Slashdot Mirror
Intel and AMD May Both Delay Next-Generation CPUs
MojoKid writes "AMD and Intel are both preparing to launch new CPU architectures between now and the end of the year, but rumors have surfaced that suggest the two companies may delay their product introductions, albeit for different reasons. Various unnamed PC manufacturers have apparently reported that Intel may push back the introduction of its Ivy Bridge processor from the end of 2011 to late Q1/early Q2 2012. Meanwhile, on the other side of the CPU pasture, there are rumors that AMD's Bulldozer might slip once again. Apparently AMD hasn't officially confirmed that it shipped its upcoming server-class Bulldozer products for revenue during August. This is possible, but seems somewhat unlikely. The CPU's anticipated launch date is close enough that the company should already know if it can launch the product."
People seem to be surprised by the delay and I have an exactly opposite reaction to that story. I remember when I was reviewing the first drafts of Buldozer (or actually Piledriver to be more specific) and I was surprised that the original date when it was planned to be released back than would have made it way ahead of the curve predicted by Gordon Moore. I was saying that it should be delayed some time so it actually is more accurate to the prediction and it has been delayed, however I will never know whether it had been done for that reason. The point is that in this industry there is something called "too good, too soon" which is not always desirable. Nevertheless, I hope both Bulldozer and Ivy Bridge will be available soon because they are both brilliant pieces of engineering.
Karma: Positive (probably because of superiour intellect)
Yes, bug K5 was a lot less complex. There's a lot more room for bugs in modern CPU designs.
There might be good reasons on both sides, but the tinfoil hatter in me believes this might have more to do with fact that both companies might want to see a little more profit out of the R&D that went into the current generation of products before obsoleting them. The performance of the current generation is high enough that it is getting harder to introduce a new generation at a price point that could both recover R&D and provide reasonable value for the customer.
Gee those delays mean the brand new shiny chips will just hapen to come out with Windows 8. Coincidence?
Not only can you finally ditch that aging Vista or XP machine, with shiny Windows 8 but now you can have a shiny new CPU too!
http://saveie6.com/
I thought the k5 never saw the light of day? K6 was the first AMD success.
Bulldozer might give Intel a run for the money and low to mid end notebook and sub notebook markets. Finally an integrated GPU that is not 7 years behind dedicated video cards. That is a plus for PC games and AMDs graphics are better than Intels.
http://saveie6.com/
The current situation is Intel is slaughtering AMD. AMD hasn't had an architecture update in a long, long time and it is hurting them. Clock for clock their current architecture is a bit behind the Core 2 series, which is now two full generations out of date. Their 6 core CPU does not keep up with Intel's 4 core i7-900 series CPU, even on apps that can actually use all 6 cores (which are rare). Then you take the i5/7-2000 series (Sandy Bridge) which are a good bit faster per clock than the old ones and there is just no comparison.
On top of that, Intel is a node ahead in terms of fabrication. All Sandy Bridge chips, and many older ones, are on 32nm. AMD is 45nm at best currently. Not only does that equal more performance but it equals lower heat for the performance, particularly for laptops. Then of course Intel is talking about Ivy Bridge, which is 22nm, another node ahead. Their 22nm plant is working and they've demonstrated test silicon so it will happen fairly soon.
The situation is not good for AMD. All they've got is the low end and that is getting squeezed hard by Intel too. They need a more efficient CPU and they need it badly. Delaying is not something they want to do, Bulldozer has been fraught with delays as it is. They've been talking about it for a long time, like since 2009, and delivered nothing.
They have every reason to want to get Bulldozer out as soon as possible and preferably before Ivy Bridge. Each generation that Intel releases that they don't have a response for just puts Intel that much farther ahead.
Now that said, Intel may well have decided to hold Ivy Bridge if AMD can't deliver Bulldozer because they don't need to. Sandy Bridge CPUs are just amazing performers, they don't need anything better on the market right now. However I can't imagine AMD colluding with Intel on this. They are not in a good situation.
we have known that ivy bridge will be released in 2012 since april... http://www.maximumpc.com/files/u69/sandy_bridge-e_roadmap_updated.jpg
The most naive question to ask if is this sort of delay is relevant to Moore's law and similar patterns. There are a variety of different forms of Moore's law. We've seem an apparent slowdown in the increase in clockspeed http://www.tomshardware.com/reviews/mother-cpu-charts-2005,1175.html. The original version of Moore's Law was about the number of transistors on a single integrated circuit and that's slowed down also. A lot of these metrics have slowed down.
But this isn't an example of that phenomenon. This appears to be due more to the usual economic hiccups and the lack of desire to release new chips during an economic downturn (although TFA does note that this is a change in strategy for Intel's normal approach to recessions.) This is not by itself a useful data point, so this is not further need to panic.
On a related note there's been a lot of improvement in the last few years simply by making algorithms more efficient. As was discussed on Slashdot last December http://science.slashdot.org/story/10/12/24/2327246/Progress-In-Algorithms-Beats-Moores-Law by a variety of benchmarks linear programming has become 40 million times more efficient in the last fifteen years and that only a factor 1000 or so is due to the better machines, with a factor of about 40,000 attributable to better algorithms. So even if Moore's law is toast, the rate of effective progress is still very high. Overall, I'm not worried.
'but, but, if you run benchmarks at 3840x2160 then the CPU is irrelevant'
it is. it is quite irrelevant. anyone who is spending money to get good performance on the mentioned resolution vicinity you speak about, is either an extreme enthusiast, a hobbyist, or a moron.
Read radical news here
Yeah. What is Rusty Foster even doing these days?
Democracy Now! - your daily, uncensored, corporate-free
A quick google would answer your own questions. K5 was released (was a competitor to the Intel Pentium & IBM/Cyrix 586), the K6 was up against PII and suffered from major thermal problems. While it did give Intel a little bit of a worry as far as sales go, the K6/K6 II's weren't exactly powerful. It wasn't until the Athlon (K7) that Intel shat themselves.
Success is a very loose term for a processor with major problems.
I remember when someone I knew (who thought he was great with computers but actually didn't know what he was doing) decided to overclock his K6.
It ran so hot that it fused to his motherboard before booting into windows.
any AM3+ motherboard will be able to support the next generation of CPUs alone is a good enough win in my book. As soon as the box is delivered I have to have have my computer upgraded in under 10 minutes
FTFY
No, but seriously, we went from AM2 to AM2+ to AM3 to AM3+, so compatibility is not that great. I just got my AM3 box last month (because going Intel would cost me about $100 more and perform a tad slower in Premiere) and I don't know if I'll be able to upgrade to Bulldozer since not all AM3 boards will work with AM3+ CPUs. An AM2+ Phenom II X4 920 owner will certainly have to buy a new mobo. Sure, it's a bit better than what Intel does, but you can only go one generation further and, frankly, I don't think it's worth it, since if you build you PC carefully, you won't really need a performance boost in such little time.
for something more useful than Starcraft. I agree you don't need 6 cores and 4G ram to read your e-mail, but today the workload of the average server or desktop, includes running virtual machines, virus scanners, full encryption, flash websites and whatnot. The laptop I was "given" 2 months ago has a brand new 4 core Intel, 4G ram, Nvidia quadro GFX and it's too slow to run my normal workload of terms, browser and VMs. Given the fact that I'm a contractor, spending a little extra on a faster CPU would probably pay itself back in less than a week for my employer. Sure, if they'd stop mandating W7 on the desktop with full encryption and on-access virus scanning, the world would be a better place for me, but most likely not for the company.
For servers, almost everything is running on VMs now. More power per CPU is very welcome there, since you can run faster/more VMs per box. The less heat you produce, the more servers you can put in a data center. Given the cost for real estate at prime interconnect sites, it's profitable to go green, even if you're not a tree hugging hippie.
I was promised a flying car. Where is my flying car?
I will continue to buy AMD. ive compared my sub $500 AMD rigs with comparable Intel rigs, I don't see why spending 2 to 4 times the amount of money for intel over AMD when AMD does a fine job. My Phenom II 945 has served me well, runs cool, runs fast, everything I put on it it takes like a champ. I have yet to stress out the Phenom. Ive run multiple games on it, audio and video work on it. The only 'advanced' thing I haven't done on it is CAD and seti@home. Why spend 2 or 3 times for the Intel, when all i'm buying is a name??? You intel fanbois go ahead, spend your money and feed the giant, duchebags
We're simply looking at an increasing gap in the demand for CPU power.
For most people (documents, spreadsheets, email, browsing) an i5 is more than powerful enough to satisfy the current demand.
OTOH there are specialized areas (visual effects, simulation,...) where even the current top-of-the-line models of Sandy Bridge Xeons reach their limit way too easily.
I thought 3dnow! appeared in K6-2?
that was a decent chip for a while(around 300mhz, for the price, they didn't oc as nicely as cellies though), the alternative was buying a celeron and overclocking it.
(disc, I got a pair of jeans with a k6-2 550mhz)
world was created 5 seconds before this post as it is.
I have to wonder how much of this is due to the stagnating economy in much of the developed nations. My recollection is that the last time the economy went south, all sorts of projects were either postponed, put on hold, or simply ended.
Nothing in the world is more dangerous than sincere ignorance and conscientious stupidity.
Oh but you could OC the living shit out of the Celery back then! With a good air cooler and a little luck a Celery could be OCed by a good 30% to 35%. Man in those days there really was a reason to replace your gear every 3 years or even less, the innovations on both sides of the aisle were just like two heavyweights duking it out....then Intel bribed all the OEMs and won through deceit what they couldn't when fairly, boo hiss.
Meh the big story as far as AMD goes isn't gonna be BD, frankly IMNSHO as a PC builder frankly PCs have been "good enough" on the CPU side for quite awhile which is one of the reasons my customers are quite happy I put my money where my mouth is and became an all AMD shop. No the really BIG STORY in 50 foot neon letters is gonna be the switch from VLIW to Vector Based GPUs, both in the discrete and in the APUs.
Imagine a chip that can not only crank out graphics like nobody's business but can actually work like a hyper FP on top of that with VERY little penalty. Today you are lucky if you get 1/4th the performance on double FP but with the new chips that will be cut down to less than half for the first bunch with the goal of native speed double FP in the next gen. I bet the math geeks are drooling at that prospect right now. And for those that switch to the APUs it'll mean an all new hybrid Crossfire where the APU can take over physics while handing off rendering to the discrete for some truly insane graphics. Sounds pretty damned sweet to me.
I just hope more geeks here at /. do as I do and give AMD some business. Unless you are one of those that literally slam your CPUs right to the bleeding edge (which admittedly there are more of those type here than on average) frankly the bang for the buck has been firmly in the AMD camp for some time now. you can get better quality boards with damned good IGPs for cheaper, The huge length of support time on the AM socket means you can go from dual to quad to six core without replacing anything but the CPU, the quads are dirt cheap right now, they just make really damned good, solid as a rock, long lasting systems.
I mean how can you not love a company that lets you get a fully loaded Black Edition dual kit for $200, a triple core kit for just $250 which BTW kicks ass as an HTPC, just swap the case for one of the "VCR style" cases and a dirt cheap 4xxx or 5xxx GPU, or a quad for $270? Oh and for those that have older machines I'd suggest a trip over to Starmicro where you can pick up cheap chips to upgrade older machines, both Intel and AMD. In my shop the Phenom X3 and the Pentium Ds are both quite popular upgrade paths for those with older AM2s or LGA775s respectively. Enjoy and go AMD!
ACs don't waste your time replying, your posts are never seen by me.
Yeah but there was a couple of dirty little secrets with the early Athlons. Number one thanks to NO thermal monitoring so they burned up, they burned up a hell of a lot and took boards with them. We used to keep a dead chip bucket at my old shop and that thing was damned near full of nothing but early Athlons. Number two it was trivially easy to rip people off by OCing an Athlon and selling it as a faster chip. Those early Athlons really didn't give you enough info either in the BIOS nor in the OS to really tell, most of the time you'd have to delve deep into the BIOS or pop off the HSF and check the chip. I don't know how many times I got told by customers "I wouldn't buy AMD, they are unstable!" only to tell them to bring in their unstable machine and sure enough, some douchebag had OC'ed it by a huge amount and sold them a chip for 30% or more higher than it was worth. Oh and to make matter worse the HSF on those early Athlons were kinda shit, which mean when they were OCed they were more likely to burn, but even without OC any dust buildup could fry the CPU and board..
We lost so many Athlons in that period that was one of the reasons I stayed with Intel plus Nvidia for so long. But when it came out that Intel was bribing OEMs and rigging their compiler (which they still are BTW, even though part of the settlement was that they end the practice. All they do now is put a little FYI buried in the bottom of the manual IIRC) and Bumpgate on the Nvidia side I bit the bullet and came back to AMD and boy not only I but my customers are quite happy as well. I'm an AMD only shop and the lower prices have made me quite popular thanks to the crazy bang for the buck and for the first time since the K5-2 I built my own AMD PC and love the thing to death. what's not to love, when I could build a quad core with 8Gb of RAM, and HD4850 GPU, 3Tb of HDD space, and Win 7 HP X64, all for less than $775? And today I could probably have knocked another $75-$100 off the price! Go AMD!
ACs don't waste your time replying, your posts are never seen by me.
45 nm is correct for the AMD Phenom II series. But the "Llano" APUs for low-end desktops are already in 32 nm. So you could say Intel is half a step ahead right now. Overall, however, I agree that AMD is under pressure and cannot afford artificial delays in their products.
What they still have are some niches where Intel has slacked off or does not compete for other reasons. The most important one right now are the APUs. AMD's Brazos platform does well on netbooks, and IMHO the LLano is a good choice for cheap consumer PCs.
Another one is (desktop) CPUs with support for ECC RAM:
Intel does not support that feature in its current "Core iX" CPUs at all, Presumably because they want to extract extra money from customers who need it by making them buy the much more expensive Xeons. Well, I'm a bit paranoid about reliability myself and thus I just ended up ordering a Phenom II instead of a (otherwise superior) Sandy Bridge CPU.
C - the footgun of programming languages
yeah my k6-2 which was stock 300mhz would only go to a bit under 400mhz oc'd, no amount of cooling helped, and the 300mhz celerons about that time did 450mhz fairly regularly, after that I went with a k7, and a duron then.
been rocking on intels for few years lately though.. because i've been content with work provided workstation laptops. they're beasts compared to what used to be usual, can even play new games well.
lately
world was created 5 seconds before this post as it is.
Geez, this brings back memories...
OCs, all the new chips, chipsets, MBs, from the early P5/P6s to the P4
But the market had its way, and it's cheaper to buy a new system (NB) nowadays. Not to mention speeds really haven't gone up as it used to be (not to mention clock speeds, but yeah the MHz is a myth)
We now do everything we need, except for the ultra needs of a few percent of the people. And you can always fire EC2 instances for sheer computing power.
how long until
Processor technology is at a state of gaining more by reducing die size than design. Because of SMP, sophisticated design changes are not needed to gain performance. Intel knows they can seriously move ahead of AMD if the next processor is successfully reduced to 22nm. They might as well wait. AMD isn't able to drop cash on reducing die size on every other release. If they can put off releasing their next processor at a point when they can afford smaller die production they will get a lot more out of it. Note: I'm not a fanboy of either company.
Having to work for a living is the root of all evil.
A couple years back I built a Socket 1366 rig - Core i7 920 overclocked to 3.8ghz, 6gb tri-channel RAM that if I felt the need could head up to 12gb relatively cheap these days, and a nice X58 Asus RoG motherboard with great features. It has basically carried me through and surpassed the Intel mainstream 1156 and compares easily to the Sandy Bridge 1155 i7-2600 etc. However, its finally getting to the end of its life - after an unheard of 3 years - and I'm trying to decide what to do next.
My first idea was to wait for Socket 2011, which is to Socket 1366 as Socket 1155 is to 1156. Its the next generation of the "Intel Enthusiast/Workstation" platform. Unfortunately, it appears that Intel is determined to put a stop to the very thing that made them successful with 1366 last time by making moronic cuts in features while increasing price. The $300 part is a lowly Quad core, with less cache. Then you have the option of a $600-700 Hex, with a bit more cache, but a locked multiplier, and finally you have to buy the Extreme to get the full package not just in speed, but in everything else as well. For those of us with Core i7 920/930 rigs it isn't worth buying a $300 processor that's marginally better than one from 3 years ago! Where are the offerings that, aside from speed, have all the features of Extreme without the pricetag? I can't justify this, and it seems a poor decision to launch with this crap, especially requiring a new X79 board and Quad-Channel RAM kit, if Ivy Bridge is right around the corner. Thus, I worry that Ivy, which stands to be a real leap forward, is going to be pushed back even farther after Intel realizes that SB-E on 2011 isn't making them as wealthy as they wished. I'd much rather them have brought Ivy out and just waited to keep 2011 until then, and follow it with the mainstream Ivy, much like the first-gen i7 1366 release.
It would be perfect for AMD to push Bulldozer and Piledriver into this gap, especially if they can put forth reasonable real-world performance, to give those enthusiasts annoyed by Intel's pricing another option. Sadly, I worry that Bulldozer is not up to the task as they've been waffling on it save for the business-level releases; by the time its out it will be near obsolete. I'd much rather they just release immediately with the 900-series chipset and then bring Piledriver out for the holidays or January. I'd really like to be done with Intel, but if there's such a huge gap between an OC'ed SB-E or IB, and Dozer/Pile as there has been in the past, I'm not sure if it will meet my needs. I can't spend money to "upgrade" to a AM3+ chip that will be less powerful than my current 1366 setup.
Again Intel seems to be swinging their weight around as they always do when they basically remain uncontested at the mid-grade and higher, while we all suffer. I love what AMD has done with their platform - great integrated, nice high-end chipset features, great prices, excellent integration with high-end PCI-E AMD GPUs, which are pretty awesome in their own right (I have a 6970 on my current system. An excellent investment, and the drivers/experience is better than Nvidia in many cases), but if the power gap continues to be so large, I don't see how I can justify what may barely be called an "upgrade". Here's hoping that AMD pulls out the stops and forces Intel to understand they're not the only game in town.
The K6 and previous had a pretty crappy FPU. And in the days of the Quake 1 and 2 engines, where 3d now was not in use yet, they made for a pretty shitty gameplay experience. No games? They were fine. But if you were in any way into 3d, they were crap.
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
Depends how you measure. I've recently gone from a Q6600 Core2 Quad, to an i7 2720 in my new macbook pro. Yes, desktop, to mobile CPU.
The i7 kicks the living shit out of my Core2 in transcoding video. Sure its a bit of a niche task, but CPUs have in general been "fast enough" for the day to day non-niche desktop crap since the original pentium, really.
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
The K6 did not include a built in FPU, like the NexGen - most PCs based on that platform came w/o it, and were low price. 3DNow was based on the assumption that the CPU had a powerful FPU and more advanced features, like SIMD. It debuted on the Athlon, which was done by Dirk Meyer's ex Alpha team. That was the first time AMD had a technological identity of its own, and not just an Intel copycat.
The entire approaches of the teams were different. NexGen gambled that most of the Windows market didn't care about FP performance, since Windows 3.11 wasn't used to do FP-intensive computations given Intel's own limitations, so they decided to maximize integer performance instead. The Athlon team, otoh, was the ex DEC team, and Alpha was always in a tough battle w/ PA-RISC and POWER for who was the leader in performance, so they came w/ the goal of giving AMD the fastest x86 processor.
As far as processor complexity goes, once you have a completely superscalar and superpipelined CPU and an ultimate performance uni-processor, it makes sense to then scale it to multi-processor units. Typically, for most workloads, one gets good performance enhancement for the first 4 CPUs, but after that, one gets into diminishing returns if one tries going into 8, 16, 64 processors and so on - unless it's a super-computing app w/ several CPUs working on different data sets locally. Similarly, a processor would also get into the realm of diminishing returns after it has a certain #registers, ALUs, and other such units.
Most of the current supercomputers are based on Opteron, which is an x86, or Intel's other x86 offerings. The only one who was offering an Itanium based supercomputer was SGI, but I'm not sure if they are any more. IBM's Watson is based on POWER7, while their Roadrunner is based on their PowerXCell8i as well as the Opteron.. But the ones from Cray are based on Opteron, and others on Intel's Xeon.
They may have been based on either DEC Alpha or HP PA-RISC, but those architectures do have to be current in order to be so used. Fujitsu's K-Computer uses the UltraSparc pretty heavily.
However, as far as servers go, you are right - high end RISCs are a third of the market. Wonder whether they'll gain marketshare as the reasons to upgrade x86s continue to get weaker.