Slashdot Mirror
New PowerMac G5s: Up to 2.5Ghz, Liquid Cooled
no_demons writes "Apple have just announced the new, completely dual-processor PowerMac G5 lineup. The models all sport an 8x SuperDrive, whilst new the dual-2.5GHz model also features an 'innovative liquid cooled heat sink,' available in July."
The clock speed is useless to compare different architectures or even different processor lines made by the same manufacturer. So, A G5 running at 2.5GHz can't be compared to an Athlon running at 2.5GHz on clock speed alone, for example.
But between two otherwise identical G5 chips, it can be assumed the 2.5GHz one will go faster than a 2GHz one. It's essentially the same chip, just running faster.
-N
I've nothing to say here...
It's been done...and more. One guy dumped all the components of his PC in a polystyrene tub full of liquid paraffin, which is pretty much what you're suggesting. However, he also stuck the evaporator of a fridge in there and got the paraffin down to stupidly low temperatures. The disadvantages: it looked crap and it smelt bad.
When I am king, you will be first against the wall.
2) Fast video cards are nice, but if I want a gaming machine I'll get a PS2.
3) Yes, only 256MB of RAM on the 1.8GHz. You need more, buy it from somewhere else. Apple overprices their RAM
The real development here is the liquid cooling. It's a big step forward, because this means that they might be able to put faster processors into the Powerbooks, and they'll be ready for 3GHz and faster processors once IBM overcomes the Voodoo Curse.
I'm a little disappointed, but since I have to save for a wedding, I'm kind of glad they're waiting to release the dual 3GHz. It will be easier to convince my future wife that we need it when we aren't dropping $5000 next week for a payment.
I'm in the hole of the broadband donut.
Macs run Windows XP a hell of a lot faster than a PC can run Mac OS X.
Post: Sigged, for your pleasure.
Its only 3 dB less. Remember that funky math that we said we'd never use? 2x the noise = +3dB. You can get more difference than this by simply switching your ghetto case fan for a good one. Unless of course Apple MEANT 1/2 the dB, which is a meaningless number w/o a reference: 1/2 the dB of a 6dB source is 3dB, or half the volume. 1/2 the dB of a 50 dB source is more significant.
So before we all drop to our knees on this one, lets consider the physics.
I want to delete my account but Slashdot doesn't allow it.
...and multiple fans: Apple does it because they want to keep the machine as quiet as possible while still as cool as possible (as opposed to being forced to do it, lest the processor become hotter than the surface of the Sun).
(They don't do it because the PowerPC 970 family is "so hot", either; the PowerPC 970, and the 970FX even moreso, run much cooler, and require less power, than even the newest generation PowerPC 74xx (G4) family processors: )
Also, new PowerPC 970FX information from IBM is now available.
Actually the breathing hole in the hard drive is critical to maintaining the cushion of air that the drive heads use to float the couple of microns over the platter surface that they need. If you plug the hole, you've got a good chance of having a hard drive crash in the most literal sense of the phrase when the heads dig into the platters on boot.
"Because Science" is one step from "Because old book". Try "Because of my experiment testing my falsifiable assertion".
Lets see...
Adding more Drives....hmmm
Well there is the firewire route, last time I checked fw800 could handle full DV...
then there is the fiber channel BTO option with an Xraid, I think 2Gb/s of drive bandwidth will do just fine too.
If you want to go oldschool you could also add one of those newfangled SCSI U320 cards which would also do a pretty good job with DV as well.
So I guess the answer to your question is a resonding yes. Since you do (very) high end work then you should be using a very high end storage system like one of the above. If you are a professional and need high bandwidth you shouldn't be relying on apples software raid anyway. But thats just my two cents.
>
Noise level (bels, often referred to in tenths of bels, or decibels) is a logarithmic measurement, similar to the Richter scale. The number of bels for a given ratio of power levels is calculated by taking the logarithm, to the base 10, of the ratio.So this is actually a reduction of just over three decibels. Doesn't sound like much, but it really is twice as quiet. Gives you more respect for the 20db case fans, eh?
Then simply stick to doing what the rest of us are doing.
internal drives in the computer are system drives.
leave the stack of external fiberchannel/scsi drives as the media drives in the raid array.
works great for AVID when dealing with HD or digitalcinema.
and I prefer that those 15Krpm drives be external... we have had to RMA 2 of them already this year for failures/ excessive noise from 2 different AVID suites here in the office.
External media drives are the only way to go.
Do not look at laser with remaining good eye.
That would be Fluorinert. Not a bad idea. NASA had a fluorinert-filled heatsink inside a mylar bag that I used once. Geez. Its still here. I need to clean out my desk more often.
PCI-X is twice as fast as PCI and is a replacement for PCI. PCI-E is a replacement for AGP and is twice as fast as AGP 8x. PCI-Express boards will have multiple PCI-X slots for sound cards, network cards, etc, and one PCI-E slot for the graphics card.
Well, actually most all of it is done in 2D with OpenGL.
GL does have a couple 2D Drawing modes, GL_ORTHO, for instance, and cards hardware accelerate them. How do you think games draw their pretty little GUI's and menus and whatnot?
Interestingly enough, nobody's ever developed a really good benchmark for cards that can accurately compare card performances drawing to ortho's. Maybe 3DMark should include a test like this. I imagine that raw fill rate has the biggest impact here, but who knows what kind of crazy optimizations card manufacturers might have in there to help/hurt the 2D OpenGL performance in favor of the 3D.
IBM went on record as saying that they underestimated the difficulties of switching to 90nm. There's a lot of current leakage issues with the process that lead to increased heat output from wasted energy.
SPEC.ORG doesnt have any recent PowerPC benchmarks, but looking at historical bechmarks (Specint95 on 500Mhz processors) PowerPC has about a 20% higher score than a pentium of the same Mhz.
I would guess this advantage has increased as PowerPC pipelining and paralellism have improved dramatically since then.
So a 2.5 GHz PowerPC should be able to crunch numbers better than a 3 GHz Intel.
The chip also has the advantage of not being constained by the 8080 architecture.
Old COBOL programmers never die. They just code in C.
I must say that I am impressed. I am really not a big apple fan but the company seems to do just the right thing again and again. Starting with USB on the iMac back in 1998 they made a lot of cutting edge decisions wich came into mainstream just because Apple made them successful (WiFi, Firewire, MP3 players, legal musicdownloads, their stereo-wifi-hubbie-thing, TFTs, DVD-Burners) and so on).
Watercooling has been around for some time but no majorplayer implemented it. I bet that two years from now this could well be standard at Dells, HPs and so on...
Did you look at the page?
The schematic clearly shows a closed circuit of pump driven fluid flowing past the CPUs getting heated and then flowing into a heat exchanger (takes the place of a holding tank, and is actually superior too one) with a fan blowing on it. This is exactly like a smaller version of the cooling system in a car. There is no phase change.
A heat pipe is completly different, it is a phase change driven system, and does not involve pumping, or traditional heat exchangers.
"I'll have a Guinness, no wait, make that a Coors Light" -Grad student I work with, who shall remain anonymous...
I'm sorry, but you have no idea what you're talking about. Radiators by all laws of physics can only cool down anything going through them to (almost) room temperature. Now the equations you want are that if the radiator can bring the water within 2C of room temperature with one CPU pumping out heat, then with two CPUs it will bring the water about withing 4C. You are correct thout that the second CPU will (obviously) be warmer than the first one.
I was a little upset when I spotted the upgrades early this morning because I just ordered a G5 1.8x2 the day before last. Imagine my delight when I got an email this morning from them offering to let me change my order. So now I'm getting a G5 2x2 for a hundred bucks less. Booyah!
Am I the only one that would rather have a dedicated controller to handle the fans as opposed to during it in software?
assuming that the new g5's work the way the current ones do in this regard (and I'm sure they do). the software overrides the hardware. if there is no software to control the speed, then it defaults to full blast.
if you install linux on a g5 then the fans all run at full because the linux people haven't decoded the fan controllers yet (or hadn't last I looked). and even with osx on it, when you turn it on, the fans blow at full for a short while until bios/osx gets loaded and takes control and brings the speed down.
so if osx crashes so hard that even low level drivers like that die, then the machine should still not overheat.
dave
A PC case of similar size has maybe four external drive bays and five hard drive bays, and five PCI slots plus an AGP slot.
A PC case with similar expandability, like the Antec Aria, is much smaller. (Though the Aria doesn't have two CPUs and can't hold 8 sticks of memory.)
While the insides of most PC cases may not look neat and tidy, it's child's play to buy or build a quiet, powerful PC that never overheats no matter what the load. So while PC cases may not (in general) have air channels, I'd say they work as well as they need to (though Intel apparently disagrees, since the BTX form factor is coming and is supposed to be designed for airflow).
I don't think you could reasonably expect a mainstream manufacturer to give up the expandability for an academic increase in cooling efficiency, although Shuttle has clearly demonstrated that at least some people are willing to give up expandibility for a slick form factor.
The G5's are beautiful though; I wouldn't mind owning one.
Some of the revisions of the old CRT based iMacs were convection cooled. They had no fans at all. The problem with doing that with a mini-tower case is that people tend to put things on top of a mini-tower... which would block the airflow... They got away with it on the iMac because its hard to put stuff on top of its curved top.
Doubling the acoustic energy (in watts) does in result in a 3db measured change in SPL, but that is a barely noticeably change in the (human) perceived volume level. Our ears are not linear devices, that is the reason that differences in loudness (sound pressure level) are represented logarithmically. 1 DB is considered to be so small a change as to be imperceptible (in nearly all cases) even though it represents a significant change in absolute acoustic energy (in watts).
This points out why it is so difficult to keep things quiet, (and why this was a significant change) you must reduce the acoustic energy tenfold in order to halve the perceived loudness.
for a quick definition of sound pressure level http://arts.ucsc.edu/EMS/Music/tech_background/TE- 06/teces_06.html
Actually the breathing hole in the hard drive is critical to maintaining the cushion of air that the drive heads use to float the couple of microns over the platter surface that they need.
Reference? I don't know of any hard drive where this is the case. The only purpose of the breather hole is for pressure equalization. The heads fly due to the aerodynamics of their physical design-- just like an airplane wing.
Ok... here goes... the cheapest dual Opteron system I can build, based on the 1.8ghz Opteron 244:
Mobo: MSI K8T Master2-FAR $220
CPU1: AMD Opteron 244, Retail $330
CPU1: AMD Opteron 244, Retail $330
DIMM1: 128MB ECC Registered DIMM $ 60
DIMM1: 128MB ECC Registered DIMM $ 60
HDA1: WD800JD 7200RPM 80GB SATA $ 75
VID: GeForceFX 5200 $ 55
DVD: 8X DVD+/-RW $ 90
CASE: Lian-Li PC-V1000 $200
PWR: Antec TRUE430 $ 70
MISC: keyboard, mouse, fans, etc.$ 50
=====
total $1540
So you're talking about a system that's about $1540 pre-shipping (which would probably run close to $100). And that's with the cheapest motherboard and RAM money can buy.
The dual 1.8ghz machine with otherwise similar specs from Apple is $1999. So you're paying a premium for quality system design and support, and software.
Lex orandi, lex credendi.
Linux has had G5 fan control for several months; it works OK in our testing.
You wouldn't get a warranty with the Opteron system, not that you get a great one with the Apple, but it's better than nothing.
OS X is definitely worth something, although if you're coming from a Linux POV you might think that OS's should be free. I'm willing to pay for some ease of use and elegance.
For my money, the Mac is a better deal, and arguably cheaper.
the rumor sites picked up on the LCD promo ending 2 days before WWDC (end of June) so they guessed Displays and PowerMacs maybe at the same time....
they also noted the fine print of the promo listed display by part number, so if a metallic 23inch came out sooner, it would not qualify for the discount per say.
though looking at the Apple site now i am only seeing a $500 off 23inch LCD with G5 purchase.... so i dont know if i am missing it of only the 23inch is on sale.
Tom Boger, Apple's Director of Power Mac Product Marketing:
"All-in-all, no we are not getting to 3GHz anytime soon"
Excuse me, but it IS a heat pipe and it's NOT pump driven.
FYI, I did RTFA and see this:
This system provides a continuous flow of thermally conductive fluid that transfers heat from the processors as they work harder. The heated fluid then flows through a radiant grille, where air passing over cooling fins returns the fluid to its original temperature.
That is clearly how a heat pipe works.
Furthermore, in relation to your question to another poster:
Mac OS X dynamically adjusts the flow of the fluid...
To control the flow of the fluid you would need a pump. You could put a valve in it, but that makes no sence, because a heat pipe adjusts it's own flow, passively, based on the heat load.
They did do it without a pump. And it's still a heat pipe. Heres how:
Mac OS X dynamically adjusts the flow of the fluid and the speed of the fans based on temperature.
By adjusting the speed of the fan that blows air through the radiant grille, they can control the cooling efficiency of the radiant grille.
So with the dynamically changed airflow, which in turn changes the temperature, the result will be a change in the pressure in the area covered by radiant grille within the closed system. This will cause a change in the speed at which the fluid flows.
Hence, Mac OS X dynamically adjusts the flow of the fluid
Read it properly next time and apply some common sense... Oh wait, this is Slashdot..
Sometimes I wish I was a plumber, then I'd know how to deal with other people's shit.
The g5 iMacs are on schedule to be released in 2 weeks at the WWDC. Is this what you're asking for?
You know that being said, I'm not sure if I ever care if Apple's prices are cheapened or not. I mean, think of it in this analogy - I love BMWs, always have. I am making a crazy attempt to save for one which will or will not ever come to fruition. But do I want BMW to come out with a $16,000 car (mini coopers do not count) that skimps on some of the quality? No, that, my friend, is not a BMW.
Part of the price included in Apple's products is research and development for its sleek designs. I would not have it any other way.
Yeah, you're paying for the software and the "experience", but frankly, most people don't care enough to plunk down the extra cash,
I do, and have since 1993. There are many that feel the same. It's just a matter of preference, there is no "correct" computer purchase.
-- (Score:i , Imaginary)
Apple has by far the best warranty experience that I have ever encountered. All systems come with 90 days phone support ("I can't get OSX to do this" or "my machine is doing this") and a full year of hardware support, over the phone ("shit broke"). I have a Powerbook G4 I bought in August of last year, and the 2 times I had to call support on it (once for the screen, it was a known manufacturing issue with the 15"s at the time, the other to fix the casing that wasn't reassembled properly, just a bit loose was all, nothing major) I was on the phone for a total of less than 10 minutes, calling during "peak" times, and was on hold for less than 30 seconds before I was talking to a Mac Genious (Apple's tech people, not someone just reading off of a screen prompt). When I got off the phone, a box was on it's way to me and arrived the next day to send it in for repair, and I got the machine back a day and a half later. No computer company that I know of, except for business-grade support can match that. And Apple's AppleCare warranty, which extends phone support and hardware warranty to 3 years, is only ~$250, which is cheaper than about any level of support that anyone else offers, and for a hell of a lot better service
I hate sigs...
That's only partially right. You are confusing the energy being generated with the perceived loudness of a sound, which is quite different.
Your figures are correct for actual power output. However, a sound 10db higher, 10 times as much energy, is perceived as being only twice as loud. Each 10db increase doubles the sound volume, but increases the energy required by 10 times. That's why a really loud stereo takes so much power; to make a sound 4 times louder, it takes 100 times as much energy. 8 times as loud requires a THOUSAND times as much energy.
The reverse is true.... to cut the apparent loudness of your case fans by half, engineers have to drop the amount of generated noise by 10 times. One fourth as loud is 1/100th the original energy. So it really does give you an appreciation for a case that is 20db quieter than another.
So Apple's actual claims could be either 'energy' or 'loudness'. They say it is 'two times quieter', which I perceive as typical marketspeak bafflegab. It's hard to interpret. If the correct interpretation is 'half as loud', then the G5 is at -10db and is generating 1/10th as much sound energy. If it is 'half as much energy', then it's about 3db as you state, and would be perceived as slightly quieter.
Judging from how hard they're pushing this feature, I'm suspicious it's the former... people would be angry if their $3K computer didn't really sound half as loud.
Ummm, no. The heads are stationary relative to the airflow. They move across the platter or are stationary. They are not shaped to be aerodynamic, and actually cause a lot of air turbulence when they move.
What makes the head float over the platter is laminar flow. This is the tendency of air to stick to the platter. This creates a molecules thick layer of fast-moving air that generally spirals out from the center of the platter to the outside, turning in the direction of the platter's rotation. As this flow is faster than the surrounding air, it draws the head closer to the platter. As the head gets very close, the laminar flow slows, until equilibrium is reached and the head floats stably.
The primary reason why most hard drives cannot be used above 10,000ft is because air pressure gets low enough for tolerance limits to be reached.
The primary purpose of the vent hole is to allow the drive to equalize pressure. Variations of pressure cause the case to flex, which can affect the head alignment against the platters.
Hope this helps...
The notion that you're looking for is "boundary layer", not laminar flow. Laminar flow is the opposite of turbulent flow, and you won't find it inside a hard drive.
Consider a coordinate system fixed to an airplane wing. Immediately at the surface of the wing, the airflow is stationary with respect to the wing skin. As you move farther away from the skin, the air moves faster and faster wrt the wing skin. If you consider the coordinate system fixed to the ground, the air far from the wing is more or less stationary (ignoring wind and the like) and the air close to the wing is moving at the speed of the aircraft.
So, the heads are not stationary relative to the airflow. The air is moving more or less at the speed of the platters (in a spiral like pattern as you described). The heads fly through this (pretty high speed) airflow.
The rest of your post is pretty well spot on.
Why yes, I AM a rocket scientist!