And Finally- A Processor worth more than its Value!

[Ray Proudfoot 9,038]

14 minutes ago, Rob Ainscough said:

You've made a terrible mistake, something faster did come out last week … but was sold out before it was actually available online. 😉  You worry too much Ray … say hello to Pete from me and enjoy.

Cheers, Rob.

I bet you got the last one Rob! 😃 Thanks. I’ll pass on your best. Pete thinks this will be the last trip to Lelystad. Even with a two night stopover the day at the show is very demanding. You should go one year. Some serious cockpit builders there.

[dave2013 1,148]

5 hours ago, w6kd said:

That Crucial white paper has been around a long time, and I found that its primary fault is that it throws a bunch of facts up on the wall, and then draws a conclusion that isn't supported by those facts...no real logical linkage from data to conclusion.  They are trying to advance the mantra that faster memory clock speeds are better, which universally holds true only if you hold the CAS latency constant.  Why?  because then they can charge premium prices for average memory by clocking it up in conjunction with higher CAS latencies.  It's analagous to the question of whether it's better to take the fast train or the slow train...faster is certainly better if the number of stops is the same, but if not, you have to do the math, because you may well find that the faster train makes more stops than the slower one, and in fact doesn't get you there any faster (or indeed maybe slower) as a result.

The memory read process involves putting the address up on the address lines, issuing the read request, then waiting for the data to be ready to be accessed on the data bus.  The process is synchronized by a sort of a metronome...the clock, and the speed at which it ticks is the memory's "speed" in MHz, or millions of ticks per second.  There is a waiting period--latency--where the memory electronics have to be given time to reliably change states and put the data on the bus to be read.  The memory controller observes this latency by counting the clock ticks from the request--CAS wait states--and then after the configured number of ticks have occurred, it reads the data from the bus.  That latency is the real measure of how fast your memory is...just as in the train example, if your clock ticks faster, but you have to wait more ticks for the data to be ready, it isn't necessarily available on the bus any faster than with a slower clock and less wait cycles.

As far as the question of whether faster memory makes a difference in P3D...I think that it can, especially with today's fastest 6-10 core CPUs.  Not in terms of frame rate, but a noticeable reduction in stutters when the CPU and GPU loads are not maxxed out.  If the sim stutters when the CPU is hitting 100%, especially on the core running the primary thread (usually core 0), then it's expected that stutters may occur due to the CPU not being able to keep up with the demand for steady frame production.  If it happens with the GPU(s) at/near 100%, same thing...the GPU is bottlenecked.  But I see stutters, especially in autogen-intensive areas like ORBX SoCal, even with CPU and GPU loads well below 100%.  I suspect those are due to the CPU being "starved" of data, perhaps when it's trying to gulp a large block of autogen objects because the memory isn't able to keep up.

I'm seeing a nice reduction in those small stutters...still get an occasional one, but much less.  Most of the conventional wisdom--that memory performance makes no difference--seems to me just parroting of experience from previous generations of hardware that were running much slower and feeding much less voracious graphics subsystems.  It stands to reason that as CPU throughput increases, memory has to keep up, or the CPU will spend a lot of time idling and waiting for data.  Memory performance hasn't been increasing nearly at par with CPU throughput, especially as more cores are brought into the mix--all of which feed from a common memory subsystem.  So I take "memory speed doesn't matter" as cliche and somewhat suspect now.  Does that mean we should break the bank on exotic high-speed DRAM?  No, but OTOH building a power system with average consumer-grade memory may prove underwhelming, too.  A balanced system, as has been said many times here, is the key.

Regards

 

Thanks for this explanation.  It has always made sense to me that faster memory would increase simulator performance.  The question is "how much" in terms of fps or smoothness, but having memory that can better keep up with the cpu is bound to benefit performance.

Dave

[joemiller 269]

12 hours ago, martin-w said:

 

Great result Joe! What sort of temps are you seeing at 5.3 GHz? Voltage? Have you delidded?

My 8700K does 5.2 GHz with HT off, it's delidded. Won't do 5.3 though. Although I haven't tried voltage at 1.5, the max Intel recommend. 

Is that 4266 Mhz Ram the XMP setting, or did you push it beyond the XMP setting? 

 

Yes, I delided my 8086K. I'm using 1.42V. The RAM was manufactured at 3600hz, but I overclocked it to 4266hz (speed and latency). 

[joemiller 269]

12 hours ago, Avidean said:

I don't recall exactly how to do the math but the Mhz value alone tells you nothing about the speed of memory.

It always has to be considered in conjunction with the Case latency. The lower the CL value the better. so indeed 3200 mhz can be faster than 4266mhz.

There is a lot of marketing hype with memory Mhz values.

Memory did make a difference with FSX. The faster the memory the less long frame stutters you would get.

I am not sure if that still holds for P3D but I would guess that it does.

Fast memory is important at the lowest latency possible.. 

[joemiller 269]

12 hours ago, Ray Proudfoot said:

The formula for determining "true latency" is:-

2000 x (CL / Tested Speed)

For a 3200 CL14 set that will be 2000 x (14 / 3200) = 8.75

For a 3000 CL14 set that will be 2000 x (14 / 3000) = 9.33

For the fastest memory - 4000 with a CL of 19 the result is 9.50

So my 3000 CL14 memory is faster than 4000 CL19. And £238 cheaper!

Well, there is a reason for making faster RAM... Means things will run faster. Latency is important as well as it determines how quickly it will hand over (data) to pass it along.

[joemiller 269]

6 hours ago, Charlatan said:

What addons are you using with P3D?

The most demanding aircrafts and  sceneries  currently in the market (pick your best) 

Edited November 2, 2018 by joemiller

[joemiller 269]

 

 Well, speaking of this animal ( I-7  8086k) I bought mine at Microcenter. For many years I used to be a Frys Electronic guy. But once I discovered Microcenter, never saw the need to go back to Frys. Why? Here is why:  (well, for those who live in the U.S)

 Every person as Microcenter is very knowledgeable  

Everyone is up-front helpful

 They offer (sell)  total protection warranty for every product. If it breaks bring it back..no questions ask. 

 And, best of all.... They sell every product significantly lower than anyone on earth or online. 

 

So, initially I was waiting on the 9900K; however, after seeing how well the 8086 overclocks past 5Ghz, I saw no need for the 9900K. And, I am extremely pleased. Keep in mind our sim loves a higher clock speed-   Not more cores!  

Edited November 2, 2018 by joemiller

[Charlatan 68]

8 hours ago, joemiller said:

The most demanding aircrafts and  sceneries  currently in the market (pick your best) 

I take it you are running orbx, pmdg, Majestic, Aerosoft Airbus etc then?

[martin-w 3,256]

18 hours ago, Ray Proudfoot said:

He’ll probably recommend G.Skill but as to what speed who knows after this morning’s discussion. 😁

Mobo? Everyone seems to go for Asus.

 

I'm feeling less playful this morning, so I will answer that without reference to alien tech from Area 51. I could become more playful later on today though.

Regarding speed... my opinion hasn't changed one iota.

As we all seem to agree, Crucial are correct when they state that "true latency" hasn't changed much over recent years, so the question to ask is simple... does the higher frequency RAM that's currently available to purchase result in greater performance? And as the reviews of faster RAM tell us, yes it does, for many applications. 

So... as I maintained at the beginning of this RAM debate and in the other thread, the faster RAM that's available to purchase can result in  "a small increase in performance". It may only be small, dependant on the application, but nonetheless, that small increase in performance is present.

Thus, the question for a would be RAM purchaser to ask is if he/she is prepared to pay a premium for a small increase in performance. Many enthusiasts would, some flight sim enthusiasts and gamers might not.

Our own member Saab340 looked at RAM speeds a while ago in FSX and determined that faster RAM is beneficial. If anyone's interested in his results a quick search should prove beneficial.

I would be very interested in Westman's finding in regard to this, as he has done more testing than we have had hot dinners.  

For my last three builds, two for me and one for my daughter, I went for GSkill Ripjaw V 3200 MHz. Samsung B-Die. Decent price and has proved very reliable for me and many users. 

 

Edited November 2, 2018 by martin-w

[Ray Proudfoot 9,038]

19 minutes ago, martin-w said:

I'm feeling less playful this morning, so I will answer that without reference to alien tech from Area 51. I could become more playful later on today though.

Regarding speed... my opinion hasn't changed one iota.

As we all seem to agree, Crucial are correct when they state that "true latency" hasn't changed much over recent years, so the question to ask is simple... does the higher frequency RAM that's currently available to purchase result in greater performance? And as the reviews of faster RAM tell us, yes it does, for many applications. 

So... as I maintained at the beginning of this RAM debate and in the other thread, the faster RAM that's available to purchase can result in  "a small increase in performance". It may only be small, dependant on the application, but nonetheless, that small increase in performance is present. 

Martin, "a small increase in performance" is quite vague. How exactly would the user see that? Increased fps? I am sceptical about that. Less stutters perhaps? Seems more likely as data is being moved very slightly quicker. But in my example on the previous page the increase is 0.57ns.  1 million nanoseconds = 1/1000th of a second. Or maybe that's 57 nanoseconds but it's still tiny.

So in my case the difference is so negligible I don't see how it could possibly affect performance. If you swapped from the very slowest to the very fastest maybe there would be a discernable change. But no-one here is talking about installing the slowest memory - just decent stuff for a balanced system.

Off to Netherlands shortly so any reply will be much later today.

Edited November 2, 2018 by Ray Proudfoot

[martin-w 3,256]

48 minutes ago, Ray Proudfoot said:

Martin, "a small increase in performance" is quite vague. How exactly would the user see that? Increased fps?

 

Yep, an increase in frame rate. It can vary quite a bit. For example Assassins Creed 1080p at 2133 CL13 to 4000 CL19 an 11% improvement. But at 1440p only 2% improvement in frame rate. 

BF1 at 1080p 2133 CL13 to 4000 CL19 only 1% increase in performance. And in fact at 1440p it loses frame rate.

So vary variable dependant on the application, your refresh rate etc. 

 

Quote

Less stutters perhaps?

 

The problem is that there is no metric to measure stutters, so the placebo affect can come into play. "Do you know I think I see less stutters" is hardly scientific. So just subjective opinion. I'm not saying less stutters is wrong, just hard to validate. 

As i said earlier, you have no issues Ray. Your RAM at 3000 MHz will be absolutely fine and I have no doubt you will be thrilled with your new rig. And as I said, if some time in the future you're feeling brave, you can nudge up the RAM a few dividers as an experiment. Mine is rated at 3200 but runs fine at 3400, I would think I could take it higher than that.

 

Quote

Off to Netherlands shortly so any reply will be much later today.

 

Have a great time, and guess what..... you are beating my daughter to it by one day. She too is on her way to the Amsterdam tomorrow!" Dash 8 Q400 I presume, so be ready for the rapid acceleration on takeoff, the noisy engines and definitely the rattles and squeaks. I do love the Dash 8. 

 

Keep away from the marijuana cookies! 

 

Edited November 2, 2018 by martin-w

[fghdgdfdfgfgf 111]

3 hours ago, martin-w said:

 

Yep, an increase in frame rate. It can vary quite a bit. For example Assassins Creed 1080p at 2133 CL13 to 4000 CL19 an 11% improvement.

 

Yeah, that falls in line with my 10% performance boost I claim to get from 3200 c14 to 4133 c17 at LAX.

3 hours ago, martin-w said:

 

 But at 1440p only 2% improvement in frame rate. 

.... And in fact at 1440p it loses frame rate.

So vary variable dependant on the application, your refresh rate etc.

Yeah, you think we might be gpu limited in these situations ...C'MON MAN!  ...Oh, and please tell us know how BF1 relates to P3D 4.3.   <g>

3 hours ago, martin-w said:

The problem is that there is no metric to measure stutters, so the placebo affect can come into play. "Do you know I think I see less stutters" is hardly scientific. So just subjective opinion. I'm not saying less stutters is wrong, just hard to validate.
 

What!?!?  (long frames can be measured)    .......and then you write this .....(see below)

4 hours ago, martin-w said:

Our own member Saab340 looked at RAM speeds a while ago in FSX and determined that faster RAM is beneficial. If anyone's interested in his results a quick search should prove beneficial.

Is Saab's analysis valid or not?????

 

Moderator  ....can I request that Martin (and other like minded persons) be forever locked in this thread ...so as to keep him (and other like minded persons) from posting else where on these sacred boards.  vbg

Edited November 2, 2018 by FunknNasty

[martin-w 3,256]

Quote

Yeah, you think we might be gpu limited in these situations ...C'MON MAN!  ...Oh, and please tell us know how BF1 relates to P3D 4.3.   <g>

 

Steady Funkmeister steady! It's not me saying that my boy, it's testing from a review... I'll find the link for you so you can complain to the author and treat him to your wisdom. 

BF1 does not relate to P3D, who said it did? That was merely an example of how variable frame rate increase is as a result of memory speed increase. Get with the program Funkyboy... it was a general debate regarding RAM speed, true latency and all manner of fascinating things. Not just relating to flight sim. Which was why I said... "So variable dependant on the application". APPLICATION!

 

Quote

What!?!?  (long frames can be measured)    

 

How do you measure stutters accurately then Funkypants? Please enlighten?

 

Quote

.......and then you write this .....(see below)

 

And THIS was about Saab340's tests that revealed "higher Frame rate" as a result of faster RAM. He may have referred to stutters in his post, but I'm not referring to stutters. Easy really.

 

Quote

Is Saab's analysis valid or not?????

 

For frame rate I would say yes. Did he mention stutters? you tell me, I wasn't referring to stutters. have you been smoking something funny today? 😉

You can either apologise or send me a nice Christmas present, I don't mind which. I'll put it down to a temporary burst of a young man's testosterone. 🙂

 

 

Edited November 2, 2018 by martin-w

[martin-w 3,256]

Quote

Moderator  ....can I request that Martin (and other like minded persons) be forever locked in this thread ...so as to keep him (and other like minded persons) from posting else where on these sacred boards.  vbg

 

Now I'm getting annoyed! There's no need for that kind of behaviour. 

 

P.S. Elsewhere is one word. 

Edited November 2, 2018 by martin-w

[fghdgdfdfgfgf 111]

15 minutes ago, martin-w said:

 

P.S. Elsewhere is one word. 

You know, I was going back and forth on that before hitting the send button.   🙂