PDA

View Full Version : "Crippled" DDR on new G5's?


erockerboy
Aug 14, 2003, 11:27 AM
I recently saw the following info posted on Unicornation.com - anyone care to comment? Please understand that I'm in a position of ignorance here, so don't shoot the messenger... the punchline seems to be that for us DAW users, it would be worthwhile to wait for the next rev of the G5's before jumping in.

Thoughts???

The first rev of the G5s may be obsolete before they even hit the streets. There are some serious questions as the the DDR specs Apple has chosen are about to be replaced by a newer and faster standard (and by standard I mean industry standard).

Of course, you are welcome to spend your money on whatever you want, but from a techie perspective the G5s (which are not going to be that much faster in real world apps) miss the mark by quite a bit and we probably won't see them where they need to be until rev b or c.

Having been a former Apple developer and HW tester I speak from experience...rev A = bad news...and I _may_ have already had my hands on one for testing (nudge, nudge, wink, wink, say no more, say no more).

Apple chose the 400Mhz DDR standard which can be addressed at 6.4 GB/s, as of right now has been replaced by a faster standard which can be addressed at 12.8 GB/s. The real issue with this is that Apple is pretending to move to an industry standard, but they are one giant step behind the rest of the industry (which is so unlike Apple). The good points are that, regardless, DDR is the way to go. The RAM will be available & relatively inexpensive. I'm sure they had a reason for choosing a standard they knew was on it's way out (hell we Mac users were using SIMMs longer than any PC user before we finally moved up to DIMMs).

My biggest issue with all of this is that they really need to compete on a speed/price level with the PC world (like Dell) and chosing an older standard is not the best way to do this. Granted DDR RAM is faster than any of the previous standrads (PC100, PC 133), so there is a significant improvement over the current standard (70.6% faster addressing) but the rest of the system bogs that down to a total of about 43% faster than a similarly rated (Ghz to Ghz) G4. That is one major hit to performance, almost 30% of the gain just given up. Not to mention the difference when going to a 64-bit (2^64) system froma 32-bit (2^32) system should be exponetial. And remember Apple's first venture in DDR RAM with the duals? They actually addressed memory *slower* than the SDRAM systems (for instance the DDR 1 GHz dual was 29% slower in memory intensive processes than the SDRAM 1 GHz dual because of this).

These are all issue that Apple need to address. And, yes given the delay in shipment of the current crop the Rev. B machines will probably be a couple months off (December/January?) by my best guess. I doubt they will upgrade the DDR/Memory architecture to take advantage of the new standard but they should at least fully support the DDR standard that they chose to use in their system to get full benefit of the potential speed gain. That alone will benefit us DAW users in a big way.

Also, a point no one seems to has mentioned. Altivec has been updated, meaning all apps that took advantage of the Altivec processor need to be updated (re-coded) to take advantage of the G5s speed. And those that don't take advantage of Altivec at all will not show as much of a gain from the G5 as they could. Case and point: some non-Altivec apps actually run a percent or 2 faster on a G3 than a G4. For instance Bryce on an iBook beats out a G4 TiBook by quite a bit. But that is another issue for another discussion.

To sum it all up: for a system that should be a magnitude of 10 times faster than the previous crop is only about 43% faster on average (again GHz to GHz). Something is wrong.

As a former Apple Developer I can tell you that I would never buy a 1st generation machine. I've seen what gets swept under the rug for "the next update".

Apple has promised that these technologies are fully supported in the new G5s (DDR-400, PCI-X, 64-bit data flow, the new Altivec code, among others), when in reality they will be fully supported at a later date -- some of them dependent upon software developers who are currently struggling to release OS X versions of their software only now -- 2 years after OS X Was introduced). Read the white papers/developer papers, it's there in black & white (black & white geek-speak, but it's there).

Lanbrown
Aug 14, 2003, 11:32 AM
Faster memory is always coming out; it's a fact of life, just like faster processors come out as well.

Moxiemike
Aug 14, 2003, 11:35 AM
*YAWN!*

complain complain complain.

go do something constructive. :D

NavyIntel007
Aug 14, 2003, 11:39 AM
Originally posted by Moxiemike
*YAWN!*

complain complain complain.

go do something constructive. :D

exactly! Shut up already.

Article was probably written by a closet PC user who is pissed that they still use windows.

szark
Aug 14, 2003, 11:45 AM
Well, this part is completely false:

Not to mention the difference when going to a 64-bit (2^64) system froma 32-bit (2^32) system should be exponetial.

and this part:

Apple chose the 400Mhz DDR standard which can be addressed at 6.4 GB/s, as of right now has been replaced by a faster standard which can be addressed at 12.8 GB/s.

is referring to Rambus which will hardly be a widely used replacement for Dual-Channel DDR (IMHO). It certainly won't be an inexpensive option.

Based on the application tests shown at WWDC, I think the performance of these systems will be more than enough for just about anyone's needs.

bousozoku
Aug 14, 2003, 11:48 AM
First of all, are 1.0 releases ever completely safe?

How many cars, computers, etc. work without a glitch first time out?

As far as I know, the G5 machines are not perfect but they're impressive. Are they going to reach their full potential right away? No! They're being released with an operating system working in a hybrid 32-64-bit mode.

As far as DDR RAM goes, saying that there's a new standard and populating motherboards with actual specimens are two different things. It was surprising to me that Apple chose RAM that was already available instead of helping develop some innovative packaging which contained much faster memory. Then again, in the initial PowerMacs, they used 72 pin SIMMs in pairs. With the PCI PowerMacs, they used DIMMs, ahead of the rest of the industry.

Newer machines will be faster and better--that's the way of things. :)

MacBandit
Aug 14, 2003, 11:50 AM
The only thing faster then DDR400 will be DDR-II which will not be a standard for probably another year. It might make it into high end machines come the first of the year but it is still very pricey.

macphoria
Aug 14, 2003, 11:53 AM
That article is probably written by same person who complained about unfairness of early test comparison between G5 and Pentium PC's.

Powerbook G5
Aug 14, 2003, 12:03 PM
I am confused because if you take just about any high end performance PC, they use DDR 400 RAM, too...so if they accuse Apple of using "out of date" tech, then accuse Dell and Gateway and Toshiba and Sony and etc, too...we just got a brand new top of the line Dell system for the family computer and guess what? It uses DDR 400 RAM, too...I don't see them bashing Dell for using DDR 400.

MacBandit
Aug 14, 2003, 12:07 PM
Originally posted by Powerbook G5
I am confused because if you take just about any high end performance PC, they use DDR 400 RAM, too...so if they accuse Apple of using "out of date" tech, then accuse Dell and Gateway and Toshiba and Sony and etc, too...we just got a brand new top of the line Dell system for the family computer and guess what? It uses DDR 400 RAM, too...I don't see them bashing Dell for using DDR 400.

I bet it doesn't used paired DDR 400. Meaning it still runs at half the speed of the memory architecture of that in the G5.

TEG
Aug 14, 2003, 12:13 PM
Rambus will never become a major player (Heck Intel doesn't even make boards that accept RIMMS anymore, and were the only ones who ever did (In great Quanity)), because they have quirks, act like simms and cost as much as the "Star Wars" Defense Project.

I would like to See DDR-II, but we all know that it may be a while before we'll have them.

With sentiments like that guy's, I know now why the Cube failed.

TEG

AngryAngel
Aug 14, 2003, 12:43 PM
The most inaccurate part is this:

"Also, a point no one seems to has mentioned. Altivec has been updated, meaning all apps that took advantage of the Altivec processor need to be updated (re-coded) to take advantage of the G5s speed."

I suggest that no-one has mentioned it because it is completely false. The Altivec unit is slightly different, but use the same instructions, I believe.

Apps have to be re-compiled for the G5 to see the biggest speed gains, but that is true of most new processor designs- and I don't think this is because of Altivec being a slightly different in the G5 (more like the older G4's, I believe). It won't be like the jump to PowerPC, where many apps were much slower on a 6100 than they were on a Quadra 840 or 950.

Sun Baked
Aug 14, 2003, 01:08 PM
When the computer was announced, people looked around and found that there really were not that many large sticks of memory to be had.

It even seemed that Apple was offering memory configurations that the market couldn't supply, yet.

It's a little hard to say that Apple is using a dead standard, when it's hard to buy 1-2 GB sticks of memory for the darn machines at DDR400.

Like, how would you expect to add DDR2 memory to the machine if Apple offered it?

With a home mortgage?

crazzyeddie
Aug 14, 2003, 01:25 PM
Originally posted by AngryAngel
The most inaccurate part is this:

"Also, a point no one seems to has mentioned. Altivec has been updated, meaning all apps that took advantage of the Altivec processor need to be updated (re-coded) to take advantage of the G5s speed."

I suggest that no-one has mentioned it because it is completely false. The Altivec unit is slightly different, but use the same instructions, I believe.

Apps have to be re-compiled for the G5 to see the biggest speed gains, but that is true of most new processor designs- and I don't think this is because of Altivec being a slightly different in the G5 (more like the older G4's, I believe). It won't be like the jump to PowerPC, where many apps were much slower on a 6100 than they were on a Quadra 840 or 950.

Actually, some Altivec commands are extremely slow (slower than it would be without the Altivec) and others don't work at all. So yes, some apps do need to be recompiled due to the change.

iPC
Aug 14, 2003, 02:38 PM
Originally posted by AngryAngel
The most inaccurate part is this:

"Also, a point no one seems to has mentioned. Altivec has been updated, meaning all apps that took advantage of the Altivec processor need to be updated (re-coded) to take advantage of the G5s speed."

I suggest that no-one has mentioned it because it is completely false. The Altivec unit is slightly different, but use the same instructions, I believe.

Apps have to be re-compiled for the G5 to see the biggest speed gains, but that is true of most new processor designs- and I don't think this is because of Altivec being a slightly different in the G5 (more like the older G4's, I believe). It won't be like the jump to PowerPC, where many apps were much slower on a 6100 than they were on a Quadra 840 or 950.
Not that big of a shift, true. But still... why do you think Adobe was there for the intro? The get to sell new "G5 optimized" stuff, offer minor bug fixes and new feature or two, and watch the faithful upgrade their 1 yr old software! :mad:

*sigh*

What can you do?

Daveman Deluxe
Aug 14, 2003, 02:44 PM
The PPC970's vector pipelines include all 162 instructions in the AltiVec specification. That said, the vector processor is poorly integrated into the unit. It seems that it was really just sort of shoehorned in there (remember, there is no vector processor in the Power4). As a result, the current version of the 970 is not any better than the G4e at vector processing, and it is even worse at some instructions.

CrackedButter
Aug 14, 2003, 02:50 PM
Originally posted by iPC
Not that big of a shift, true. But still... why do you think Adobe was there for the intro? The get to sell new "G5 optimized" stuff, offer minor bug fixes and new feature or two, and watch the faithful upgrade their 1 yr old software! :mad:

*sigh*

What can you do?

Not make a purchase?

gopher
Aug 14, 2003, 03:31 PM
Oh come on, we have the fastest bus on the planet now. 1 Ghz! Dual independent buses at that meaning data doesn't get crammed waiting to go in while other data goes out! Apple developer specs say with 2 GB PC-3200 modules (about $1100 through Pricewatch.com) you can get 16 GB of RAM into the machines. The system runs so much cooler because of the great design of the locations of the fans. Cooler means faster. To top it off, no longer are we limited to ATA/133 ATA, we have Serial ATA at 1.5 GB a second on the hard disk. The biggest slowdown on any machine is the hard disk, not the chips. The 1.6 Ghz G5, nice entry level machine, but not the one I'd get. I'd either get the 1.8 Ghz or the 2 Ghz. The 1.6 doesn't even have PCI-X. This coming off Macs that had buses at a maximum of 167 Mhz while all the people on this board were whaling that the bus made a huge difference. You got your faster bus, you got your faster hard drive, you got your memory capacity, you got your faster PCI slots, you got your faster AGP slot. You even got your digital audio ports. The way people carry on in here, they are expecting the M-5 from Star Trek to actually have been made 30 years ago and available in the desktop today. PCers are just jealous. We have the most secure system on the planet, and now the fastest for under $3000 with built-in developer tools. The tests were run at 32 bit. Imagine what it will be like at 64 bit. Oh and did you notice, the Opteron has to be rebooted to switch from 64 bit to 32 bit applications. The G5 does not. Imagine the time savings there!

Powerbook G5
Aug 14, 2003, 03:57 PM
I personally cannot wait for the G5s to make it into maturity across Apple's different product lines.

ddtlm
Aug 14, 2003, 04:10 PM
Yeah whoever wrote that article is out of touch. But anyway, since everyone already knew that...

MacBandit:

I bet it doesn't used paired DDR 400. Meaning it still runs at half the speed of the memory architecture of that in the G5.
Dual channel DDR isn't unusual in PC land, and in fact dual Opterons can support 4 channels of DDR (although current dual Opteron boards only have 2 channels). Tyan has such a board coming, but all those DIMM slots (16) make it huge and, I bet, expensive.

TEG:

Rambus will never become a major player.
Rambus lost their first battle for the desktop, but the war isn't over yet.

crazzyeddie:

Actually, some Altivec commands are extremely slow (slower than it would be without the Altivec) and others don't work at all. So yes, some apps do need to be recompiled due to the change.
I think they all work, just some memory commands require that the pipeline be flushed, as far as I know.

gopher:

Cooler means faster.
A processor at a given clockspeed is exactly the same "fast-ness" at every temperature that it can run at without error.

Oh and did you notice, the Opteron has to be rebooted to switch from 64 bit to 32 bit applications.
No, it does not. At this link you can see AMD making repeated claims about "simultaneous 32- and 64-bit computing".

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_8826_8832,00.html

Cubeboy
Aug 14, 2003, 04:11 PM
Most of the article is crap but he does have a point about the memory, the G5's 1 GHz bus (8 GB/s peak) is going to be "bottlenecked" by Dual Channel DDR-400 (6.4 GB/s peak). Not that it matters much as the G5 is fast enough already. It's just that it (the G5) will become significantly faster when Apple implements PC4000 DDR-500 (which is out by the way) or Quad Channel RDRAM. Remember, more speed is always a good thing. :)

Powerbook G5
Aug 14, 2003, 04:24 PM
This is still a brand new system...give it time to mature and we'll all be amazed with the capabilities, I am sure.

AngryAngel
Aug 14, 2003, 04:27 PM
Originally posted by crazzyeddie
Actually, some Altivec commands are extremely slow (slower than it would be without the Altivec)

How does a processor without an altivec execute an altivec instruction?

and others don't work at all. So yes, some apps do need to be recompiled due to the change.

I've never heard anything to that end: link?

Lots of the "970 is crap at altivec" comments are from people who have heard comments from people who read the first two Ars Technica 970 articles. There were mistakes in their discription of the unit, which are corrected here:
http://www.arstechnica.com/cpu/03q2/ppc970-interview/ppc970-interview-2.html

But the original article was just criticising how the unit managed dispatches- not that it couldn't handle some of the altivec instruction set. If it wasn't compatible with the existing instructions from the G4, then it would be pretty useless to Apple. The 970 was designed by Apple and IBM together. I think Altivec compatibility was a major target for them.

The recompiles are for optimisation for the new (to the Mac) architecture of the rest of the chip that evolved from the POWER4- not for the only part of the 970 which came from the Mac's existing processors.

KentuckyApple
Aug 14, 2003, 05:20 PM
Not sure that this is a pertinent, but... I sometimes use my parents' Pentium4 which is 1.4 ghz. It has the aforementioned rambus memory, and is ridiculously slow. I had to order the RIMM from the internet b/c nobody in town sells it. It is complete garbage. Thankfully, I have my 400 mhz imac right next to it on the same desk. I am so much more productive thanks to Apple and OSX. I can't wait till my G5 ships so I can see just how pitiful the pentium really is.

TeraRWM
Aug 14, 2003, 06:36 PM
Don't believe everything you hear.

A lot of what you just said is completely false.
Rambus sucks, altivec hasn't changed in the G5 (it's only gotten more room to work with it now), etc etc

Cubeboy
Aug 14, 2003, 09:32 PM
Rambus doesn't "suck" it was just too expensive. It wasn't until quite recently that (old PC-1066) RDRAM was surpassed by (new PC3200) Dual Channel DDR in terms of performance. Before that, Rambus has always held the performance crown on the P4 platform.

Frohickey
Aug 14, 2003, 09:34 PM
If anything, the DDR I based 400MHz is pretty leading edge in memory technology. Ask the former Apple developer and HW tester why he/she is a former Apple developer and HW tester. ;)

DDR-II is a completely different beast. Its definitely a year out, even more. As it is, DDR-II dimms won't even fit in the DDR-I slots. :eek:

illumin8
Aug 15, 2003, 01:23 AM
Originally posted by Cubeboy
Rambus doesn't "suck" it was just too expensive. It wasn't until quite recently that (old PC-1066) RDRAM was surpassed by (new PC3200) Dual Channel DDR in terms of performance. Before that, Rambus has always held the performance crown on the P4 platform.
This is not 100% true. Rambus has a faster total throughput, but the latency is much higher than DDR, so short operations are much slower. Now that dual DDR is approaching 1ghz. (2x 500), there is no reason to buy the much slower overall and more expensive RDRAM 1066 sticks.

MacBandit
Aug 15, 2003, 01:26 AM
Originally posted by ddtlm
MacBandit:


Dual channel DDR isn't unusual in PC land, and in fact dual Opterons can support 4 channels of DDR (although current dual Opteron boards only have 2 channels). Tyan has such a board coming, but all those DIMM slots (16) make it huge and, I bet, expensive.



I realize that and was commenting on a previous posters comments on the top of the line Dell they just got.

ddtlm
Aug 15, 2003, 02:34 AM
illumin8:

This is not 100% true. Rambus has a faster total throughput, but the latency is much higher than DDR, so short operations are much slower.
Interestingly, a very similar thing can be said about the high clocking FSB's on the G5's. Higher throughput than anyone else, but possibly the highest latency around too (although such things are hard to know for sure based on the limited infromation available).

Now that dual DDR is approaching 1ghz. (2x 500), there is no reason to buy the much slower overall and more expensive RDRAM 1066 sticks.
DDR-1 is hardly approaching 500mhz in the mainstream system RAM market. Its definately not a standard, and I'm unaware of any plan to make DDR-1 above 400mhz into a standard.

dethl
Aug 15, 2003, 02:41 AM
Wait a sec....Apple's implmentation of DDR isn't crippled. You have to install the RAM in 2's in order to saturate the FSB- eg. 2 x 6.4 G/s = 12.8 G/s (am I correct? )

So all this new DDR implementation is just a movement to a 128-bit archetecture, something that wasn't available when Apple designed the board. I'm pretty sure the new DDR will be put in when Apple cranks up the machines to beyond 2 ghz, most likely (in my opinion) when they get to 3 ghz.

ddtlm
Aug 15, 2003, 02:45 AM
MacBandit:

Oops, didn't mean to tell you things you already knew. :( But anyway, not that I mean to carry such a minor arguement too far, but I just checked at Dell's site and apparently of the 5 models of PC for "home" use, the top 4 all have dual channel DDR RAM, the top two (Dimention XPS and 8300) being dual DDR-400 and a 800mhz FSB. I must say that I'm suprised to see just how dominant dual DDR is in their lineup.

Daveman Deluxe
Aug 15, 2003, 03:44 AM
One thing that's being forgotten in all of this is the G5's use of direct memory access (DMA) technology. Every component has an independent 6.4 GB/s bus to the RAM banks. That takes a LOT of stress off of the system controller, which normally has ALL RAM operations routed through it.

Wait a sec....Apple's implmentation of DDR isn't crippled. You have to install the RAM in 2's in order to saturate the FSB- eg. 2 x 6.4 G/s = 12.8 G/s (am I correct? )

I did the math, and your math is impeccable, albeit for the wrong reasons. While it's true that the RAM controller will access two banks of SDRAM at once, the controller still only moves 128 bits of data at a time. By my math, that's (128/8)*400 000 000 * 2, which should be 12.8 GB/s. I'm not sure why it's not.

The other thing is that the ability to read/write to two banks at a time has little to do with data speeds--the bus can only handle so much regardless. It seems like it would instead reduce latency. I had a good idea going for why, but it escaped me. Hopefully somebody else can prove/disprove my thought.

As for one of the original complaints that 64-bit memory addressing should mean much more memory is addressable--the problem with that idea is the fact that a 22-bit initialization vector needs to be appended to the address, leaving 42 bits to actually address a RAM location.

Sun Baked
Aug 15, 2003, 04:23 AM
They make the simplest math confusing.

There's a reason they call DDR400 memory PC3200. ;)

If the memory controller is looking at two banks of the stuff PC3200 & PC3200 -- it magically equals Apple's estimate of GB/s, imagine that. :p

Also the reason for needing memory in matched pairs.

Daveman Deluxe
Aug 15, 2003, 05:09 AM
Sun Baked-

I didn't multiply by two because the controller reads from two banks at once--I multiplied by two because the controller transfers data on both the rise and the fall of its clock cycle.

(128/8) to convert from bits to bytes...
Multiply by 400 000 000 because that's the clock frequency of the memory controller...
Multiply by two because the controller transfers data on the rise and fall of the clock.

Which works out to 12.8 GB/s. There must be something I'm missing.

Cubeboy
Aug 15, 2003, 08:08 AM
Originally posted by illumin8
This is not 100% true. Rambus has a faster total throughput, but the latency is much higher than DDR, so short operations are much slower. Now that dual DDR is approaching 1ghz. (2x 500), there is no reason to buy the much slower overall and more expensive RDRAM 1066 sticks.

Ahh, but it is, the fact that Rambus has a faster total throughput allows it to keep up with fast FSB's (like the G5's and at it's time the P4's) better than DDR. This was largely the case with the P4 where systems equipped with PC1066 RDRAM consistently outperformed equivalent systems equipped with DDR333/266 SDRAM. I'm afraid you won't find anything in my post that isn't true, i850e/PC1066 RDRAM was Intel's flagship chipset/memory until the arrival of Dual Channel DDR solutions. RDRAM doesn't suck, it only needs a cpu who's FSB is fast enough to take advantage of it's strength (but not that of DDR). Morale of the story, if DDR is fast enough for your cpu's FSB, stick with DDR, if it's not and you have deep pockets, move to Rambus.

Cubeboy
Aug 15, 2003, 08:16 AM
Daveman Deluxe:

SDRAM/DDR processes information in 64 bit data paths, I think the reason Apple said 128 bit DDR400 was to signify that it was dual channel. Either that or the G5's using (proprietary?) memory thats unavailable anywhere else in the industry.

G4scott
Aug 15, 2003, 09:02 AM
I remember reading an article about RAM some 3 months ago, and they didn't like DDR400, because it was 1.) expensive, and 2.) only showed improvement in certain areas.

They still recommend slower RAM, because DDR400 is still new, and expensive

When BestBuy starts to advertise DDR400 memory in their ads, or even carry the stuff in their stores, then it'll be old.

Besides, rambus is expensive. If Apple were to add it anytime soon, the G5's cost would probably go up a good $500, and don't expect to buy any memory for cheap, especially from Apple.

ddtlm
Aug 15, 2003, 11:49 AM
Cubeboy:

Although the terms get misused a lot (by myself even), the difference between one 128-bit memory channel and 2 64-bit channels is in how they can be accessed. Dual channel implies that memory can be read separately on each channel, 128-bit implies that that they operate as one (they share address lines). I think most times people say dual-channel they really mean 128-bit, and not two imdependent 64-bit channels.

Daveman Deluxe
Aug 15, 2003, 12:17 PM
Originally posted by Cubeboy
Daveman Deluxe:

SDRAM/DDR processes information in 64 bit data paths, I think the reason Apple said 128 bit DDR400 was to signify that it was dual channel. Either that or the G5's using (proprietary?) memory thats unavailable anywhere else in the industry.

That's good to know, but the fact remains that data will still travel 128 bits at a time, twice for each clock cycle. Unless I'm missing something important about how dual-channel actually works, my math stands.

topicolo
Aug 15, 2003, 01:16 PM
Originally posted by Daveman Deluxe
Sun Baked-

I didn't multiply by two because the controller reads from two banks at once--I multiplied by two because the controller transfers data on both the rise and the fall of its clock cycle.

(128/8) to convert from bits to bytes...
Multiply by 400 000 000 because that's the clock frequency of the memory controller...
Multiply by two because the controller transfers data on the rise and fall of the clock.

Which works out to 12.8 GB/s. There must be something I'm missing.

You are missing something. Each bank is only 64bits wide so it should be
64/8 to get 8 bytes/cycle
multiply by 400 000 000 cycles and then by 2 (2 banks) to get 6.4Gb/s

Or, you could've just multiplied the 3200 in PC3200 by 2 and reached that much faster.
The 400 in DDR400 stands for the megahertz of the ram, while the 3200 in PC3200 stands for the throughput (in mb) of the ram.

Hope I cleared everything up

MacBandit
Aug 16, 2003, 12:41 PM
Originally posted by ddtlm
MacBandit:

Oops, didn't mean to tell you things you already knew. :( But anyway, not that I mean to carry such a minor arguement too far, but I just checked at Dell's site and apparently of the 5 models of PC for "home" use, the top 4 all have dual channel DDR RAM, the top two (Dimention XPS and 8300) being dual DDR-400 and a 800mhz FSB. I must say that I'm suprised to see just how dominant dual DDR is in their lineup.

Hey no problem. I didn't know about the current lineup of top end Dells. That's cool for the industry. I would say this is pretty new though probably just since the introduction of the 800MHz FSB P4s.

windwaves
Aug 16, 2003, 07:50 PM
I also had a terrible experience with RAMBUS at work. we needed a really fast machine for some computing intense stuff and some guys decided to go with one RAMBUS based. WTF'g piece of crap. I could not believe it. I have never seen a computer so unstable in my life, not a Mac, not a PC.

acj
Aug 16, 2003, 10:50 PM
Originally posted by windwaves
I also had a terrible experience with RAMBUS at work. we needed a really fast machine for some computing intense stuff and some guys decided to go with one RAMBUS based. WTF'g piece of crap. I could not believe it. I have never seen a computer so unstable in my life, not a Mac, not a PC.

Concluding that rambus is crap from your experience with one machine is ridiculous. One of the many PC's I've used that have never once locked up uses Rambus. It's overpriced though. Why spend 100% more for 3% more performance?

nuckinfutz
Aug 16, 2003, 11:29 PM
The DDR is fine. This guy sounds just dangerous enough with his misinformation to do damage.

The G5 Dual 2Ghz is not using the full bandwidth available. Each FSB to the processor could consume the whole 6.4GBps throughput. Unfortunately while each processor has it's own FSB they both meet the Memory controller which "talks" to the bank of DDR so you're halving the memory anyways.

The ideal situation would be either 2 Dual channel banks of RAM with two system controllers or ondie memory controllers. But we're talking alot more $$$ if you doubled the ram slots and added a beefier controller.

The statements on Altivec are just plane false. You don't have to recompile because the Altivec unit is the same as the 7400 chip. The G4+ had a slightly tweaked unit but they all function the same. Where the optimizations will likely be is taking advantage of the inflight intruction advantage and out of order execution benefits of the G5.

Perhaps the PPC 980(mythical) will have a improved Altivec unit but it may not as the dual FPU units will provide a nice boost to apps recompiled to take advantage.

Catfish_Man
Aug 17, 2003, 12:49 AM
There are a few vector cache touch instructions (DCBsomething) that are slow on the G5 and should be avoided, but overall it has shown (in benchmarks) to be roughly equivalent per clock to the G4 on vector code.

windwaves
Aug 17, 2003, 07:33 AM
acj, you do have a point - I simply did not say that after we bought our Rambus pc, we heard of many others having a bad experience with it. It was not just our PC that was crap. Still, I am not stating that this constitutes statistical evidence. However I am implying that MY experience has been crap. And in professonal computing world you may very well understand that individual experiences are often all it takes to determine one's feeling towards a product. And my feeling remains, it is crap.

In any event, I am glad your particular experience has been so good.

ZeppelinArmada
Aug 17, 2003, 08:11 AM
I worked at a company, which will remain nameless, and we had unending troubles with rambus. It was one of my duties to every monday at 9 A.M. to replace the rambus in our servers that died over the weekend, which it always, without fail, did. At least four design computers on our floor had their rambus smoke every week. It cost said company at least $10,000 dollars a week in rambus; this is no exaggeration. Pleading with management to get new computers was wasted effort as it would make them look like they had made a mistake in ordering them. Eventually one of the sticks of rambus fused to its slot and we got a new enterprise server. One down 11 to go.

Moral of the story bleeding edge is not always the best place to stand. DDR 400 is out it works and it is fast.

benixau
Aug 17, 2003, 10:24 AM
Originally posted by ZeppelinArmada
DDR 400 is out it works and it is fast.

makes my late celeron with PC66 look a bit old dont it?

makes an emac with only PC100 look a bit slow dont it?

makes a DP G5 with DDR400 dual channel RAM and all that 64bit stuff look like a good buy. now here do they put the monitor??????


:D really, i know my sh:t but i also know that the faster the FSB is and the faster the RAM is the faster the rest of the system will perform asa general rule.

HasanDaddy
Aug 17, 2003, 11:57 AM
plain and simple ---

a 1 ghz BUS makes a much FASTER computer!!!