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."
Now, these numbers were completely extrapolated from the key cracking rates I saw generated on my Athlon 1200, and estimates based on published power consumption. But it pointed out to me that these distributed contests are not good for us, and they're not free. It personally cost me about $40.00 / year in electricity. So, I don't play the distributed computing games any more.
John
"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.
Great, now my transmeta stock is going to go negative.
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.
I would be a little paranoid if I had a 'Venice' core and was using water cooling, what with the rising water and all...
Wow...I figured that this wouldn't impress you much...:P
You are thinking Pre-2000. AMD has run cooler than Intel, cycle per cycle, for a while now. This is in addition to getting more work done per cycle. The days of AMD being used to toast marshmellows was a loooong time ago. I have both in servers. My dual Xeon server is a fine example of a great system that requires 6000btu of AC to keep a room at 72F. Great box, great cpu, unreal heat.
Tequila: It's not just for breakfast anymore!
... shouldn't it also reduce the heat produced by processors, therefore extending processor life?
:)
Or, for an overclocked machine, extending the amount of time it takes for the processor to die?
Eviscerati.Org: All Hail the Eviscerati
Lowering the power consumption per core is a first step to upping the number of cores. I imagine that CPU power consumption for desktops will level out in the 100 W range and makers will add cores, cache, and clock speed to maximize performance within a given power budget. I could also see some innovators creating new cooling technologies to boost the power budget and thus boost the permissible CPU performance within that expanded budget.
Two wrongs don't make a right, but three lefts do.
"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."
Erm? As more cpus? Or cpus with stupidly high power usage.
Someone once told me that 7/10ths of the world doesn't have a phone line, let alone a computer. Now your telling me that the power usage of the world has increased due to all these people getting computers? I seriously doubt it.
How about all these people are finally getting electric to their houses? They finally have eletric kettles, ovens, irons, microwaves...
Im not saying that a lower usage cpu wouldn't make a difference, but im saying its going to make a very small difference compared to somethings.
Plus its going to be a LONG while before we see any difference. The only chips really to take the pi££ when looking at powerusage are the top end P4s, not like teh A64s etc are as bad as these?
As newer low powerchips are already out i doubt the p4's are going to make much of a impact either way.
- http://www.milkme.co.uk
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.
Do they still let users overclock their cpu's? I know intel locked thier CPU's. I wonder if AMD still lets people play with their products more.
Rosco: "If brains were gunpowder, Enos couldn't blow his nose."
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."
Your logic about AMD following the leader used to be unequivocally true.
Today, not so much. AMD really trumped Intel with the 64 bit architecture, and AMD 64 bit chips are the CPU of choice for huge numbers of gamers these days (after all, who else notices the raw speed of a processor like a gamer?)
It was a joke! When you give me that look it was a joke.
You laugh, but it's true. I spent some time living in Kenya, and Internet cafes are everywhere. The power grids in these countries are already so stressed that to have a chip that drastically reduces power consumption in these places would be a tremendous help. And not just for Internet cafes, but for point-of-sale terminals, businesses, etc. (And yes, they really do have point-of-sale systems in the "third world.")
Ah, this is obviously some strange usage of the word 'safe' that I wasn't previously aware of.
I think the #1 problem AMD must overcome is the relationship Intel has with Microsoft. AMD makes clone chips, Intel makes chips that fit into Microsofts OS. Intel and Microsoft share information about how the chip will work with the software.
Right. Which is why Intel and MS have both adopted AMD's x86-64 stuff. Intel are no longer leading, they are following.
---- Den ene knappen er powerknapp, den andre er Bender voice knapp "Bite My Shiny Metal Ass"
While a typical home user probably does have other, larger energy hogs, we have almost 300 systems between desktops and the compute farm. This would be a huge savings for us, both on the front end (direct power to run computers) and on the backend (air conditioning).
For someone with a huge sim farm (ATI, Nvidia) or other giant compute farm (google, MS), it's a phenomenal win.
By the way -- Sony sucks!
Gamingmuseum.com: Give your 3D accelerator a rest.
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!
"I think the #1 problem AMD must overcome is the relationship Intel has with Microsoft. AMD makes clone chips, Intel makes chips that fit into Microsofts OS. Intel and Microsoft share information about how the chip will work with the software."
AMD came up with x86-64. Microsoft was only willing to support one 64-bit extension to x86, so that's what Intel chips use; they are the clones now. And Intel is the one with compatability problems (eg DMA is broken with Intel x86-64 chip, which seriously hurts performance).
I don't support one over the other. They trade performance and price/performance crowns regularly and I'll buy whoever's ahead this quarter. Just sayin' that AMD not "just a clone" anymore.
I rarely criticize things I don't care about.
If AMD had some brains they would hire a few engineers to submit optimization patches to gcc for AMD processors. They could get an edge OVER intel by having the best compiler technology avaliable publicly as opposed to ICC which is difficult to integrate into open source projects as GCC is pretty much the standard.
Religion is a gateway psychosis. -- Dave Foley
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
In other words, cautiously we project the current power consumption of all computers running somewhere in the order of at least 20 Hoover Dam power plants
.04% of the annual US oil consumption.
If 9000Mw/hrs are the equivalent of 4 Hoover dams and current estimate is 20 Hoover dams, then current consumption by CPUs is around 45,000 Mw/hrs.
This site quotes 10.9 cubic meters of oil per megawatt/hour.
If my math and sources are right, then CPUs alone, worldwide consume the equivalent of nearly 500,000 cubic meters of oil each year.
According to this site, one American barrel of oil is 0.15899 cubic meters.
That means that the power consumption of all the CPUs in the world equate to over 3 million barrels of oil/year.
Perspective? The US currently uses a bit over 20 million barrels of oil/day. So CPUs worldwide are using around the equivalent of
That's what you get for being a cheap-ass and buying the cheapest toilet you can find. If you'd buy a high-quality (over $125) 1.6gpf toilet, it'd flush the crap just as effectively as any 3.5gpf toilet, and probably better.
I've had lots of 3.5gpf toilets clog on me; does that mean they all suck too? The high-efficiency toilets have gotten a bad rap because stupid house builders, who buy the cheapest crap they can find in order to maximize their profit, installed cheap toilets. So now that everyone's stuck with them (and they're apparently all too damn cheap to go to Home Depot or Lowes and get an American Standard Cadet II for $150 or so), they sit around whining about government regulations instead of blaming their builder.
The government probably should have instituted a minimum performance test when they instituted the 1.6gpf requirement.
So the cranking isn't totally free for those boxes, but worth it because I get to be cool and ranked in the top 99.725% (top 15k of 5.4m) of Seti contributors ;)
Um, you're in the top 0.275%, buddy.
-mkb
You need to get a solar chart for your area of the world, and look up the equivalent insolation in terms of hours. Around here, we get an equivalent of 3.5 hours of maximum sunlight per day, averaged over the course of the year. Assuming your numbers are similar, you'll need about (24/3.5)*200 watts worth of solar panels -- that's 1370 watts. Assuming you get a great deal, you might pay $2.25 per watt, uninstalled cost, so that's over $3000 just for the panels. You'd also have to build a mounting system and possibly install a small motor to keep the panels pointed in the optimum direction.
On top of that, you need a battery system to provide power during hours of darkness. I could continue BS'ing the numbers to figure out how many batteries you'd need but would rather not. Needless to say, it's going to be several thousand dollars for the whole system.
(Yes, I've done this before)
Back in 2000, duing the California power "crisis," Amory Lovins of the Rocky Mountain Institute was asked what things citizens could do to conserve power. His response: "Conserve water. The lagest consumer of power in California is electric water pumps. So if you save water, you'll save power."
Still, every little bit helps. By residents switching over from incandescents to screw-in fluorescents duing the power "crisis," California reversed approximately 8-10 years of power consumption increase (according to some estimates).
Schwab
Editor, A1-AAA AmeriCaptions
For some time now Intel has relied on slick marketing and big numbers while AMD did the same thing... better. Efficient computing is where AMD has gained a nice edge over the years. Intel is playing catch-up at this point. Keep it up. Competition helps us all.
Don't forget that in may large server rooms you actually end up paying twice:
1) the first time to power the chips
2) the second time to remove the waste heat in the server room.
the pay off in some cases may be more than originally anticipated.
putting the 'B' in LGBTQ+
I think you've made a huge leap there. You've tried to imply that CPUs are what's causing the increased demand for power. That's the logical fallacy of Correlation implies causation. I'd be willing to bet that computers use very little of the additional power consumed. Think about if you lived in a developing country and had limited resources to spend, but increasing energy supplies. Would you be more likely to spend money on a PC, air conditioning, a laundry washing machine, or a TV? And of those, the PC probably uses the least energy already.
Software sucks. Open Source sucks less.
Yes, that all sounds about right but has absolutely zero relevance to today and AMD.
The past few years, I have started meeting some people who are fanatical about AMD, how it is better than Intel. And it is no coincidence, many of these people are die-hard linux users as well. But I remembered the old AMD k-5 chips that used to overheat. My logic was "AMD is following the leader, making imitation chips, they will never be in the lead".
I'm not sure how your "logic" follows, but AMD has had several firsts over the last few years - first to 1 GHz., first with on-chip memory controller, and first with x86-64 instruction set. AMD chips are also the highest performers on pretty much every workload except media encoding/decoding. If you're a gamer, they are the best performers these days. Plus, AMD64 CPUs are rock-solid stable, use less power, run cooler, and cost less for the same level of performance. What's not to like?
I think the #1 problem AMD must overcome is the relationship Intel has with Microsoft. AMD makes clone chips, Intel makes chips that fit into Microsofts OS. Intel and Microsoft share information about how the chip will work with the software.
I have no idea what you're talking about here. Microsoft has been up-front about preferring AMD's 64 bit technology, and is using for all their 64 bit servers. Further, AMD is absolutely compatible with Intel, and there haven't been any publicized compatibility issues for quite a while.
And, I guess it is also an issue of name. To this day, I still buy Sony because their TV's were the cadillac of TV's when I was a kid.
Basing your purchases strictly on a company name is a good way to waste money. Do some research and buy the best product. In my opinion, on the PC CPU front, that's AMD.
Oh, one last point about AMD's current lineup - you can purchase a socket 939 motherboard today, and use an inexpensive Athlon 64 CPU for now, then later do a firmware upgrade and install a dual-core replacement once prices come down. Intel has no such upgrade path for its products.
Galileo: "The Earth revolves around the Sun!"
Score: -1 100% Flamebait
...because less heat means less fans, smaller enclosures (doesn't have to have so much room for the air to flow) i.e. __quieter__ machines.
oooh sweet...
"A language that doesn't affect the way you think about programming, is not worth knowing" - Alan Perlis
Surely he's in the top 99.725% also then?
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
Don't confuse energy with price. Electricity is about the most expensive form of energy. If your heating runs on natural gas or petrol, you pay more for heat generated by your computer than for heat generated by your dedicated heating system ;).
And if you use a heat pump, or are connected to "urban heating" (sp?) the computer looks bad even energy-wise (yes, the computer's heating efficiency is indeed 100%, but heat pump is more than 100% efficient because it works by sucking additional energy out of the ground...).
The efficiency loss is at the electrical generation side.
Exactly.
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.
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.
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.
"Here's a recently updated performance benchmark on the G5."
No, it's not. It's an Apple marketing piece. As is the other iterm that you linked.
DO NOT trust manufacturer benchmarks. They are always manipulated - usually by careful choice of the tests run.
Let me guess, you work for Apple?
"The Opteron can certainly win on synthetic benchmarks that test memory bandwidth due to the memory architecture, but most people don't actually need that bandwidth."
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.
"When testing actual customer code we're usually the same or better in performance, with lower power draw and less heat generation. As always, your mileage may vary."
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.
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.
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.
Whale oil was a luxury good. Petroleum is the foundation of our society. We will hit the peak sooner or later, and we need to deal with that fact. I don't think there have been too many new breeder reactors built in your town lately.
"A language that doesn't affect the way you think about programming, is not worth knowing" - Alan Perlis
Maybe it's somehow getting converted into mass. Have you weighed your PC lately?
"So, for 20% to 30% more performance, according to your numbers"
20% to 30% PER CPU. The DL585 is a 4-way (soon to be 8-way) box with iLO, 64GB memory capacity, and a whole swath of other features that the XServe G5 cannot match. You cannot run an enterprise-class DBMS on the XServe - it doesn't even have redundant power.
You *cannot* compare a 64GB, 4/8-way Opteron server with redundant power and cooling to the XServe G5. Hell, can you even replace the fans in the G5 while it's on?
"But, I do think it illustrates my point that the Xserve often has better price/performance when you factor in power and cooling expense."
The XServe is not an enterprise-class server. It lacks the features that even many low-end PC servers have - important things like redundant power, SCSI, and large memory capabilities.
Downtime is not acceptable. It's not OK to have the DB server crash because the PSU crapped out. It's not OK to have to take it offline to replace dead fans or dead disks.
Oh, and again, our customers want Linux or (in some cases) Windows. Not a "BSD-based" commercial OS.
Try getting Oracle to run on OS X server.