Slashdot Mirror
AMD Says Power Efficiency Still Key
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.'"
But will it run Vista? (j/k)
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?
Tequila: It's not just for breakfast anymore!
The article is very light on details but the one picture implies power control at the core level. For example if core-1 is running a 100% workload and core-2 has a 50% workload, core-3 and core-4 can be halted resulting in a power load of only 45% the total 4-core max load.
... run OS/2?
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!
64K code+data.
Not 32/32.
Tom [not official...]
Someday, I'll have a real sig.
Well cover me in egg and flour and bake me for 14 minutes!
games journalism blog
In other words, we can expect the quad core AMDs to still get throughly spanked by Intel's Core Duo technology as far as performance goes. Not that it's necessarily a bad thing, extra performance may not be necessary and the power savings may make it simply more cost effective.
:-)
Of course, most people purchasing server technology probably won't care about power consumption. It's not their budget being wasted on power, it's some other overhead account they don't have access to.
I expect AMD is trying to work to convince people that their lower power consumption is worth the performance lost over using an Intel part. Sort of like the "megahertz myth", but instead more of a "total cost of ownership" deal. Like how Microsoft explains that Linux is REALLY more expensive than Windows, or something.
I really have no idea where I'm going with this. Hopefully someone can figure out WTF I'm talking about and make it coherent.
what the usage graph will look like on actual consumer machines with typical currently available software?
I'm betting that 99% of the time when the user's running some CPU-intensive application it'll look exactly like the image in TFA: One core busy with a single-threaded app, one core slightly occupied with background system tasks, and the other two cores twiddling their thumbs with nothing at all to do.
Sure will be nice when we have more software that can make use of two or more cores!
A company announcing upcoming features in order to create hype for their product? Who'd have thought of that.
is to reduce the distance between their transistors from 90nm to 65. Intel started shipping their 65nm chips nearly a year ago (OCT 2005), while AMD has yet to ship any. AMD isn't expected to be fully converted to the 65nm process until mid-2007, and by then Intel is expected to start shipping their 45nm chips. AMD is playing catch-up these days and it's hurting them bad.
AMD now provides a TSC (cycle counter) that doesn't vary in speed when the core speed changes. This greatly helps with timekeeping.
As for race conditions: that is pretty well taken care of already. SGI has Linux on a 2048-way system now.
It's "Yes, it's like 'goldy' and 'bronzy' only it's made out of iron.". If you're going to borrow lines from Blackadder to such bad effect, at least get them /right/.
Well, at least if you are in a data center.
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'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?
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
How does this tally with a previous story about multi-core architectures being ideal for realtime ray-tracing in games? Is anyone working on a Ray-Tracing evuivalent of OpenGL?
**TODO** Steal someone elses sig.
AMD and Intel are last century. bor-ing! I am looking forward to some cheaper mobos that have both POWER and CELL onboard. That's my idea of "dual". *IBM* is where cutting edge happens.
This was one of the marvels of the BeOS. It excelled at one thing above all else: Multithreading. Almost every part of an application was a lightweight thread, and the kernel's scheduler was a work of art. It's really a shame that the original architects never got a chance to finish their "sticky threads" impelmentation; cache trashing was a real bottleneck on heavily loaded machines.
Women are like electronics: you don't know how damaged they are until you try to turn them on.
AMD Says Power Efficiency Still Key
I'll be happy with these new processors as long as I can still efficiently heat my apartment with them.
I got my Linux laptop at System76.
Windows NT 4.0 and run it on a multicore MIPS or PowerPC based CPU.
Compared to everybody else on the list (see Intel), AMD looks pretty clean.
Founding member: He-Man Windoze Hater Club
AMD already can halt the two cores in an X2 setup independently.
This saves power, but better yet would be if you could (halt and) power down the cores independently. AMD cannot do this yet, but Intel can.
http://lkml.org/lkml/2005/8/20/95
I didn't know AMD could do that, as I run Windows and under Windows the two cores speed up and slow down (several steps from 1GHz to 2.2GHz) at the same time.
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 unless the other one slows down too.
So you'd rather run two cores at 1.7GHz and 1.2V instead of one core at 2.4GHz and 1.4V and one core at 1GHz and 1.4V. The power usage goes down with the square of the voltage, so the former case is saving 25% more power than the latter.
http://lkml.org/lkml/2005/8/20/95
You're right about Unix/Xenix having long existed in both 32-bit and 16-bit. I was just referring to the long delay between the debut of the 386 in 1986 and OS/2 2.0 in 1992 and WinNT in 1993.
http://www.xbitlabs.com/articles/cpu/display/amd-k 8l.html
Power efficiency was the reason cited by Apple for dumping IBM/PowerPC for x86. If AMD can clean Intel's clock with coolness, I wonder how long until Apples are Intel's biggest competition again.
--
make install -not war
no, he's merely trolling for page hits, as you can see from here: http://slashdot.org/~applix7
I doubt that AMD has launched a new process within 6 months of Intel in the last decade. It's a almost a basic fact of life. Intel has the money and experience to manufacture at smaller dimensions sooner than AMD. Sometimes 6 months, this time about a year.
It's not necessarily "hurting them [AMD] bad" though. The Opterons are still quite competitive in power use versus the latest Core 2's (certainly not as bad as the P4 vs the Opteron). Plus, there are time and cost benefits to letting Intel work out the kinks in the system -- obviously Intel doesn't tell AMD what to do, but there are changes that affect shared suppliers that benefit everyone.
Modern companies believe the features of their products are merely an excuse for their marketing campaign statements. Like with HD DVD / Blue Ray.
Speed is apparently no longer something you need to care about: all CPU's are about fast enough for most uses.
Cores... I swear a modern OS (Vista including) can simply make no use of more than 2 cores.
Power efficiency: they are all more or less the same, unless you have a Pentium4 / Celeron (P4 based) on a laptop/desktop system, in which case you may upgrade.
64-bit -> that's the last selling point. And AMD offers this for a long time now. Intel will sell plenty of Core 2-s since they are late on the train.
But then what? They should either invent totally new way to use a CPU or scale down dramatically (both AMD and Intel).
"AMD Says Power Efficiency Still Key" .. funny, I thought customers tell to vendors what is key. When did this turn the other way around?
I know that the individual VIA CPU does not need much power, but it is also significantly weaker in performance than an AMD or Core 2 Duo dual-core.
So how many VIAs do you need to replace one AMD or Intel?
C - the footgun of programming languages
AIUI, a scheduling challenge with hyperthreading was ensuring that two simultaneous threads didn't compete for the same types of execution units [int or fp or vector?] at the same time. That wouldn't be a problem with separate cores, each with their own execution units.
Actually the new low power X2-3800+ (ADD) now has the performance/Watt crown.
It has idle power of 8 watts and full load of 25 watts.
They measured the performance of various benchmarks and also the energy required to perform the benchmark to get a figure for performance/Watt and the
X2-3800+ came out miles ahead of the Conroe (or any other chip). http://www.lostcircuits.com/cpu/low_e/6.shtml
When it started to cost more for the electricity to power a server for a year than it cost for the server.