TalkChess.com

corres wrote:Thanks for the detailed answer.
Based on the lot of tests what can be read about this subject I think that:
1, Behavior of a given PC is decisively determined by the CPU and the motherboard used and not the RAM.
2, In the case of very memory sensitive tasks the effect of RAM parameters on the power of a given PC is not more than 10 %.
3, Chess programs are more sensitive to the parameters of CPU and CPU cashes than the parameters of RAM. The only one parameter of RAMs
what important to a chess program user is its bigness in GB.

Best regards

Robert

I mostly agree with your statements. But I will note that #2 is not always true. In some case the latency or bandwidth, or a combination of both can be the dominant impediment to higher performance. In those cases higher ram performance equates to higher over-all performance. The CPU manufactures know this and are constantly looking for new ways to improve the performance of their CPU's by increasing the performance of their memory subsystems. e.g. if you were forced to use the same ram as was in common use 30 years ago modern CPU's wouldn't perform that much better than much older CPU's simply because they would be memory bound all the time. At least with the newer memory technology they're only memory bound most of the time.

Regards,

Forrest

Dann Corbit wrote:I am waiting for Naples.
With baited breath.
Give me Naples, or give me death.

Ok, maybe not that serious.

Uff for a moment you had me worried about you

Zenmastur wrote:The CPU manufactures know this and are constantly looking for new ways to improve the performance of their CPU's by increasing the performance of their memory subsystems. e.g. if you were forced to use the same ram as was in common use 30 years ago modern CPU's wouldn't perform that much better than much older CPU's simply because they would be memory bound all the time. At least with the newer memory technology they're only memory bound most of the time.

Please stop writing nonsense, this is getting more and more ridiculous.
Today's CPUs have for order of magnitude more cache than maximum addressable memory 30 years ago.

corres wrote: Based on the lot of tests what can be read about this subject I think that:
1, Behavior of a given PC is decisively determined by the CPU and the motherboard used and not the RAM.
2, In the case of very memory sensitive tasks the effect of RAM parameters on the power of a given PC is not more than 10 %.
3, Chess programs are more sensitive to the parameters of CPU and CPU cashes than the parameters of RAM. The only one parameter of RAMs
what important to a chess program user is its bigness in GB.

1. Correct
2. Correct in at least 90% use cases.
3. Yes up to the point of 16GB. Due to the way memory is managed in CPUs, more hash than 16GB often hurts performance, and improvement when going from 4GB to 16GB even when using large pages is not measurable.

Milos wrote:Please stop writing nonsense, this is getting more and more ridiculous.

And your supporting evidence for this ad hominem attack?

The fact is, there is nothing nonsensical or ridiculous about my post. What I wrote is a matter fact and can be easily checked by reading any of the various pages on the web that deal with the history of computer memory.

Milos wrote:Today's CPUs have for order of magnitude more cache than maximum addressable memory 30 years ago.

Yeah...

...Normally I would ask what your point was.

But considering the content of your previous posts, it seems like your ad hominem and glib response is specifically designed to hide your lack of knowledge on the subject.

I see no point in arguing with an unarmed person. So, I think I'll call it a rap on this thread.

Regards,

Forrest

Zenmastur wrote:

Milos wrote:Please stop writing nonsense, this is getting more and more ridiculous.

And your supporting evidence for this ad hominem attack?

The fact is, there is nothing nonsensical or ridiculous about my post. What I wrote is a matter fact and can be easily checked by reading any of the various pages on the web that deal with the history of computer memory.

Milos wrote:Today's CPUs have for order of magnitude more cache than maximum addressable memory 30 years ago.

Yeah...

...Normally I would ask what your point was.

But considering the content of your previous posts, it seems like your ad hominem and glib response is specifically designed to hide your lack of knowledge on the subject.

I see no point in arguing with an unarmed person. So, I think I'll call it a rap on this thread.

Point is discrediting charlatans that are giving ppl wrong information and pretending to be some experts but actually knowing nothing about the subject.
As this thread progressed your answers and "explanations" become more and more as if they were coming from some parallel universe. At the end you even ran out of those imaginary "facts" so the only remaining thing is to run around and shout ad hominem, ad hominem. Sorry but when someone says 1+1=5 and I shout nonsense that has absolutely nothing to do ad hominem. Your response just reflects your hurt ego.

I've read the entire thread, but found no tests run about chess performance. Have I missed something or hasn't anybody with high speed RAM bothered to check just how much they're getting out of their investment?

That's from over a year ago.

At the moment we're stuck with a few memory-chip makers that, given the current price levels, have no incentive to do anything.

I am fired up about DDR5. Hopefully they will make it.