Should be PC66, but PC100/PC133 are almost always backwards compatible.
hello guys and gals,
i hope everyone is having a good thanksgiving week. i will try to make this short and sweet. i am in the process of doing a memory upgrade. i am working on a aptiva. it has a 233cpu, 64 meg ram, and so on. the memory is 168 pin sdram. my question is, is it pc 66 100 or what. of course, unless you know something about reading the chips it does not tell you anything.if you need to know anything else please post. i will be watching real close. thanks for any advice in advance. DALE
"THE PASSION OF THE CHRIST"
thanks for the quick reply.
so are you saying that if it is pc66 and i install 100, it will still work but will run at 66. thanks again
"THE PASSION OF THE CHRIST"
It will run at whatever the speed is set to in the BIOS. In this case the memory that is in there is most likely PC66, but could be PC100.
Either way, a 233 CPU has a 66mhz bus, so that's what the memory should run at and may be the only available setting.
The most popular misconception that I've seen regarding memory is that its running at a certain speed is an active process, like the memory determines the speed at which it runs.
The clock generator on the mobo determines at what speed everything runs. As long as the memory meets or exceeds that specification, it's fine.
The memory is just along for the ride, so to speak and doesn't decide what speed to run at. Some motherboards can set the memory bus based on what the memory identifies itself as, but that's another issue...
A comment on hiredgoonz reply:
Although you are correct about the misconception that the memory defines what speed the CPU operates at, in many cases, the speed of the memory WILL affect the CPU speed. Let me elaborate with an example.
I'll use the good ole' 8086 uP since that's the one I'm most familiar with.
If your processor is running at 10MHz, then each clock cycle would be 100ns. If performing an internal operation on the processor (say an add or multiplication of internal registers), then your theory that the memory speed does not affect the speed of the CPU is true. A simple addition of two registers will take about 3 clock cycles to perform and is 300ns regardless of your memory's speed.
When interfacing with memory, however, we must take into consideration all of the timing delays of memory to read and/or write data from/to the uP. The minimum delay for a memory read/write cycle is 4 clock cyles to complete. This is called a bus cycle. With a 100ns clock period, the entire read/write operation must be performed within 400ns of the start at the first clock cycle. As long as we have a fast enough memory (about less than 250ns is a good estimate depending on the propagation delays of address latches, data tranceivers, etc.), then we can disregard the speed of the memory and no matter how fast it is, it will not affect the CPU's bus cycle.
It is when we have a slower memory that the CPU bus cycle suffers. This is a fundamental problem (and a good place for computer manufacturers to lie about their computers' overall performance). If you have a slower memory device that will not meet the minimum timing requirements of a single bus cycle, then you need to do one of three things to force the read/write operation to meet the timing requirements. You can:
1) Slow the clock speed down a bit
2) Get a faster memory device, or
3) Add wait states in which the CPU does nothing
Most often, wait states are added in between the third and fourth clock cycles in a bus cycle since it's the most cost effective solution which doesn't affect the ever-so-important clock speed specifications. Essentially, these wait states are used up clock cycles eating up precious CPU time to allow the memory to read/write valid data. If we have a very slow memory device which takes, say 400ns to do a complete read/write, then we'll probably have about 3 wait states added to the bus cycle causing a 400ns bus cycle to increase to a 700ns bus cycle. That's almost halfing the speed at which read/writes can occur! In this case, obviously the speed of memory will affect the overall speed of the CPU. Since memory reads and writes are the essence of what a x86 does best, memory speed actually does affect the overall system speed.
So, since I've made this way too long now, I'll add the note that this was an example from the very first IBM PC so the timing analysis won't be comparable to current technology (or even old technology from those blazing fast 486s!). The example is still relevant; however, since these basic problems are still the principle problems with any RAM/ROM devices. In fact, as CPU clock speeds have exponentially increased, memory speeds are lagging even more nowadays.
Hiredgoonz, you are definately correct when you tell us that the clock speed on the motherboard defines the speed at which everything runs, but that only corresponds to the clock cycle speed, not bus cycle speed. Hopefully this hardware oriented reply was of some use to understanding the issue.
Damn, I have way too much time on my hands.
And this affects the user in what way when determining between PC66 and PC 100 memory?
Your lengthy explanation really has no bearing on what the original poster was asking.
As I stated in my post, I was referring only to the speed of the memory being set by the clock generator IC on the mobo.
There are a number of issues that I could have addressed, such as the various latency settings for memory, but they would have NOTHING to do with the question or in any way help tad214 determine if he needs PC66 or PC100 memory...
Your post may very well help someone understand memory better, but for my part, I try to answer the question that was asked. Since I work in the IT field, odds are I have a lot of information that might be of limited interest to a poster, but I try to stick to answering the question concisely in order to best help the person having trouble...
You're absolutely right that it has no bearing on the original posting. When I was skimming the thread, your statement struck me when you said "The most popular misconception ... like the memory determines the speed at which it runs.", because I know it does if you dig deep enough.
I wasn't trying to attack your explanation, and I certainly didn't think it would help anyone determine what memory device to choose. But being a discussion group, I was merely trying to add some discussion to the board so-to-speak. It just dawned on me that since I'm such a hardware oriented person, I may be looking to start discussions such as this in the wrong place. The main reason for looking at the posts is for my own general knowledge and I hoped that other members were able to contribute what they know just as I did.
I'll take your advice to post relevant info to the poster's thread from now on and be in the search of other discussion boards for my interests. And I'm new on this site, so I'll work on the length a little
I'm not trying to discourage you from posting here and I apologize if I offended you in some fashion.
However, with respect to what determines the clock speed at which memory runs, it is the BIOS setting, not the memory itself. Although as I stated, some systems can set this speed based on what the memory identifies itself as...
This has nothing to do with clock cycle length or wait states.
I was just trying to point out that a lack of information in a reply may have less to do with what an individual knows or doesn't know and more to do with simply answering the question at hand.
Rereading my post I can understand if it seemed rude, so again, sorry for that and dpn't let me discourage you from posting here...
Also, WPIEngineer, keep in mind that it is often a good idea to start your own thread to discuss hardware issues or any other topic that interests you. There are plenty of people here who have the knowledge and experise to follow what you are talking about and engage you in interesting discussion. (I don't happen to be one of them, I didn't follow a lot of your post) This forum is a good place to get general discussions about computer topics going:
I am interested in what you have to say from your perspective, even if I won't always understand it.
This is what you call "A true blue discussion forum"..and I certainly enjoyed the exchange between savvy individuals. As a person just started out into the field of "computer savvy", I want to say "keep it up!". It broadens our understanding of the nuts and bolts of the computer.
May you all have a happy continuence Rudedog
There are currently 1 users browsing this thread. (0 members and 1 guests)