Slashdot Log In
AMD Says Power Efficiency Still Key
Posted by
ScuttleMonkey
on Sat Sep 02, 2006 06:05 PM
from the teaser-trailers dept.
from the teaser-trailers dept.
Larsonist writes to tell us that even though AMD's new architecture wont be released until mid-2007 they are still letting people in on what some of the new features will be. From the article: "While clock speeds have not been revealed, each of the four cores will integrate 64 KB L1 Cache and 512 KB L2 cache. The native quad-core architecture will also include a 2 MB shared L3 cache, which may increase in capacity over time. The processor will have a total of four Hypertransport links - up from three today - that provide a total bandwidth to outside devices of 5.2 GB/s. AMD is also thinking about integrating support for FB-DIMMs 'when appropriate.'"
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
Almost obligatory statement... (Score:5, Interesting)
Is Vista going to support 4 cores, or like XP Pro and 2k, limit it to 2 "cpus" so they can charge more for the server version?
Re:Almost obligatory statement... (Score:5, Informative)
In other words, MS counts sockets, not cores.
Parent
Re:Almost obligatory statement... (Score:4, Interesting)
Parent
Re:Almost obligatory statement... (Score:5, Informative)
So, let's look at the two markets seperately.
Desktop, the users are likely to not care too much, provided the "Per core" cost is low enough. When we start seeing 4/8/16 core CPUs, a $10 per core fee will add up quick, but most home users will be using OEM copies and won't see that cost. Most business will have site licenses and won't care. But, some home users will, and some businesses will care, and they'll seriously consider alternatives (Maybe, maybe not switching). Microsoft would much rather "lose" money by not fleecing people, than have them even CONSIDER switching, so management's going to ditch the idea for desktops and workstations.
Server market... they need any advantage they can get. They main competition is Solaris, BSD, and Linux. Linux and BSD are *free*, and Solaris has a bunch of good features which are pretty much Solaris only, even still. Charging per core would be suicide in this market, too.
So, what market would charging per core be a good idea for Microsoft? None. Say what you will about their software writing abilities, but nobody should doubt their marketing prowess.
Parent
Re: (Score:2)
At this point, they are the de facto standard, so they don't need good marketing.
Re: (Score:3, Interesting)
1 CPU = most laptops and desktops, low end servers
2 CPU = high-end workstations, average servers
4 CPU = high-end servers
As the number of cores ramp up, as you said to 4/8/16, then charging per core would be like charging per GHz or per L2 cache size - it doesn't make sense, adding cores will just
Re:Almost obligatory statement... (Score:5, Informative)
I work for a very LARGE bank. I guarentee you we have more boxes running MS Server 2003 than all others combined.
We have some HPUX, IBM, and SUN sprinkled in there but several of the vendor apps I've worked with lately has dropped all support
for SunOS and is now requiring Windows Server for their apps.
We use to run Novel for our file servers but that was dropped for MS Active Directory.
The next leading server OS we use is probably zOS. No Linux yet but I know the powers-that-be are looking at it.
(BTW, I'm no MS fanboy, I'm just making a point.)
I'm sure there are many industries in which MS does not have the majority of the server market but
large financial groups are not among them.
Parent
Re: (Score:3, Funny)
Is that because you need more Server 2k3 boxes to get the job done?
Re:Almost obligatory statement... (Score:4, Interesting)
Real reason: politics. One of the problems with server apps is that they tend to be "critical" to normal operations of the business.
This means if you tell someone, "Hey, your server is under-utilized so were going to put this other groups app on there...",
the shit is definitely going to start flying. (ex. "that server came out of my budget..." etc.)
Ask anyone where I work (or probably where anyone works) and they would say that it would be the end of the world to have to share a server.
In truth, it's all too easy for someone to mess up and bring a server to it's knees.
The group I work in, we have probably 80 servers (yes, they all run MS), and none of them can be shared because it violates the SLA we have with Cisco. They won't support the app if anything else is running on the server.
(Cant say I entirely blame them.)
They solution to this problem is virtualization; I've been telling my co-workers that the future of servers is virtualization.
Until we can really make it work politically, we'll be running racks and racks of servers that mostly run 90+ percent idle.
Parent
Re: (Score:3, Insightful)
But _that_ is not politics, it's because Cisco understands windows servers and how badly they handle more than one task.
Re:Almost obligatory statement... (Score:5, Informative)
I was giving an example to counter the GPs claim that MS doesn't have respectability in the server market.
Not only do they have respectability, I see more and more vendors requiring them and most custom apps we write is on MS (desktop and server), although a lot of our web stuff is IBM (websphere, etc).
I work in telecommunications; Aspect, Avaya, Cisco (not cisco routers/switches) have ALL migrated from other OSs to MS in the past 5 years for their server and desktop apps.
Yes, Novel, Sun, HP, and IBM are huge in the server business. Except for IBM, they will steadily lose share to MS over time.
One of the main reasons this hasn't happened faster (like in the desktop arena) is because there are a LOT of home-grown apps that companies have built over the past 15 years that still require a particular OS to use.
At the bank I work for, we have a home-grown desktop app that is "critical". It was originally written for OS2. Around 1997 the team was forced to port it to MS windows NT.
The conversion was brutal and it took them forever to get all the bugs worked out.
Cut to 9 years later. WERE STILL USING THAT APP! (and no one is happy about it.)
They've tried 4 times in the past 5 years to either have it rewritten or replaced with a vendor product, but every project has failed.
It's so bloated and patched it's a nightmare to maintain, but even tougher to replace.
Now apply that story to the thousands of custom server apps in thousands of companies around the world that only run on SunOS, HPUX, Novel, etc.
The number of custom web apps alone that have been written on Sun servers could keep that company going another 20 years.
We run a lot of MS server OS at work because we've been busy migrating stuff over the past 5 years. (a LOT of stuff.)
So, my point is, MS Server OSs are not only respectable, but many large companies are actively moving towards them because they see it as streamlining. (Large companies like to be consistent and stick to "industry standards", even if it doesn't make good business sense.)
MS knows it's going to take time; it's not like replacing a desktop OS.
Parent
Re: (Score:2)
Re: (Score:2, Funny)
Re:Almost obligatory statement... (Score:4, Insightful)
For example, which is worth more (none of these CPUs exist) - a quad CORE 2 or a eight core Athlon XP?
I'd expect the eight-way to be better at some tasks that are easy to multi-thread, perhaps database servers. I'd expect the quad CORE 2 to be better at most general tasks, mostly because current OS and compilers aren't designed for multiple cores.
Parent
Re:Almost obligatory statement... (Score:4, Interesting)
I'm not sure if Oracle still does this, but they used to have almost voodoo math to figure out how much you owe oracle. It was something like X/CPU, then that value multiplied by a scale for the type of CPU (at the time RISC vs CISC), and then another multiplier by the amount of RAM on the box.
When I heard that, I always suggested under specing the box and then silently upgrading it after the Oracle guys left. I believe the technical term is sliding scale which means the more you can afford, the more you will pay.
Parent
Re: (Score:3, Interesting)
Most SMP code is tested on CPUs of equal clock speeds so odds are this is going to bring out all sorts of fun race conditions in Vista, Linux and *BSD and I'm personally not so sure I'm going to touch this until the resulting dust settles.
I'm not saying it's a bad idea.. it looks like a good one but this will take time for the software to mature.
it's better actually (Score:4, Informative)
As for race conditions: that is pretty well taken care of already. SGI has Linux on a 2048-way system now.
Parent
Re: (Score:3, Interesting)
Re: (Score:3, Informative)
Different clock speeds should not present much of a problem. In terms of performance it should be a non-issue - right now you can get variable performance out of the same code depending on other factors like memory contention, cache pollution by other processes, etc so if a cycle takes 1.00ns or 0.50ns isn't going to be anything new.
Only the scheduler is going to care about frequency differences, and considering that we already have the abili
Re:Almost obligatory statement... (Score:4, Informative)
You're kidding right? I can't imagine any software which depends on the timing of coperative CPUs.. MPI and general divide-and-qonquer work-clusters could care less about the performance level of peer threads/co-processes. Hell process interrupts due to pre-emptive multi-tasking is enough to guarantee lack of symmetry.
Now perhaps you're referring to scheduling problems in the kernel.. I'm sure that AMD would be generous enough to provide kernel patches as are necessary.
Parent
Re:Almost obligatory statement... (Score:4, Informative)
I find that the two processors in my dual-core Athlon X2 run at slightly different speeds (according to AMD, this is expected). That in fact did cause the Linux kernel some problems, since it was trying to balance handling the interrupts between the two. The problem happened when the timer interrupt bounced between the two, as within an hour or two of startup their tick counts became significantly separated. This made the system clock start running forward at a rapid rate. A Linux patch fixed this issue. So I can definitively say that Linux does run on SMP cores at different speeds.
I'm not sure how Windows will do it, but they'll probably figure it out if they haven't yet. The real challenge is a new scheduling algorithm for variable CPU capabilities (although we do have that to some extent with frequency scaling on single CPUs).
Parent
Re: (Score:3, Informative)
Re:Almost obligatory statement... (Score:5, Informative)
The Symmetric in SMP refers to the fact that each process can run the same tasks.In an assymmetric setup, there may be a processor dedicated to the kernel or other tasks.
It most definitely has nothing to do with speed.
Parent
Re: (Score:3, Informative)
Spin-locks or not, cores running at different clockspeeds aren't going to expose any more race conditions than regular usage. Even on a current SMP system, the processors wil
Power control at the per core level (Score:5, Informative)
Re: (Score:2)
Re:Power control at the per core level (Score:5, Informative)
My cpu speeds idle at around 1GHz, and then spring back up to 2.4GHz when under load, then slowly drop back to 1GHz when not in use. Independently, I might add.
Parent
you don't want to do that... (Score:3, Insightful)
But you don't want to do it anyway. If you look at the voltages to the cores, as the cores slow down, the voltage goes down too. This reduces power used and also reduces leakage (which is large at 90nm and can be large at 65nm). The problem is that both cores receive the same voltage, so you can't reduce the voltage to the one core that is slowing down
Power isn't the only criterion (Score:3, Insightful)
Re: (Score:3, Interesting)
Exactly. Intel calls this EPI throttling in one of their recent papers.
Re:Power control at the per core level (Score:5, Informative)
Mathematically that's what you'd expect but 50% CPU load doesn't mean 50% the power - it completely depends on what state a 50% load puts the CPU in (lowered clock or same clock / sleep-state during idle / etc). Plus you have to remember that a halted-state CPU still consumes more power than a CPU that is completely powered down. This is why it takes 45% rather than 37.5%. You can't apply such simple principals to power calculations.
Parent
But will it... (Score:3, Insightful)
Joking aside, lately I've been pondering AMD's next move in the everlasting Intel vs. AMD chess game.
I'm here for hoping they can pull ahead again and force Intel to do the same.
Always Remember: competition is good!
128KB L1 (Score:2)
Not 32/32.
Tom [not official...]
Amazing (Score:2, Funny)
A company announcing upcoming features in order to create hype for their product? Who'd have thought of that.
Biggest way for AMD to save power (Score:5, Insightful)
Re: (Score:2)
That, of course, if you ignore current leaks...
Re: (Score:2, Interesting)
Intel shipped a few 65nm processors in 2005, but didn't really get started until 2006, and full conversion might not have happened yet, although all the important plants should have migrated by now.
AMD have been behind on the process node, but that's not the only issue when it comes to making chips, although it is the most major. SS + SOI are other technologies that AMD is far ahead of Intel on, and they help reduce power sign
Re:Biggest way for AMD to save power (Score:5, Informative)
Nah. Here's what Intel's 45nm page [intel.com] says: This important milestone demonstrates that we are on track for 2007 to manufacture chips on 300mm wafers using the new 45nm (P1266) process, in accordance with Moore's Law.
Intel shipped a few 65nm processors in 2005, but didn't really get started until 2006, and full conversion might not have happened yet, although all the important plants should have migrated by now.
Even if true, AMD has yet to ship a SINGLE 65nm processor. By this measure alone, I'd say the claim that they are a year behind is quite adequete. But by the speed at which AMD is producing fab plants, I'd argue that they are or soon will be an entire chip generation behind.
AMD have been behind on the process node, but that's not the only issue when it comes to making chips, although it is the most major. SS + SOI are other technologies that AMD is far ahead of Intel on, and they help reduce power significantly - hence AMD's low power 90nm processors compared to Intel's 90nm, and even Intel's 65nm P4s, and AMD aren't doing too badly in terms of performance/Watt right now either.
Traditionally Intel has won the absolute performance title. As you said yourself it is the biggest factor when it comes to performance/Watt statistics. If been following any of the Core 2 Duo reviews, Intel is now dominating in that arena too.
Parent
Power is always key (Score:5, Informative)
There are two huge concerns in a typical data center enviornment: Heat and Power. These two areas are key because of the density of servers today. We're cramming so much processing and storage into 48U that people 10 years ago couldnt have even dreamed of even existing. Delivering enough power to run 48 servers can be difficult if each server is pulling 4 amps each (thats 192 amps). Considering most circuits are 20 or 30amps, thats alot of circuits to fit in one rack.
This was always the biggest reason why Dell servers were not as popular with the companies that I have worked with. Quite simply, AMD was kicking Intel's ass with heat and power. I heard many people say they'd start ordering Dell servers by the pallet if they sold AMD processors (looks like they finally listened).
I don't get the point of this (Score:4, Interesting)
I'm sceptical that this technique will be very useful. (Of course, AMD is full of smart people and I'm just some net.moron.) I don't think it will be very common for the load on a 4-core processor to be somewhere the middle like 1.5. It's either going to be mostly idle (load close to 0) so you might as well power down the whole chip, or going full blast with the load as high as I think will give me the most throughput. For example, when compiling (and that's when I wish I had more cores) I'm gonna "make -j n" and my load is going to be about n, and that number is going to be chosen to be one more than the number of cores I have (or something like that). If I have a 4-core machine, do you think I'm going to make -j 2? No way.
I can't think of many situations where I would have one core running at 100% and another at 50% and the others idle, for any significant length of time. I can imagine a desktop user clicking on something and maybe for a few milliseconds that load is somewhere around that, but then the work gets done and you're idling again. Or the user asked it to do something "hard" so all cores are near 100% (except maybe while waiting for I/O) for a "long" time.
Am I wrong? What kinds of things does your computer work on, which are a little parallelizable but not very much?
Re:I don't get the point of this (Score:4, Interesting)
Another of my computers is basically used to play games. Most games don't seem to do much on the SMP side, so I doubt it would much matter how many cores there were as far as the game's concerned. They do tend to peak one CPU pretty much all the time though, while another core might end up servicing OS calls. Again, it couldn't hurt to let those sleeping processors/cores power down while they're not doing much of anything.
Parent
Re: (Score:2)
You are right that in certain high load applications you may not need it. But remember for every live server in the world there are dozens of test boxes which take power just the sa
Re: (Score:3, Interesting)
How about games and media playback? In those applications, you have X amount of work that needs to be done every (say) 33ms... there might be more work to do than one core can handle, but not so much that you need all 4 cores.
OpenGL equivalent for Ray Tracing (Score:3, Interesting)
Efficiency, yes (Score:5, Funny)
I'll be happy with these new processors as long as I can still efficiently heat my apartment with them.
Re: (Score:3, Interesting)
Re: (Score:2, Funny)
Re: (Score:2)
Lots of us do care about power. We fill racks with at least two VIA chips per 1U slot. Errr... VIA isn't made by AMD. Well, good luck to them!
Re: (Score:2)
Re: (Score:3, Informative)