Slashdot Mirror
AMD 'Venice' Core Shows Big Drop in Power Needs
dtjohnson writes "
Lost Circuits has carefully
measured the power consumption of four recent Athlon 64 cores and
has found that power consumption has been dramatically reduced in the
new 'Venice'
core from the relatively-low (compared
to Intel P4) numbers of the original 2003 'Clawhammer' core to less
than 30 watts under load and less than 10 watts for Windows at
idle. This huge power reduction was apparently accomplished by a
combination of 90 nm die shrink, Silicon-on-Insulator
technology, and something called 'dual-stress
liner technology' As Lost Circuits points out, power
consumption worldwide has been exploding as more CPUs come online and
the CPU power requirements increase so a significant power reduction
will reduce the burden on electrical grids everywhere."
Power density is one of the biggest reasons why you don't see massive colocation facilites. Unless they themselevs are colocated with a power plant. :)
The mirror of the article is here
"I'd rather be a lightning rod than a seismometer." -Ken Kesey
"a combination of 90 nm die shrink"
No, the Winchester core preceding it was 90nm. There was no die shrink with Venice.
Still a great core, but this is a blatant error on the front page.
I wonder how many cars of coal have been used to read /. ? While every watt counts, I could do much better replacing my light bulbs with lower wattage. In California it's like installing a low flush toilet and save 1000 gallons a month when the central valley uses 80% of the water for watering crops.
is this you?
-mkb
Alas not. Hard drive power consumption is pretty small. 10-12W in normal action. Now compare to CPU and GPU usage. These are the buggers that can take the juice to stop your drives firing up.
While it doesn't really make that much of a difference, the core lines gon Diego (1MB L2) andV enice (512kB L2).
Clawhammer(754)->Clawhammer(939)->nothing->Sa
Newcastle(754)->Newcastle(939)->Winchester->
But whatever. I'm sure the extra cache doesn't make too much of a difference.
You can't buy AMD from Dell because Dell doesn't sell computers with AMD chips in them. Kinda how you can't get cheese puff type snacks from the Dannon company.
As for where to get them elsewhere, try http://pricewatch.com/ I start every net search there.
Pulp Audio Weekly - Geek News and Reviews
This is a pretty good decrease in consumption, but according to http://computer.howstuffworks.com/monitor10.htm "CRTs are somewhat power-hungry, at about 110 watts for a typical display, especially when compared to LCDs, which average between 30 and 40 watts."
HP, IBM, etc, etc... Just a few of the companies that sell AMD64 machines.
Do you use electric heat in your house? Do you live in a cold climate?
If you live in, say, Norway, I suppose there's a good chance the answer to both questions is yes.
Otherwise, your argument doesn't stand. If you live in a warm climate, for at least part of the year, the CPU heat is at best not welcome or at worst increases your A/C load. In the winter, the production of the CPU heat may result in more energy use/pollution than what would have been produced by, say, a gas furnace, depending on the source of the electricity. In the case of a fossil fuel electrical plant, the heat your CPU generated is represents roughly 1/3 of the total heat generated (the other 2/3 is 'waste' produced at the power plant).
Simple Unexpected Concrete Credible Emotional Stories
Um no? Your CPU isn't a Heater, it's just a byproduct and since 100% of the energy going into your CPU isn't being put off as energy, it's nowhere near 100% efficient.
Where is the other energy going then? Disneyland?
I think the 2nd law of thermaldynamics may apply here. It doesn't just disappear.
Tequila: It's not just for breakfast anymore!
While these chips are "lower power" they really aren't "low power". There are some good, powerful, truely low poer chips out there that burn 5Watts that would be perfectly fine for POS systems and web browsing...
Their Athlon 64 FX-5x line is unlocked, designed for the enthusiast crowd. Their Athlon 64 xx00 series is multiplier-locked, but you can still play with the FSB.
$8.95/mo web hosting
Xbitlabs found that Venice uses slightly more power than Winchester (the older 0.09u core) around a month ago. They tested cores at the same speed unlike Lostcircuits, and while LC is a good site, xbit is generally better. Not to mention the guy at LC blew up a few MBs before "finding out" how to do his measurements. Aslo Xbit is the only site I know that has an accurate video card power consumption database. http://www.xbitlabs.com/articles/cpu/print/athlon6 4-venice.html
Look here (EVERY SLASHDOTTER SHOULD READ THIS):
http://www.lifeaftertheoilcrash.net/
Section called "Won't the Market and the Laws of
Supply and Demand Address This?"
The answer is "No". Basically, to retrofit all the systems to use other sources of energy will take a lot more years than we'll have by the time the market forces factor into it.
The market is also not perfectly rational, nor slow to react... if the problem is perceived to be serious, the market will explode overnight.
The oil companies are even faster than the market.
(for example, they are currently swimming in record profits because they upped the prices at the first sign of crude oil's price jump, despite the fact that they were still selling gas they produced at much lower prices of crude oil).
I love the textbook answer, but I think we're placing too much faith in the almighty market.
"If you could only see what I've seen with your eyes..." - Roy Batty
According to this page, Hoover Dam generates 4e9 kilowatt-hrs, or 4e6 MWH per year. 20 Hoover Dams would account for 8e7 MWH per year. Using your conversion factors, that comes out to 5.48e9 (~5.5 billion barrels) of oil a year, or 15 million barrels a day. Scarcely a drop in the bucket, eh? Worldwide, that is 3/4 of the American oil burn rate being consumed by CPUs (note that much electricity is generated by coal and nuclear).
Double check your arithmetic.
Galileo: "The Earth revolves around the Sun!"
Score: -1 100% Flamebait
Please read up http://en.wikipedia.org/wiki/Heat_pump on how a heat pump works, because you are apparently ignorant of their operation.
A PC emits heat at roughly the same rate as a resistive heater (think Lightbulb or electric baseboard heating). A heat pump emits significantly more heat than that (two to three times more), by extracting heat from the outdoors and moving it inside.
No laws of thermodynamics are broken in the process.
And the worms ate into his brain.
Maxwell's demon. Deep.
"A language that doesn't affect the way you think about programming, is not worth knowing" - Alan Perlis
WTF??
Two other people have already said the same thing, but I'm going to repeat it. Either you live in some weird country where toilets are all different, or you're just ignorant. Here in the US (should be obvious since I mentioned Home Depot and Lowes), toilets are all the same. There's a supply line on the lower left side (viewed from front), and a drain on the bottom. There's two bolts on either side of the drain.
To replace, you disconnect the supply line, remove the nuts from the bolts, and just lift the toilet up. Then, replace the wax ring on the flange (costs about $4; wear gloves), sit the new toilet down on the flange/ring and seat firmly. Put the nuts back on (and decorative covers), connect the supply line (a flexible line, ~$10, is a good idea), turn on the water, and you're done.
Of all the plumbing jobs out there, replacing a toilet is one of the easiest. With faucets, you have to worry about whether the sink has 1 hole or 3 holes or whatever. But toilets are all the same. Even the fancy pressurized toilets still install exactly the same way as the regular gravity-operated ones, although their internal operation is completely different.
Don't forget SUN
If you really are looking into getting a backup generator, you may just want to look into getting a CHP (combined heat and power) system for your shop instead. A CHP system will generate electricity for you, as well as allow you to use the waste heat from the generator to heat your shop, hot water, or use an absorption cooling device to cool your shop (although I've never seen one of these for a small-scale application). The CHP system would essentially eliminate or drastically reduce your need for energy from the grid, and probably save you a few bucks in the long run as well. There are only a few manufacturers out there right now though for smaller scale CHP systems (namely Honda, but I think there are a one or two others), and it also may depend on where you're situated (sometimes they will not let you purchase one unless your power company is actively involved in supporting CHP systems).
:)
Overall it would likely save you a ton of money as you could just purchase natural gas instead of electricity. When you factor in generation costs, demand costs, transmission and hookup costs, peak load costs, etc from the electric company, it quickly makes sense to just generate the electricity yourself. Also, unless you have an electric heater for your home, you're probably already having to buy natural gas anyway so there's no additional hookup charge from the gas company.
Another thing to keep in mind is that if you had other shops/stores/residences near you that were interested in doing this it would be fairly simple to install a larger system, split the cost of the system, then just split up the gas bill accordingly (depending on how confident you are that your neighbors will pay on time of course
Actually it's more like 3 to 4 times. Since the server room is cooled by a heat pump (refrigerator) and not by convection (blowing the heat out and replacing it with fresh outside air). The heatpump is only ~30% efficient. i.e. it takes 3 joules of energy to remove 1 joule from the room. That means 1W for the PC and 3 more to remove the heat. The same goes for anything (lights etc) you leave on in your house with the air conditioning on. You'd think big servers would all be moving to the northern states or to Canada ;-) Or in Europe, up there to Sweeden.
Therefore, using gas heat is about 3X more efficient overall than using resistive electric heat. In most areas, energy prices reflect this. (Your CPU is effectively resistive electric heat.)
Heat pumps counteract the inefficiency of delivered electricity by extracting a couple of watts of heat energy from the outside air and moving it inside for every watt of electricity consumed. Therefore heat pumps can be competitive with gas heat (unless it gets too cold to effectively pump). CPUs are not heat pumps, however.
Bottom line: waste electrical heat is not a cost-effective way to heat your house vs. your furnace or a heat pump. In most areas it's cheaper to not create the waste heat in the first place and use your furnace to heat your house instead.
All of this obeys all rules of thermodynamics.
Laugh or not, this country is heavily dependent on coal. Not that I think that any money spent by W. Shrub is going to have any effect other than fill some corporate coffers, if there's a real chance that they can produce more efficient plants or cleaner stacks, then we ought to explore it. Coal emissions are responsible for much of our atmospheric mercury, as well as a number of other pollutants.
John
x86-64 is a crappy 64 bit hack onto the x86. I wouldn't call it innovating, ms picked it because it was only architecture that could run 32bit x86 decently.
The proof is in the pudding.
The x86 itself is just a big pile of hacks, so I don't see the problem with hacking another addition on top of it. But all the performance figures I've seen show the x86-64 performance to be very good. What are you more concerned with? How elegant the architecture of your CPU is, or how fast it performs and at what price (and what power consumption)?
Intel tried doing an all-new 64-bit design, called Itanium, and it's been a disaster. Who uses them? Not very many. They're extremely expensive, they consume tons of power, and their performance is very lackluster. Intel claims the problem is with the compilers, or with the software developers, but the real problem is they tried to do something all new with their EPIC architecture and it was a terrible idea. But AMD's 64-bit chips are selling like hotcakes, and in their price range perform excellently.
In the marketplace, what matters is making products that people want. Innovation is useless if no one wants the results of that innovation (especially when they don't even live up to their prior claims).
No, it's not. It's an Apple marketing piece. As is the other iterm that you linked.
:)
:)
:)
:)
Actually, the other item I linked was the support article showing the actual power draw and BTU's generated. I guess you didn't actually click through and read it. That's OK, it's Slashdot.
Let me guess, you work for Apple?
Yes. And you?
It can also win on applications that use lots of memory bandwidth. Databases, for example, are almost always bandwidth-hungry. So are distributed filesystems. Many technical and scientific computing applications are also memory-bound.
True. What I said is also true, especially since it was about my experience with my customers. Empirically true. How about you? In your experience what's the mix between memory bound and cpu bound applications?
Our mileage does vary. My company has compared PPC970, Xeon, and Opteron using the SAN solution that we integrate, and Opteron is the clear winner. For database systems as well, Opteron is 20-30% faster than Xeon and PPC970.
Yes, indeedy, just like I said. How about price/performance versus cost of heating and cost of power? This was an article about power draw, after all.
Also, you have a CPU-bound SAN solution? Tell me more.
It's great that you're picking the real-world winner based on testing. That's exactly what you should do. In my experience we win sometimes and Opteron wins sometimes, but we usually win on price/performance when factoring in the cost of power and cooling for large systems.
Moreover, XServe doesn't support more than 8GB of memory. That's simply not enough for our customers. Heck, the 64GB provided by HP's DL585 *still* isn't enough.
You are incorrect. I would link you to the correct specs but you'd dismiss them as a marketing piece.
Look, Apple has some nice products, but without a true commercially-supported Linux distro, it's hard to sell your product. People buying servers want Linux or Windows, and they want something that is supported by the vendor.
Funny, I thought we were selling a commercially-supported UNIX distro, and that our server sales were going up. I guess our customers are buying things they don't want!
Have a great day, and try to relax a little.
- "When you want something with all your heart, the entire universe conspires to give it to you" -Paulo Coelho
Which models of Opteron would have these improvements?
Opteron 252 (Troy core) and all dual-core Opterons. If you're building servers, you'll want the dual-cores.
From here:
Throw in several percent generation and transmission loss and you typically end up at around 1/3 total efficency. IIRC, in addition to that quotation, high-tech "combined-cycle" plants which use a gas turbine whose exhaust is used to heat a boiler for a steam turbine can push 60% efficiency before electrical losses. However, that's not yet in widespread use.At any rate, those efficiencies aren't really "incredibly low" given the constraints on heat engines dictated by the laws of thermodynamics. For example, typical automobile engines are a good deal worse than any of the above numbers.
Slashdot note:
...so it continues to be marketing-speak vs. real-world experience.
Todd Dailey works at Apple as a pre-sales server and storage technical contact in the sales group...(info from here).
This sig donated to Pater. Long live
Well, actually sales and marketing are quite different. And pre-sales technical SE and marketing are quite a bit different.
We get involved in a lot of practical deployments, like most SE's in reseller organizations, and most corporate enterprise SE's in large companies.
So it's fair to say that I have less real-world experience than, say, an Oracle DBA doing large enterprise deployments, it's completely unfair to say that I never get my hands dirty. I talk to customers every day and help them with real world deployment issues.
Hi Dan!
- "When you want something with all your heart, the entire universe conspires to give it to you" -Paulo Coelho
Don't get me wrong: distributed.net's RC-64 challenge did a great service by empirically demonstrating several things: the power of lots of computers; the ability and willingness of people to donate to a worthwhile cause; that brute-force can break a cypher; they empirically tested that the amount of brute force required was right where they expected it to be (IIRC 75% of the keyspace was searched before a solution was found); the contest was not shut down by poisoned packets sent by malicious hackers; it verified Moore's Law (looking at the charts of keys-per-second-per-machine over time); it showed how well fat binaries could be used to optimize the hunt on a per-processor basis; and the list of benefits goes on.
But it's over. RC-64 already proved all of these things. It already proved everything that could be proved about a large key-cracking contest. RC-72 will prove nothing new. All it will do is deplete the country of energy. It's stupid in the extreme to continue that contest.
However, proteome folding is NOT a contest. It's research. There are no prizes for the winners, only the satisfaction of knowing you've helped contribute to furthering medical / chemical knowledge. I have no problem with the WCG taking on the noble challenges and helping solve them.
There's also a "grey" middle ground: SETI@home. I guess this one is a "take it on faith" effort -- if you believe (or if you "want to believe" :) in little green men, flying saucers, Vulcans, Klingons, an ascended Marconi or even Cylons, fine. I'm not even complaining about using energy to look for things like this.
My complaint is that d.net has degenerated to a simple "random number hunt." If d.net switches gears, and sends a new client to everyone saying "gosh, we've realized we're just burning electrons here, your clients are now joining the WCG" that would be fine. But they're not. They're continuing what is essentially computational masturbation, and they've got a large chunk of Slashdot readers suckered into continued participation. And I don't understand the appeal, because it's not logical to continue it. No benefits can be gained.
If it's prize money you're after, you're something like 2^23 times more likely to win the powerball lottery by investing a single dollar in lottery tickets than you are in investing that same dollar in electricity to feed the RC-72 contest.
John