israel threw the Pentium 4 in the garbage, and started back at the Pentium III to design the Pentium M as an efficient mobile processor (which ushered in the confusing line of Centrino mobile platforms, that had Pentium M, Intel chipset, and Intel Wifi, not an actual Centrino processor). Intel eventually threw the Pentium 4 wholesale in the garbage, and expanded on Pentium M to form Core/ Core 2, and then from there i3/5/7 family. The rest is history.
I like to root for AMD as the underdog. When I bought my laptop 8 years ago, though perhaps not as good on battery, my AMD was best bang for the buck, and the low end graphics it had blew away Intel's junk i945. Sadly last year when I went to spec a new desktop, I wanted to want AMD, but Intel simply blew it away with performance. On normal desktop functions I feel single thread performance plays a major part in the CPU bottle neck. With a quad core, additional execution units are going to give marginal incremental improvements if you're not running a highly parallel workload. My i5-4690 stood out as the obvious choice.
AMD was riding high when Intel made a major mis-step with Pentium 4, but those days are long behind them, and they seem to be circling the drain. Spinning off, outsourcing, always behind on Fab technology, and a fraction of the R&D budget of Intel it's hard to compete.
You'd want to look at a 5960X, if you can afford it. Particularly when overclocked (and they OC well with good cooling) they are the unquestioned champs for that kind of thing. They have plenty of power to be able to run a game well, plus have cores left over for good quality encoding.
Are you kiddinig me?
Intel was overselling AMD with slower, pricier, power hungry chips 4 to 1.
Compaq refused to take AMDs chips FOR FREE.
R&D, my ass...
Ultimately, the likes of Dell, although they might have, of their own volition, used AMD, were always going to be Intel shops.
So, Intel was paying Dell essentially up to $1 Billion a year to not carry AMD just for fun? They were not going to go AMD anyway, even though they were so much faster, cooler and even cheaper?
Back in 2003-2004 we wanted to buy a few dozen servers for our lab at my University. My professor who had gotten the grant had gotten offers from various companies, Dell offering Xeon-based ones and others (HP and Sun I think?) offering Opteron-based. I was given remote access to a sample Dell server and a sample Opteron server (sorry I forget exactly what it was), and was asked to benchmark them. So, I benchmarked with the actual software we were using. The fact that the Xeons back then were 32bit while the Opterons were 64bit added to the speed difference and made a devastating difference. Sure, our natural language processing suite written in Perl was not getting much of the 64bit benefit, so it was "only" about 40% faster on the AMD server, but any of the computational-biology C code was over 2x faster. I gave my report and the Professor was going to go with an AMD server, when Dell came back and gave him a quote at less than half the price. They were actually selling each server less than what just the CPU was supposed to cost. I could not understand how they could do that. Well, now we know how.
Oh, to finish the story, the Professor told me he would go with Dell and buy more since they were so cheap and we would end up with more processing power. I warned him about the fact that the Dells were producing more heat and having even more of them would be a problem. It was hard to say no to such a good deal though, so the school bought the Dells. Then they had to wait for an upgraded electrical and air-conditioning system to be installed since the head/power requirements went up. It was a State University so it took almost a year. Also, the Dell file servers had a bunch of disks too close together and a dual Xeon throwing hot air over them, so a second disk would probably fail before the RAID array could restore from the previous disk failure. But most of those 100+ Dell servers are still running. Sure they are about as fast as a P4 that your knowledge-able friends made fun of even back then, but at least there is a LOT of them...
To return from the digression, my point when I started replying was that if the market was free, Intel would not have all that money and all those years (eternity for CPU manufacturing) to surpass AMD. AMD would have been the one selling and making money in the meantime, which would have meant more R&D for themselves to maintain a lead.
Anyway, let's hope we will never return to the thousand-dollar prices for desktop CPUs we were enjoying back in the good ol' Intel-only days.
Violence is the last refuge of the incompetent. Polar Scope Align for iOS
Part of it would depend on the relative OCs, of course. Also it would depend on if your encoder could use AVX2/FMA3 and if so, how much speedup it provides. For things that it matters on, there have been near 2X speed gains, but I don't know how applicable the instructions are to H.264 encoding.
Another option is if you can find an encoder you like that has a CUDA version, you could give it a video card to run on. However you'd want to check the implementation to make sure its quality is comparable. Also you might need to get a video card that has better double precision performance, as I'm given to understand single precision math isn't enough for top quality H.264 encoding. So like a GTX 480 or a normal Titan, the newer GPUs generally have less DP cores (to keep power/heat down).
Only applies if the encoder you want has CUDA support, of course, and if it knows how to use DP math.
No true. AMDs biggest contribution was making a 64 bit x86 processor that could run existing 32 bit code, something that Intel's Itanium could not do. Eventually, Intel ditched Itanium and even long before that followed AMD on the 32/64 bit combo. AMD is still hampered by the reputation of being a clone shop, as clearly shown in your post. Nobody gets fired for buying Intel and even for OEMs Intel has a much bigger pull on consumers than AMD has, yet AMD is across the board more affordable providing the same processing power. I do agree that AMD should have done more with the ATI acquisition, but shedding the chip making side of business was the right thing to do. An AMD owned fab would be hard pressed to find additional business from what might be competitors. Spinning this business off into Global Foundries was the right thing to do. The Middle Eastern investors are also a logical partner, right now they have a lot of cash available from their oil based economy, but they know that once the wells run dry they can't live off the desert. Investing the money in a diversity of industries is their plan for the future. Global Foundries signed production contracts with many companies and they operate factories in the US (upstate NY) and Europe (Dresden, Germany). Especially Dresden is a good choice as it was the leading center of microelectronics design and research even as far back as the 70s and 80s in East Germany. AMD's biggest challenge is the undeserved bad rap they continue to get. If I had money I would buy AMD shares...and Intel, and Broadcom, and and and and....
I ran across this link to a transcript about the history of the 68000 after my post here. There is a discussion thread on the Real World Technology forums about why IBM chose the 8088 where this transcript was linked:
http://archive.computerhistory...
http://www.realworldtech.com/f...
The additional cost of the 68000 over the 8088 was an even larger factor than I remember.