So what really is faster...3.8GHZ CPU but NB at 2000 or 3000MHZ?

A teaser showing what i have been upto...1333 6-6-6-24 1T but with the Nb clock at 2000, 2200, 2400, 2600, 2800 and 3000 with CPU @ 3.8GHZ as this is where the bulk of air and watercooled rigs will land.

Have 1600 tests going on the bench next and then will do the whole lot again looking to see if HT clocking has an effect. I will add 3Dmark vantage into HT clocking also if needed.

As mentioned i have only done ram at 1333 to start, super pi 32m and AutoGK encode DVD FILM 720X480 800MB VOB up scaled to 1280x960 AVI 160MP3 sound 350MB file.

time is converted to seconds for both test, lost the decimal places off PI as they were not needed ;)

TEST WITH RAM AT 1333 6-6-6-24 1T

http://dl.getdropbox.com/u/920660/NB...ncode_1333.png

TEST WITH RAM AT 1600 7-7-7-24 1T

http://dl.getdropbox.com/u/920660/NB_clock1600_test.png


1333 RAM RESULTS:
32M pi 35secs faster with NB clock at 3000 over 2000
encode 47secs faster total job NB clock at 3000 over 2000

1600 RAM RESULTS
32M Pi 36 secs faster with NB at 3000 over 2000
encode 57secs faster total job NB clock at 3000 over 2000

Finally i know now 1333 6-6-6-24 1t is faster for 32m Pi than 1600 7-7-7-24 1t ;)

made life easy for you...shows results a little clearer now adding cas5 1333 And 1600 CAS6-6-6-24 1T now done.

http://dl.getdropbox.com/u/920660/AM3_Nb_test.png

5-5-5-24 1t wins it here for me as its easier to do with most ram.


Now for those thinking HT clocking matters for 3D....i have a shock for you. 3870X2 stock clocks. 3dmark vantage
CPU @ 3.8GHZ, NB@ 3GHZ, RAM @ 1600 7-7-7-24 1t

http://dl.getdropbox.com/u/920660/HT_3dmark.png


Decided to do some CPU-NB speed Vs voltage testing, board is the Asus M4A79-T with Vista 64. Note using this OS its much harder to OC, especially once I installed all updates. if you are building a gaming rig Vista 32 or W7 32 maybe be a better bet for easy clocks and low voltages with 4GB ram.

Anyway, specs are:

stock 3.2GHZ at 1.20V on the CPU
2000 to 3000NB
memory @1333 6-6-6-24 1T 1.5V fo all tests kit used was 1600Platinum Animal

http://dl.getdropbox.com/u/920660/CP...e_Vs_speed.png

As you see voltage scaled linear till i got to 2700 where it needed a massive jump, I got away with 1.4125 on V32 so it just shows v64 is harder on the CPU's.

really nice work.

Quote:

---

Originally Posted by demonkevy666

really nice work.

---

Xcellent work Tony!:clap:

very interesting :up: .. it was the thing i wondered about over many days :D I guess 2,8ghz NB is good enough for >4ghz

Nice work. Next I'd like to see some memory benchmarks with ram at high fsb and high latencies vs slower fsb and lower latencies.
I appreciate it when people take the time to do this sort of investigation. Very useful information.:up:

yea seems to scale pretty linearly.

Great work! Tony seems to be falling in love with his Phenom II. They seem simple at first glance, but are in actually very complex and rewarding to tune. The Northbridge sweet spot was a pleasure to find back when I first got mine in January. NB does just as much or more for performance as core clocking does. I will bump mine from 2.4 to 2.8 on air to see how it handles it soon.

But how fast is your ht-link? You know you want to run it a 3ghz with that nb speed;).

Funny story about that though, normally I run my system at about 4.1ghz with the nb and ht-link around 3.1ghz. I came home from college last week and because of space issues I have to run my computer in our office rather than my room. Now these settings worked for an entire semester, 2 months staying at full load with gimps for prime 95. Turns out my computer doesn't want to turn on because it doesn't get enough power from the wall for everything.

But back on topic, good work Tony. I bet if I was running some ddr3 like you I would be hitting like 15gbps in everest for memory bandwidth instead of only 10gbps. Until then I am limited to my "slow" ddr2.

HT Link speeds don't add hardly any performance after overclocking. I keep mine at 2Ghz and that's as much as I need to push it.

Quote:

---

Originally Posted by Mechromancer

HT Link speeds don't add hardly any performance after overclocking. I keep mine at 2Ghz and that's as much as I need to push it.

---

dont even touch mine i havent found a time when overclocking it really ever helped. id rather have more stability.

Quote:

---

Originally Posted by Oldguy932

But how fast is your ht-link? You know you want to run it a 3ghz with that nb speed;).

Funny story about that though, normally I run my system at about 4.1ghz with the nb and ht-link around 3.1ghz. I came home from college last week and because of space issues I have to run my computer in our office rather than my room. Now these settings worked for an entire semester, 2 months staying at full load with gimps for prime 95. Turns out my computer doesn't want to turn on because it doesn't get enough power from the wall for everything.

But back on topic, good work Tony. I bet if I was running some ddr3 like you I would be hitting like 15gbps in everest for memory bandwidth instead of only 10gbps. Until then I am limited to my "slow" ddr2.

---

HT frequency actually hurts performance up high without very very high clocks...

Tony I hope you don't mind me adding to this........

Here we have 2 sets of ddr 3 clocked........one set is using 1.65v........the other set is a favorite of benchers :rolleyes: D9's..............using 2.0+v obviously one set is well within safe ranges for 24/7 use and is a 2x2g set so perfect for the daily driver......

The other set is a 2x1g perfect for benching if you like $200 + sticks....... lets see what that extra $120 dollars spent on a set that will collect dust unless benching will net you.....in a synthetic bench.........

Please take not I could not match them perfect and the high volt sticks have a NB frequency advantage which could have swayed results just a little.

Low voltage 2x2g
http://members.cox.net/wmdieselmc26/efficiency.JPG

Stupid voltage
http://members.cox.net/wmdieselmc26/efficiency2.JPG

Just what I wanted to see. Thank you, Tony.

A quick question. In the encoding graph between 2800MHz and 3000MHz - I assume the reason performance didn't increase is due to memory frequency? I'd imagine DDR3-1600 with tight timings would continue to improve performance in a linear fashion.

Quote:

---

Originally Posted by Oldguy932

But how fast is your ht-link? You know you want to run it a 3ghz with that nb speed;).

---

HT Link doesn't matter that much, its the same as back with K8 its overdimensioned. Matters for servers but not desktop users... My HT link currently is 1700mhz-ish something...

OT: Nice work Tony, eagerly awaiting more results! Can't you try out those DDR2 Blade modules at 1200mhz and watch the difference too?

Great info! Thanks for putting these vs. into perspective for a noob like me :D

Great stuff Tony! :up:

Thanks for the info, going to have a play and see what I can get out of mine when I get home. Interesting about the HTT maybe thats why my scores seemed lower with a higher htt with CPU @ 3.9GHz than with default htt.

priceless Tony!!

nice Tony...my idea is NB about 2600-2800MHz.

Quote:

---

Originally Posted by Smartidiot89

OT: Nice work Tony, eagerly awaiting more results! Can't you try out those DDR2 Blade modules at 1200mhz and watch the difference too?

---

+1 I would also be very curious to see similar numbers for DDR2 since I run 8Gig of it at 1066 atm and am on the fence if I want to switch to AM3.

Glad to see you so active in here btw Tony, thanks for all these numbers!

I will do the DDR2 stuff using the DFI also, for now im concentrating on the AM3 boards as this is what AMD are pushing and i like to keep partners happy ;)

i will have to reinstall on the DFI as I have W7 64 on there now...AM3 is running Vista 32bit so to keep apples to apple work is needed...OR are you guys not bothered?

Quote:

---

Originally Posted by Tony

OR are you guys not bothered?

---

yes we're! :D

first do 32bit, and afterwards 64bit ;)

Quote:

---

Originally Posted by Nedjo

yes we're! :D

first do 32bit, and afterwards 64bit ;)

---

+1

64-bit is more interesting imo thou.

so you guys going to fund all this work? each loop takes over 1 hr at each speed...6 loops = 1 whole day.

Quote:

---

Originally Posted by Tony

I will do the DDR2 stuff using the DFI also, for now im concentrating on the AM3 boards as this is what AMD are pushing and i like to keep partners happy ;)

i will have to reinstall on the DFI as I have W7 64 on there now...AM3 is running Vista 32bit so to keep apples to apple work is needed...OR are you guys not bothered?

---

I for one am simply curious at what speed for DDR2 (perhaps up to 600Mhz) the NB stops giving much return. I want to clock the NB as high as I can when still getting results, but if it's net result is simply increasing my temps I would rather keep it modest.

Updated to include 1600 results

Quote:

---

Originally Posted by Tony

Updated to include 1600 results

---

Thank you much again, Tony. It's exactly what I wanted to see. I think I'm starting to get to know about PII. And probably you've known about it all along. :P My hypothesis regarding PII architecture from your tables above:

1. When data can be replenished from L3, K10's performance keeps increasing to NB frequency, until it reaches the same frequency as the core frequency.
2. When data needs to be fetched from system memory, increasing NB frequency alone will give diminishing returns unless memory can keep up with it. Ideally it should be [NB frequency = 2 x memory frequency] (dual-channel)

So in my imagination, a 'perfect' 955BE would be 3.2GHz core / 3.2 GHz NB / dual-channel DDR2-1600 or higher.

Does this make sense to you, or am I totally wrong?

Quote:

---

Originally Posted by Tom128

I for one am simply curious at what speed for DDR2 (perhaps up to 600Mhz) the NB stops giving much return. I want to clock the NB as high as I can when still getting results, but if it's net result is simply increasing my temps I would rather keep it modest.

---

If my theory is correct the efficiency peak would be 2400MHz NB with DDR2-1200, although there would still be benefit of faster NB or faster memory. (more bandwidth -> lower latency)

As you see from Tony's table, pi calculation improves along with NB despite memory lagging. So if what matters is the traffic between core and L3, the higher NB the better the performance.

On the column of tables, we see the performance flattens with DDR3-1333 past NB frequency 2.8GHz, which is rectified by DDR3-1600. So it tells me DDR3-1333 is not feeding enough data to NB running @2.8GHz. I'd assume DDR3-1600 would be good up to NB 3.2GHz and at that point we would see diminishing returns once again..

tony can you test cas 5 at 1333 ? :)

Quote:

---

Originally Posted by Enoc

tony can you test cas 5 at 1333 ? :)

---

cas6 at 1600 is easier but the results will be the same...

higher the NB clock the more performance...2600/2800 is enough for 1333, 2800/3000 is enough for 1600.

Silly think is 2600NB is dead easy...why AMD don't preset the NB to 2600 i have no idea :shrug:

Quote:

---

Originally Posted by Enoc

tony can you test cas 5 at 1333 ? :)

---

I tested cas 5 1400+, its fast.......but does not scale as well....I think your better off with tony's reccomendations ;)

Updated to include HT clocking and 3D

nice funny to see that HT overclocking lowers the scores. doesnt make sense to me but w/e. im gonna try underclocking my HT and see what happens. also btw the left side of the HT graph is messed up. should say score instead of HT overclock.

Quote:

---

Originally Posted by roofsniper

nice funny to see that HT overclocking lowers the scores. doesnt make sense to me but w/e. im gonna try underclocking my HT and see what happens. also btw the left side of the HT graph is messed up. should say score instead of HT overclock.

---

check again ;)

Quote:

---

Originally Posted by Tony

check again ;)

---

lol cool. got an explanation for why the scores get lower as the speed goes up? also you planning to do any more tests?

will try quad Xfire next to see how that does...if its the same Sampsa is wasting his time clocking HT high for his 3d runs

just ran the canyon test of 3d mark at 1400 1600 1800 and 2000 and all of those were normal on my 4870. scaled linearly from 1400 to 2000.

very nice work :)

Tested 2x 2870X2, No gains with HT, tested PCIE OC single card to 125, tiny gain, with HT OC no gain. Dual cards PCIE 115 no gains, with HT no gains.

Weird thing is 115PCIE would not boot with HT 2000, at 115 dual cards but would run OK with HT at 2600...i think there is something wrong with bios on the Asus...at least PCIE and SATA are decoupled as the SSD had no issues even at 125 PCIE.

Not got good single cards...im wondering if 2x 4890's would gain thru PCIE OC and then HT OC on top.

well the ht thing is interesting that as you overclock it it loses performance. from the simple things i did from 1400-2000 it gained every step. i have seen others lose performance too by overclocking but maybe if i push it past 2000 i would start to see performance drops.

Lets go back a yr, 780i HT link between NB and SB, running 2600 or 2800HT along with PCIE at 141 with dual cards made a massive gain...BUT things were different.

1 the NB supplied 16X and the SB supplied 16x PCIE so an overclock to the PCIE bus would yield gains if the bus connecting the domains had a frequency increase.
2 790FX supplies both 16X slots, so the ONLY gain we should see with the video cards would be PCIE clocking until the transfer of data from PCIE to HT was the bottleneck then an increace in HT will be needed from the NB to the CPU.

So...why does increasing HT lower 3D performance...if anything an increase in HT bus should do nothing if the bus is already not the bottleneck?

Something is not right here

really i dont get how overclocking can ever lower performance. i know people have had issues with running the nb too high sometimes and getting less performance and just like you and others have tested with the ram. even if u run 1600 at c6 its still slower than 1333 c6. this phenom II platform has just been confusing.

Quote:

---

Originally Posted by Tony

Lets go back a yr, 780i HT link between NB and SB, running 2600 or 2800HT along with PCIE at 141 with dual cards made a massive gain...BUT things were different.

1 the NB supplied 16X and the SB supplied 16x PCIE so an overclock to the PCIE bus would yield gains if the bus connecting the domains had a frequency increase.
2 790FX supplies both 16X slots, so the ONLY gain we should see with the video cards would be PCIE clocking until the transfer of data from PCIE to HT was the bottleneck then an increace in HT will be needed from the NB to the CPU.

So...why does increasing HT lower 3D performance...if anything an increase in HT bus should do nothing if the bus is already not the bottleneck?

Something is not right here

---

it's not just connected with nb/L3 cache between cores ???

there is some inter connecting Ht between cores right.

HTT speed increase on should change those, but if can't by pass the nb/l3 cache it isn't going to improve anything.

anyone got some old bios to turn off the L3 cache ? maybe then we can see lol

Looks like I can keep my NB speed high enough isn't bad idea:D

Quote:

---

Originally Posted by demonkevy666

it's not just connected with nb/L3 cache between cores ???

there is some inter connecting Ht between cores right.

HTT speed increase on should change those, but if can't by pass the nb/l3 cache it isn't going to improve anything.

anyone got some old bios to turn off the L3 cache ? maybe then we can see lol

---

yea i think when you overclock HT it also overclocks the HT connecting the cores as well.

Quote:

---

Originally Posted by lopri

Thank you much again, Tony. It's exactly what I wanted to see. I think I'm starting to get to know about PII. And probably you've known about it all along. :P My hypothesis regarding PII architecture from your tables above:

1. When data can be replenished from L3, K10's performance keeps increasing to NB frequency, until it reaches the same frequency as the core frequency.
2. When data needs to be fetched from system memory, increasing NB frequency alone will give diminishing returns unless memory can keep up with it. Ideally it should be [NB frequency = 2 x memory frequency] (dual-channel)

So in my imagination, a 'perfect' 955BE would be 3.2GHz core / 3.2 GHz NB / dual-channel DDR2-1600 or higher.

Does this make sense to you, or am I totally wrong?

---

Your theory would be wrong.

I tested equal NB to cpu ratio.......the NB must not match or exceed cpu frequency or it takes a rather large performance hit.... for whatever reason it is not happy with a 1/1 ratio......give me a few and I will show you...

Once again I hope you don't mind Tony

For the sake of saving my poor CPU I used at PH II 925 for example.....

So here we have 1/1 ratio with CPU and CPU/NB efficiency = 67656 ( result time multipled by cpu speed ) <--lower is better

http://members.cox.net/wmdieselmc26/1to1%20ratio.JPG

And here we drop the NB down to 2600....This should be slower in theory..... efficiency = 67436 ( result time multipled by cpu speed ) <--lower is better

http://members.cox.net/wmdieselmc26/1to1%20ratio3.JPG

And here is where the Real sweetspot truly is...... efficiency = 67411 ( result time multipled by cpu speed ) <--lower is better

http://members.cox.net/wmdieselmc26/1to1%20ratio2.JPG

i wonder chew*, does it continue that trend as you raise the CPU frequency? not sure how high you can clock your NB

Quote:

---

Originally Posted by eRazorzEDGE

i wonder chew*, does it continue that trend as you raise the CPU frequency? not sure how high you can clock your NB

---

I would be glad to do it on phase but.....thats not going to help you guys on air /water....Trying to avoid that as although its fun....its not much of a help to those on air/water.

I managed to test it on my 940 at 3000+ at one point, I know with ddr II it took an even bigger performance hit.

From my testings, I found that the higher the NB clock frequency, the more performance up untill your NB is clocked higher than the cpu itself. At that point, the performance takes a real hit.

Thanks for the insights regarding the memory overclocking. It seems that we'd all be better off using low latencies rather than high frequencies, which is pretty much the same I noticed when comparing the AM2 platform to the AM3 platform. Highly overclocked DDR2 memory (620+) combined with tight timings (CL4-4-4) will be able to match any DDR3 configuration with timings equal to CL7-X-X and looser. This reminds me of the Msi P45 DDR3, which overclocked better at DDR2 timings!

Regarding the HT Link findings: are you sure nothing went wrong during the tests? I've done the EXACT same tests in combination with my 4870X2 card and an increase in HT Link frequency showed me a clear increase in 3D performance (3DM01 - Nature). This HT Link frequency has, in addition, given me fantastic performance increases when benching the HD3300 IGP at 1.1GHz.

Quote:

---

Originally Posted by massman

Thanks for the insights regarding the memory overclocking. It seems that we'd all be better off using low latencies rather than high frequencies, which is pretty much the same I noticed when comparing the AM2 platform to the AM3 platform.

---

well, bandwidth still do counts... games do benefit from BW

Sure, but the bandwidth is just the same:

http://www.madshrimps.be/?action=get...58&articID=926

Check the Everest chart and compare GD70 to the M2RS in the 3.6G test setting. 600 CL4 beats 800 CL7 :)

Quote:

---

Originally Posted by massman

Sure, but the bandwidth is just the same:

http://www.madshrimps.be/?action=get...58&articID=926

Check the Everest chart and compare GD70 to the M2RS in the 3.6G test setting. 600 CL4 beats 800 CL7 :)

---

Yes, but because at that point the bottleneck is NB. It's my hypothesis that when NB is a bottleneck (i.e. NB frequency < 2 x memory frequency), higher memory frequencies are somewhat wasted. When you run Everest, keep an eye on memory read bandwidth and L3 read bandwidth. You will see both of them moving together very tightly. It seems to me, NB can only process as much as memory can feed it, and memory can send data only as fast as NB can process.

My hypothesis is in the post #27 and #28 in this thread. In your testing, if NB was clocked @2.8GHz, DDR3-1600/CL7 would be pulling ahead, leaving DDR2-1200 behind.

You can see it from Tony's first post in this thread. In the encoding test, look where the performance stops improving with DDR3-1333, yet keeps improving with DDR3-1600. It's somewhere around 2700~2800MHz NB. So DDR3-1333 is good enough to feed 2600MHz NB, but going higher you'll need a comparable (good timings) DDR3-1600 to keep the linear scaling.

So depending on applications - some apps will need to access system memory more often and others don't - NB frequency and memory frequency both matter, and ideally you'd want to have enough memory bandwidth for your NB even if some of it is wasted. If your NB doesn't clock any higher than, say, 2400MHz, then DDR2-1200 or DDR3-1333 will be all that's needed.

That seems to me as a pretty solid hypothesis (not the numbers, but the concept)!

When I find the time, I'll try to rerun the everest bandwidth benchmarks and check with a higher-clocked NB frequency (if the AM3 boards allow me to).

Quote:

---

Originally Posted by lopri

Yes, but because at that point the bottleneck is NB. It's my hypothesis that when NB is a bottleneck (i.e. NB frequency < 2 x memory frequency), higher memory frequencies are somewhat wasted. When you run Everest, keep an eye on memory read bandwidth and L3 read bandwidth. You will see both of them moving together very tightly. It seems to me, NB can only process as much as memory can feed it, and memory can send data only as fast as NB can process.

My hypothesis is in the post #27 and #28 in this thread. In your testing, if NB was clocked @2.8GHz, DDR3-1600/CL7 would be pulling ahead, leaving DDR2-1200 behind.

You can see it from Tony's first post in this thread. In the encoding test, look where the performance stops improving with DDR3-1333, yet keeps improving with DDR3-1600. It's somewhere around 2700~2800MHz NB. So DDR3-1333 is good enough to feed 2600MHz NB, but going higher you'll need a comparable (good timings) DDR3-1600 to keep the linear scaling.

So depending on applications - some apps will need to access system memory more often and others don't - NB frequency and memory frequency both matter, and ideally you'd want to have enough memory bandwidth for your NB even if some of it is wasted. If your NB doesn't clock any higher than, say, 2400MHz, then DDR2-1200 or DDR3-1333 will be all that's needed.

---

You are bang on the money here ;) as it stands there is absolutely NO NEED to run 1600 ram 7-7-7-24 1t with any AM3 CPU at stock, even if AMD up the NB clock to 2.6GHZ you still would do better with 1333 6-6-6-24 1T.
Also remember i did these tests at 3.8GHZ, if i get time I will rerun them at stock 3.2GHZ to see where that pans out although i have a feeling it would be very close to the same.

Today I will run Elpida Hypers (cas 7 blade) at 6-6-6-24 1t 1600 just to see how this one performs and add a new graph.

Overall my conclusion from this testing to date is AMD need to drop 2000NB clock today and implement 2600mhz on BE. There is no need for them to leave it at 2000 and they would see nice gains in performance when compared to Intel CPU's. Regards where OCZ can help with this I have already asked for a 6-6-6- kit at 1600, I will now also ask for a 5-5-5-kit at 1333 for those not OCing as high but wanting the best at that lower clock.

Keep your eye on the Blade and platinum series for these kits

Mine doesn't like loading windows with the NB set to 2600, have to use the MSI multistep OC tool that applies the setting once the os has loaded :confused: Shows the windows is corrupted screen otherwise :down:

That means your system is not running stable so you have to raise NB-vid a tad if possible.

Mind you that raising CPU-NB volts makes the cpu run hotter though.

Btw, does msi allow you to raise CPU-NB freqency in windows? I thought it wasn't possible without a reboot. :shrug:

yeah, I have to sy, on my board at least, 2600Mhz NB is a no go. Windows 7 64 :banana::banana::banana::banana:s itself on reasonable voltages..

I don't see AMD raising the NB that hig for some time. They have to supply a CPU that will work in any motherboard, at stock voltages, on any OS :) I predict a 3.4Ghz 2.2Ghz NB part won't be too far off if their SOI process tweaks pay off as well as "usual"

Just my opinion.

Quote:

---

Originally Posted by SKoob

Mine doesn't like loading windows with the NB set to 2600, have to use the MSI multistep OC tool that applies the setting once the os has loaded :confused: Shows the windows is corrupted screen otherwise :down:

---

same i get the windows corrupted screen even if i set the clocks back. and i have to add volts if i want my nb to run over 2200 and just adding volts alone but no overclock will corrupt windows. kinda sucks and its a little confusing for me.

Quote:

---

Originally Posted by roofsniper

same i get the windows corrupted screen even if i set the clocks back. and i have to add volts if i want my nb to run over 2200 and just adding volts alone but no overclock will corrupt windows. kinda sucks and its a little confusing for me.

---

Hmm when you add volts for Higher NB it may be conflicting with the HT .


Try 2600 with the higher NBvolts and HT at same speed ,boot and Lower HT if it fails to load. Youd be surprized if it swhat i think it is.

Sometimes that the HT setting in a Bios makes adding Volts to NB unstable not the NBfreqency.

Even with the M3A79-T when i set a certain HT freqency and a certain NB Freqency it will fail loading at windows ( But lowering HT one or two stes its 100% stable at same Clocks). In some way the added volts is affecting my Bios while running on a certain HT divider with High Nb and it only does this when i raise the Nbvolts some or not enuff.

The cure for this while keeping the NB High with volts. I simply lowered the HT one or two settings.

Note: Mine does this because the Bios is bugged on some settings when i use certain Ht dividers with High NB.

Its worth a shot:D

ehhhhh ill try it out but i need to get a new dvd drive first incase it corrupts again. my last dvd drive died 2 weeks ago.

updated first post adding cas5 1333 in the mix...cas5 1333 hits back ;)

Quote:

---

Originally Posted by Tony

updated first post adding cas5 1333 in the mix...cas5 1333 hits back ;)

---

hmmm interesting you gonna try a 1600 c6? may want to think about putting a sticky or something on this. it is useful info.

Quote:

---

Originally Posted by chew*

I would be glad to do it on phase but.....thats not going to help you guys on air /water....Trying to avoid that as although its fun....its not much of a help to those on air/water.

I managed to test it on my 940 at 3000+ at one point, I know with ddr II it took an even bigger performance hit.

---

hey, i'm on phase :p:

i'm gonna do some of my own testing here for s#!ts and giggles as i run 3200MHz NB 24/7 and with the M4A79 DDR2 board.

i really respect you guys doing all these indepth tests and drawing these pretty graphs. i get so impatient with it that i just settle for what seems ok most of the time. u, on the other hand, really push it... love every minute of it. :D

yes sir cas6 next. I may go back and push the CPU to 4000 also for all tests just to get that last piece of data where I think we may see a things change...its all time though so we will see

Quote:

---

Originally Posted by eRazorzEDGE

hey, i'm on phase :p:

i'm gonna do some of my own testing here for s#!ts and giggles as i run 3200MHz NB 24/7 and with the M4A79 DDR2 board.

i really respect you guys doing all these indepth tests and drawing these pretty graphs. i get so impatient with it that i just settle for what seems ok most of the time. u, on the other hand, really push it... love every minute of it. :D

---

The point is this, there are way to many who quote what they think is fast...without actually knowing what is fast.

Check my sig

lol, well seeing what's been posted around here in the xtreme sections, i know i'm no where near "fast"

Was not a personal dig m8...please don't take it as one, but many just set systems up and don't actually test what is fast...

Also consider it may be easier for OCZ to make 1333 5-5-5-24 than 1600 6-6-6-6-24 and we now know cas5 1333 is faster than cas7 1600 so its a no brainer ;)

Also looking at the graph 2400 to 2800NB is where most will land, and 1333 5-5-5-24 is killer here with air cooled CPU's from i would guess 3.6 to 4GHZ...so again its no brainer.

i don't take things to heart, tony, no matter what you say to me... and i didn't think anything wrong of what was being said until you brought it up. i was just going a different direction off your comment, taking a stab at myself ;)

was anyone else able to hit 3000MHz on the NB under air or water? i never could on my M3A79-T, but once i got the M4A79 it as almost easy.

Quote:

---

Originally Posted by roofsniper

yea i think when you overclock HT it also overclocks the HT connecting the cores as well.

---

I think ACC will/should show that.

however I also remembered we can split the hyper transport links going to the cpu in two half speed duplex.

though that seem pointless, unless we can find the inter core speed Ht links somewhere.

(the phenomenons of phenom's )

Quote:

---

Originally Posted by demonkevy666

I think ACC will/should show that.

however I also remembered we can split the hyper transport links going to the cpu in two half speed duplex.

though that seem pointless, unless we can find the inter core speed Ht links somewhere.

(the phenomenons of phenom's )

---

yea whats the option called that lets you split it? im too lazy to go look. and how would it be useful for inter core links?

Quote:

---

Originally Posted by Tony

I will now also ask for a 5-5-5-kit at 1333

---

Quote:

---

Originally Posted by Tony

updated first post adding cas5 1333 in the mix...cas5 1333 hits back ;)

---

thanks tony :up:

Have you already figured out why the performance decreases when you increase the HT Link frequency? I have a completely different dataset, so I'm quite interested in your findings :-)

Quote:

---

Originally Posted by Tony

The point is this, there are way to many who quote what they think is fast...without actually knowing what is fast.

---

I've actually been thinking this as well. I have not really verified on my setup what is the best speeds. My current DDR2 sticks can either hit 570Mhz 5-5-5 or 430 4-4-4. I am curious which is faster, and if I can tweak a few things to try and reach 600Mhz 5-5-5 or 500Mhz 4-4-4 (unlikely).

Quote:

---

Originally Posted by massman

Have you already figured out why the performance decreases when you increase the HT Link frequency? I have a completely different dataset, so I'm quite interested in your findings :-)

---

post it up

Quote:

---

Originally Posted by massman

Have you already figured out why the performance decreases when you increase the HT Link frequency? I have a completely different dataset, so I'm quite interested in your findings :-)

---

i have been told that it just gets the timing off and can cause cache misses and stuff. so with the HT link find the best clock not the fastest.

This is my stock cooled X3 720BE @ X4. Interesting fact that every mem kit i had (Crucial 8500F D9HXT; CSX Diablo DDR3-2000 D9GTR; OCZ Platinum DDR3-1800 8-8-8-x 1.95V) caps out at 855 CAS8 8-8-24 1T, 830 CAS7 7-7-21 1T.

And yes, 2700MHz is the hard limit for NB speed for given settings and poor cooling.

System:
Phenom X3 720BE @ X4 (ACC enabled) - stock cooled
2x512MB Micron D9HXT
ASUS M4A78T-E
PoV 8400GS 256MB PCIe
LCPower 420W

Settings:
Core voltage = 1.35V
CPU/NB voltage = 1.375V
DIMM voltage = 2.1V
HTT voltage = 1.24V
NB voltage = 1.44V
NB 1.8V voltage = 2V
SB voltage = 1.2V

CAS 6 hardly worth the effort? I have re-runned CAS7 with fresh reboot, opened SPi automatically on start-up to make sure no wazza-like effect happened in between.

WinRAR 3.4x benchmark gives only about ~ 20kB/s increase, almost the same as squeezing up the Trc.


CAS 6-7-6-18-34 1T:
http://www.shrani.si/f/3L/12n/XIPJdH...2695-ram81.png

CAS 7-7-6-18-34 1T:
http://www.shrani.si/f/q/36/37xUqXVP...2695-ram81.png

Quote:

---

Originally Posted by Tony

so you guys going to fund all this work? each loop takes over 1 hr at each speed...6 loops = 1 whole day.

---

Just paypal'd you a few quid. Get back to work! :p:

Heh at those speeds try running your HT lower ;) and recompare.

Quote:

---

Originally Posted by Tony

post it up

---

Test settings:
Core frequency: 3000MHz
HT frequency: 200MHz
Northbridge frequency: 1800MHz
Memory frequency: 400MHz (1:2 divider)
Memory timings: 4-4-4-10 2T Ganged

Variable:
HT Link frequency (200 - 1800MHz*)

http://i154.photobucket.com/albums/s...rformance-.png

I'm running a lot lower frequencies though, but the general trend seems to be upwards to me rather than no increase at all let alone decrease. Could you do a run matching the settings I've used and HT Link at 1G and 1.8G?

I have no solid data of the HD 3300 IGP findings, but it seemed that increase the HT Link frequency to 2.4G+ helped me gain quite a bit. Even in the 3DM-V benchmark, I went from 674 to 716.

Phenom II Processors and DDR3 Memory: http://ixbtlabs.com/articles3/mainbo...2-ddr3-p1.html

No benefit at low clocks, at 3.7 GHz noticeable difference with DDR3 1600

Quote:

---

Originally Posted by massman

Test settings:
Core frequency: 3000MHz
HT frequency: 200MHz
Northbridge frequency: 1800MHz
Memory frequency: 400MHz (1:2 divider)
Memory timings: 4-4-4-10 2T Ganged

Variable:
HT Link frequency (200 - 1800MHz*)

http://i154.photobucket.com/albums/s...rformance-.png

I'm running a lot lower frequencies though, but the general trend seems to be upwards to me rather than no increase at all let alone decrease. Could you do a run matching the settings I've used and HT Link at 1G and 1.8G?

I have no solid data of the HD 3300 IGP findings, but it seemed that increase the HT Link frequency to 2.4G+ helped me gain quite a bit. Even in the 3DM-V benchmark, I went from 674 to 716.

---

so your not OCing HT you are starting from an under clock and pushing to stock speed....i was overclocking HT to see if it gained the same way NB does.

Apples to oranges m8...clock HT from 2000 to 2600 and show us what you find

Quote:

---

Originally Posted by salad

Phenom II Processors and DDR3 Memory: http://ixbtlabs.com/articles3/mainbo...2-ddr3-p1.html

No benefit at low clocks, at 3.7 GHz noticeable difference with DDR3 1600

---

they were no where near as thorough as they needed to be.

hence you only reported what you did from their review

Updated to add 1600 6-6-6-24 1t.

1333 cas 5-5-5-24 1t is the winner IMO...faster for PI and right with 1600 6-6-6-24 1t for encoding.

Who needs 1800+ ram 8-8-8- on AM3....? No me ;)

Tony, you think I could run this
at ddr1333 cas 5 with some voltage without killing them?

Tony this is absolutely amazing work!!

I have two requests if possible?!

Could you add voltages for achieved frequencies and latencies in the graphs legend?

Could you add DDR2 1066 CL5 (or CL4 if it's squeezable), and DDR2-800 CL4?

Tony, can you strap your OCZ Cryo-Z on your Phenom, clock it to around 4.3GHz and re-compare? :yepp:

anyone have any idea what the performance of 1333 c5 or c6 would compare to ddr2 1066 c5?

Quote:

---

Originally Posted by roofsniper

anyone have any idea what the performance of 1333 c5 or c6 would compare to ddr2 1066 c5?

---

hopefully Tony will give us the answer ;)

Quote:

---

Originally Posted by Nedjo

hopefully Tony will give us the answer ;)

---

lol tony has done so much already. i think this thread should get a sticky or something because lots of the info is pretty nice.

Quote:

---

Originally Posted by shaolin95

Tony, you think I could run this
at ddr1333 cas 5 with some voltage without killing them?

---

Not got those...will ask what IC and see if I have something close.

Quote:

---

Originally Posted by Nedjo

Tony this is absolutely amazing work!!

I have two requests if possible?!

Could you add voltages for achieved frequencies and latencies in the graphs legend?

Could you add DDR2 1066 CL5 (or CL4 if it's squeezable), and DDR2-800 CL4?

---

I have voltages mapped out BUT the cas5 1333 and cas6 1600 is using Hypers and they make ram clocking very easy for the CPU-NB so it may not be usefull for those with PSC or Micron.

Quote:

---

Originally Posted by tiborrr

Tony, can you strap your OCZ Cryo-Z on your Phenom, clock it to around 4.3GHz and re-compare? :yepp:

---

watercooling only...I can do 4GHZ but i posted where most will land with 955/945 and 925's and thats 3.6>4GHZ...so I picked the middle ;)

getting back to voltages, i can post some work with the Animal Kit i have here as i know thats LV Micron..BUT my cpu maybe better or worse than what you guys have

Quote:

---

Originally Posted by Tony

Not got those...will ask what IC and see if I have something close.

---

Thats awesome Tony!
Thanks for the support...I expect to get a Cryo soon too! :up:

Quote:

---

Originally Posted by roofsniper

yea whats the option called that lets you split it? im too lazy to go look. and how would it be useful for inter core links?

---

it's in the advance hyper transport options

2x lckl mode

( I can't remember my rib got botched by ati 9.4 drivers I can't boot in to windows :/ )

Quote:

---

Originally Posted by Tony

I have voltages mapped out BUT the cas5 1333 and cas6 1600 is using Hypers and they make ram clocking very easy for the CPU-NB so it may not be usefull for those with PSC or Micron.

---

My local supplier have LVs (DDR3-1600; 2x2) on stock but how can I know will they run DDR3-1333 CL5? :shrug:

All this low-latency overclocking made me fire up the GD70 rig again: 700MHZ CL5 1.65v easy

http://www.madshrimps.be/vbulletin/a...1&d=1242765269

http://www.madshrimps.be/vbulletin/f...51/#post238204

I'm really struggling to get my NB past 2.6 and even at 2.6 is seems to need over 1.3V at the NB :eek: I noticed my 3dmark 06 score @ 3.9GHz Core, 2.6 @ NB and HTT @ stock scores had gone down in comparison to 3.9Ghz Core with HTT @ 2.4 :confused:

NB clocks:

2000 0.95V
2200 1.125V
2400 1.2V
2600 1.25V
2800 1.3 to 1.35 depending on memory IC
3000 1.4125V

watercooled

CPU clock on MSI GD70 and Asus M4A79-T

4000 1.4 to 1.425V
3800 1.3 to 1.35V
3600 1.25V
3400 1.2v
3200 1.1v

4000 and 3800 depend on NB clock also and memory used.

Clocking on the DFI 790FX M2R:

CPU clock

3200 1.05V
3400 1.125V
3600 1.2V
3800 1.25V
4000 1.35V

this with the new LV Blades...

The DFI rocks for 955 clocking using multies and HTT to around 250.

The GD70 and Asus will do HTT over 300.

how did u get 4ghz on stock volts?

On the DFI...tweaking I suppose or just an awesome CPU with 10C water cooling it.

I will say this...you guys need to set the CPU's to stock speed and find the lowest voltages needed for stable operation. This took me 2 hrs.

Then leaving CPU speed stock with the lower vcore you found (Important) increase NB clock 1 multi and test again noting the voltage needed. 1 hr more testing.

Then NB up another ...then another....then another...you get where i am going here ;)

I did this all the way to 3GHZ NB.

Now set the ram at say 1333 6-6-6-24 1T, set the Nb at 2000 with the voltage you found stable and find the lowest voltage for the ram thats stable....i bet now if you wanted you could lower the NB voltage 1 notch and still be stable ;) I was able to get 1333 stable with 1.5v on the MSI and then lower the NB voltage a little.

Do this all the way up the NB clock range...Vdimm will change as you clock up. Make notes as this took me 1 day.

So now you have 1333 voltages for the memory at each different Nb clock speed...do the same for 1600 7-7-7-24 1T.

Once you have all that data down, you now need to find what voltage is needed at each CPU half multiplier up you set, keeping in mind you may need to adjust Nb voltage and vdimm again...but if anything its only tiny changes...remember make notes.

Also consider an increase in 790Chipset voltage to around 1.25>1.3V...it helps.
I also found SB voltage can add stability going up 1 notch.

So now you map out Nb clock with the CPu at stock and now CPU clock with the NB at stock...lots of work.

Next map out CPU and NB clocking with ram at 1333...then at 1600 and make notes of all settings.

I did all of this and found the DFI clocked the 955 like a demon, i actually managed 1200 with 4GB ram 1.8V or so with the CPU at 1.25v 3.8GHZ in Vista 32 bit.

Here is the killer punch to all this though, what you need in 32bit OS is totally different to what you need in 64bit OS...64bit needs more juice.
I am testing W7 64 now and 4GHZ is not coming any where near as easy on the DFI as it did with V32...keep this in mind.

so now you know how Tony does his testing ;)

Quote:

---

Originally Posted by SKoob

I'm really struggling to get my NB past 2.6 and even at 2.6 is seems to need over 1.3V at the NB :eek: I noticed my 3dmark 06 score @ 3.9GHz Core, 2.6 @ NB and HTT @ stock scores had gone down in comparison to 3.9Ghz Core with HTT @ 2.4 :confused:

---

Have you tried with HT @ same as NB?

Also sometimes you must ADD some volts to the NB(CHIPSET) sometimes labled CORE PCI-E voltage. Usually when the Speed drops off on the NB or HT@ a given Higher NB freqency its one of two things,, not enuff volts to IMC or NB is getting to Hot @ that freqency.

Tony has displayed the Volt ranges for NB/freqency. If your using More than 1.3v on the CPUnbvolts,, it could be the MOBO just needs more for it to work. Some will use less and others More. In genaral you should be able to come close to same volt range if your temps are not too shabby.

Now while your playing around with volts you can try some more CPUx Multi's . See which one lets the system run faster while using 2600NB with closer to same ram timing. Once you find it Then try raising the HTT 1mhz at a time. Soon youll be going to 2700/2800mhz on the NB. small steps not leaps;)

If your clocking HT @ 2600 with NB @ 2600 using NBmulti . try to go one step lower on HT . Remeber to Give the CHIPSET a hair volts over the base volts when pushing High Clocks and High HTT . This just may get you 2700/2800 when you add HTT to clocking the NB higher..

leave HT at 2000, no need to adjust it.

Gotta try 800 CL5-6-5-18 on my memories when i get home! :up: