TR Forums

Yeah, I'm ritchan on Hardforums BTW

It's stable at 285*14=3990MHz, but every now and then while running Prime95 or LinX it throttles itself to 7.5x multiplier. This also seems to happen at lower speeds.

The CPU with the brand new heatsink was reporting temps of 50C. That's 10 degrees less than the stock heatsink, so that obviously wasn't the problem. I read that overheating VRMs could also be the problem. On the M4A88TD-V, one row of (plastic things) is covered by the heatsink, but another row closer to the CPU isn't. The heatsink started to hurt my finger after 2 seconds, but the ones closer to the CPU hurt my fingers within a second. Lowering voltages...

I settled with good old 270*14. I turned off CPU LLC to ensure the CPU doesn't get anywhere near 1.6V now - mostly to make sure the plastic things around the CPU socket don't heat up as much. I set the voltage to 1.53V in the BIOS, but CPU-Z reports 1.524V idle and when under load, it drops to 1.452V. So it seems I may be able to lower voltage a lot more, but if I do that then the drooped voltage will also decrease with it. Definitely a delicate balancing act.

Obviously 1.53V is overkill for an idle CPU, but if I enabled CnQ it should keep the CPU's power usage down when it's idle, while ramping up the voltage and speed for when it's loaded. So if I turn CnQ back on I should see around 1.3V-1.4V idle, and ~1.4V under load. Makes sense, right?

I'm just starting to OC my 1090T I need a good system benchmark for before and after the overclock, any recommendations?

EDIT : Hey guys, I know this is a silly question but when overclocking what is more beneficial, turning up the multiplier or the "FSB" at stock my mem can reach its max of 1600 mhz. What do you think?

What is more beneficial is having a high NB frequency. It does not matter what your HT link ref is (FSB in Intel terms). So depending on how high your NB multi (for instance mine only goes to 10) goes and what speed your MB accepts RAM at should determine how you want to attack OCing the machine.

i.e, If your MB only accepts memory speeds at 800, 1066 and 1333 you may want to run a HT link ref at 240 and increase the multi from there, and set your RAM speed to run at 1333. (240*10)/3 = 800 * 2 = 1600 RAM speed.


Benchmarks depend on what you want. i.e, Gaming or Scientific
You can use Y-chuncher as a good system benchmark or even LinX. http://www.numberworld.org/y-cruncher/

I second mmmmmdonuts21 thoughts. Get your NB speed to 2400MHz and your DDR3 running at 1600MHz. The Black Edition CPU's are supposed to support that right out of the box. Then when you fiddle core speeds try to pick combination's that do not bump those any higher. There are a couple of good reviews over at Lost Circuits where MS shows that there is not much benefit for NB beyond 2400MHz.
e.g....
http://www.lostcircuits.com/mambo//inde ... mitstart=1

they just reviewed the Pheom II X6 1100T Black Edition here...
http://www.lostcircuits.com/mambo//inde ... &Itemid=42

Also, you might get a better than stock fan. The Coolermaster Hyper 212+ is good about clearing the DDR and it can be mounted vent in any direction. I found that my chipset heatsink had to have a blast of air on it to make my memory stable at those settings. That may just be peculiar to my motherboard. But I am running 4 sticks and that makes it a bit more difficult.

Ok, I've been mixing Overclocking up between my mobo bios and the AMD overdrive utility. Overdrive is awesome but I have this issue. Settings I establish in overdrive aren't retained after a reboot. I'll tweak settings the run prime and some of the built in stress tests, restart and everything is put back to bios defaults. I tried setting stuff in the bios but voltages don't seem to be retained. I'll put CPU voltage at 1.4625 and the overdrive utility shows it as being 1.3 I'm very confused.

I tried to use Overdrive but every now and then after I use the utility to change something my system would BSOD. So I've stuck to OCing via the BIOS. What does CPU-Z/HWMonitor have to say about the voltage?
EDIT: But of course settings from Overdrive aren't retained after a reboot. It's just a Windows program after all, doesn't change your BIOS settings, which is what really matters.

OH! I was really liking the immediacy of the process, not having to boot so frequently, not to mention the far more comprehensive adjustments you could make. I noticed that in the bios you don't have a separate voltage for CPU cores, vs the socket and you also don't have a separate voltage adjuster for turbo core tech. I was really liking overdrive. At least with allot of those GPU windows overclocking devices the overclocks stick better.

I found what the issue was. In overdrive I was adjusting the individual cores. I'm disappointed in the fact that overdrive can't adjust bios settings, but I realize that in the end it has way more inputs than the bios so it probably isn't possible.

One reason I like my motherboard is that I can put the settings that work into the bios. For some reason (maybe because OS is XP64 or maybe the motherboard) AMOD does not seem to work well on my system. I use it mainly for the stress test and the nice graphics display. I asked MS over at Lost Circuits for the black edition memory settings for my memory (he works for OCZ) and put them in by hand.

Just found out today: LinX with 10000 problem size, at 10 runs isn't enough at all! I was wondering why my system wasn't rock solid even though it passed 10 minutes of Prime95, Memtest86+ and LinX - but when I upped the problem size to 15000, the CPU quickly showed its true colours. It really doesn't like being anywhere near 4GHz, I think.

So now I'm at 270x14=3780MHz, down from 275x14=3850MHz. I ran 54 runs of LinX with problem size 17000 on it - now I'll just use it for a week or two without OCing my GTX470 and see if any problems crop up. Then I'll keep the cores at 3.78GHz or less and try to up the bus speed to 285MHz to get the uncore working faster.

Crayon Shin Chan wrote:

Just found out today: LinX with 10000 problem size, at 10 runs isn't enough at all! I was wondering why my system wasn't rock solid even though it passed 10 minutes of Prime95, Memtest86+ and LinX - but when I upped the problem size to 15000, the CPU quickly showed its true colours. It really doesn't like being anywhere near 4GHz, I think.So now I'm at 270x14=3780MHz, down from 275x14=3850MHz. I ran 54 runs of LinX with problem size 17000 on it - now I'll just use it for a week or two without OCing my GTX470 and see if any problems crop up. Then I'll keep the cores at 3.78GHz or less and try to up the bus speed to 285MHz to get the uncore working faster.

For me the ultimate stability benchmark is DC projects, especially Folding at Home. I have had many OC's that have passed LinX at sets around 20-25k (15-25 times) and prime 95 for 8+ hours or so and then fail at F@H within an hour. I can't remember a time when F@H didn't fail but it failed LinX or Prime 95. Here is a link to a F@H Benchmarking app that does not require you to register for the project but it can give you a great idea with stability. http://www.avid-edge.com/fah/FAHbench_v12.zip

Errr... actually... I ran F@H all the time in the background, along with F@H GPU3. And it didn't throw any errors. It was only when I fired up LinX that I would find any errors, or some weird glitch would happen like my DVD drive suddenly disappearing. I also got two SESSION_HAS_VALID_POOL_ON_EXIT BSODs, which clued me onto the CPU (I also OCed my GPU). They were misleading though since they only occurred while playing Crysis. I didn't give much time to Prime95 though.

Wow thanks donuts! I can actually see how fast my system will run a bigadv! Also if my CPU screws something up it won't count against me when it gets uploaded!

Your GPU and or drivers maybe an issue and not your CPU. Do any errors happen when your GPU is not overclocked? I found this with that check bug error: http://msdn.microsoft.com/en-us/library/ff560119(VS.85).aspx and here http://social.technet.microsoft.com/Forums/en-US/w7itproperf/thread/5185cb9d-2b20-45d4-a054-7f49e971623a.

Both of these sites seem to suggest a problem with the GPU / Drivers (which I am guessing because its overclocked that maybe the case) rather than the CPU. I don't know if your running both at full bore when trying to isolate the problem but keeping your GPU at stock would be a great place to start.

My 1090T system has been showing stability issues at stock speed with Turbo core turned on(random crashes or lockups). I get a more stable cpu overclocked to 4.1 Ghz with it off than a stock clock with it on. What do you all think is the issue here? Would you say my CPU or mobo is messed? Since I've already fiddled with OCing it I can't return the new toys but I was wondering your thoughts on my issue?

Turbo is pushing your CPU beyond 4.1GHz, and your CPU can't go faster than 4.1GHz with the amount of cooling/voltage you're throwing at it. Nobody overclocks a CPU with Turbo on anyway - why artificially limit the number of cores that can run at a specified frequency if you can run all of them at a number close to or equal to that specified frequency?

I see... thanks, I didn't know about those resources. Yes I was running it at 810MHz with 912.5MHz RAM. It's funny because I folded with the GPU3 client at those speeds, and I didn't get a NaN, although I did get a EUE once. That's over about... oh, 2-3 days? So that's about.... 30 WUs at least, with a TPF of 1min 14s. It didn't crash at all. Then I ran Crysis, and after some time it will crash with that BSOD. I could hear the audio still going on in the background... oh yes, that should mean it was the GPU after all, sorry. I would feel reluctant to press reset since the audio was still going on and I thought I could save it, but then I would reset it.

Since it occurs after some time (like 20 minutes), I'm holding out hope that this is a cooling related issue, which will be solved when my Gelid Icy Vision arrives.

I haven't had time lately to play more Crysis. Plus the alien levels kinda diminished my interest.

Crayon Shin Chan wrote:

Turbo is pushing your CPU beyond 4.1GHz, and your CPU can't go faster than 4.1GHz with the amount of cooling/voltage you're throwing at it. Nobody overclocks a CPU with Turbo on anyway - why artificially limit the number of cores that can run at a specified frequency if you can run all of them at a number close to or equal to that specified frequency?

I'm sayin that at stock settings 3.2 Ghz the comp freezes and stuff with turbo clock enabled(no OC), I leave it set to auto and the comp crashes in a few hours to a days time. I can easily OC to 4.0 ghz across all cores,or higher with turbo core disabled. I was just wondering if turbo core crashing was a sign of an error in my CPU or mobo.

I'm watching this thread with interest. I've been tempted to upgrade to a Phenom II X6 1090T.
Check out
pages 7-8 of the Lost Circuits Sandy Bridges review
versus page 5 of the Tech Report Sandy Bridges review.
Look at the two memory bandwidth graphs from the two reviews. Take the Phenom X6 1100T. That 15.6 versus 13.4 is the difference caused by using black edition memory timings DDR3 at 1600MHz and NB at 2400MHz in the Lost Circuits review. I get the same boost with my X4 955 BE.

Even running no core overclocking, it gives a nice boost to most things. Once you get that working then mess around with core frequencies. The AOD benchmark will fail every time if you don't get the timings and voltages right when pushing the memory. But for my X4 955 BE, once you have the memory working and code that into the bios, you can fiddle the cores without messing up the memory. I fear long term strain so I just run the memory timings and don't bother overclocking the cores.

I installed Sandra 2011 Lite and it gave me only 9.4GB/s for memory bandwidth! That made me feel sad... and I OCed my uncore by 700MHz too! Whatever... it still feels incredibly fast. Those tighter timings in the TR review probably helped them get those numbers.

kamikaziechameleon: I really wouldn't know, I'd say maybe trying to reset all settings in the BIOS might help. But since 4GHz is stable, I wouldn't bother personally. I wouldn't suspect the CPU.

Mr Bill wrote:

I'm watching this thread with interest. I've been tempted to upgrade to a Phenom II X6 1090T.
Check out
pages 7-8 of the Lost Circuits Sandy Bridges review
versus page 5 of the Tech Report Sandy Bridges review.
Look at the two memory bandwidth graphs from the two reviews. Take the Phenom X6 1100T. That 15.6 versus 13.4 is the difference caused by using black edition memory timings DDR3 at 1600MHz and NB at 2400MHz in the Lost Circuits review. I get the same boost with my X4 955 BE.

Even running no core overclocking, it gives a nice boost to most things. Once you get that working then mess around with core frequencies. The AOD benchmark will fail every time if you don't get the timings and voltages right when pushing the memory. But for my X4 955 BE, once you have the memory working and code that into the bios, you can fiddle the cores without messing up the memory. I fear long term strain so I just run the memory timings and don't bother overclocking the cores.

I will try and run some benchmarks when I get home for you on this on my 1055T. I got 4x4gb of RAM so it will be interesting to see how that would compare to a 4x2gb. It should be quite a bit faster.

Just out of curiosity do you use things that will be able to take advantage of all six cores? Don't get me wrong, I think the Thubans are fantastic chips but for a small gain in memory bandwidth, I would not be able to justify upgrading unless I was rendering/encoding/DC projects, etc. and using all the of the cores the majority of the time (and the additional memory bandwidth). The only reason I say this is if your upgrading for gaming there are like 20 or so games that take advantage of six cores http://www.pcgameshardware.com/aid,794274/From-Medal-of-Honor-to-Civ-5-17-Games-that-already-benefit-from-six-cores-CPUs/Practice/ (granted more will be coming out in time) but I guess I am not to sure about that upgrade unless two additional cores really really benefit you as much as updating your GPU for around the same amount of money as a 1090T (or way less) and seeing much better FPS results.

My next question is what about waiting for Bulldozer in a few months? You already have everything except the MB and CPU you would need. I am just curious on your perspective I guess.

My main application is Adobe Lightroom 64-bit. I take a lot of digital photos and spend hours at a time examining the results. Lightroom actually goes quicker with more cores and sure the X4 955 BE renders in a few seconds but cutting that even more would be nice. MS over at Lost Circuits benchmarked Lightroom in his X6 1090T BE review and it is noticeably faster. The X6 instead of bulldozer because I can upgrade one last time for a ~$200 without changing anything else in my system. Finally, I might be able to sell the X4 BE 955 for ~$100.

Edit: I should add that Lightroom is very IO bound and that really cramped it as a benchmark. I have a SSD dedicated to photos so I can fiddle with setting quickly. My OS is on a separate SSD so I thought I would be able to get quicker results. As it turns out, Lightroom lets you set a cache directory but it insists on having renders saved in a directory on the OS drive. I try to install applications on a separate partition but I still have limited space on that partition (Due to numerous OS updates, especially NET!), I am constantly getting notifications of space limitation and have to go in and delete those cache files. I only have about 10GB free on the OS drive IIRC.

Oh okay. I was just curious what application you were planning on using it for. Got to make sure those extra cores are being utilized

Mr Bill, I ran the Sandra benchmark for memory on my system you listed above and got 13.67 GB/s. I forgot what my latency bench was. I can probably look when I get home. Any ways my memory is G. Skill RipJaws 1600 (4x4gb) @ 9-9-9-24-2T and NB@2400 HT@2100. I don't know if that sways you in anyway but I figure I post it anyways.

mmmmmdonuts21 wrote:

Mr Bill, I ran the Sandra benchmark for memory on my system you listed above and got 13.67 GB/s. I forgot what my latency bench was. I can probably look when I get home. Any ways my memory is G. Skill RipJaws 1600 (4x4gb) @ 9-9-9-24-2T and NB@2400 HT@2100. I don't know if that sways you in anyway but I figure I post it anyways.

This system? So what version of Sandra did you use?
MB: Gigabyte 870a-ud3
RAM: 4x4GB GSkill RipJaws 1600

CPU: 275*14 (3850) @ 1.5V
NB: 2750 @1.35V
RAM: 1466 (9-9-9-24-2T)
HT: 2000

I have to use an XP64 version...
Here are my SiSoftware 2009.SP3c results:

Benchmark Results
Aggregate Memory Performance : 15.8GB/s
Int Buff'd iSSE2 Memory Bandwidth : 15.8GB/s
Float Buff'd iSSE2 Memory Bandwidth : 15.8GB/s

Benchmark Results
Aggregate Arithmetic Performance : 41.7GOPS
Dhrystone ALU : 46GIPS
Whetstone iSSE3 : 37.42GFLOPS

Benchmark Results
Aggregate Multi-Media Performance : 119.29MPixel/s
Multi-Media Int x16 aSSE2 : 147.77MPixel/s
Multi-Media Float x8 iSSE2 : 90.8MPixel/s
Multi-Media Double x4 iSSE2 : 49.6MPixel/s

Benchmark Results
Memory Latency : 73ns
Detailed Benchmark Results
1kB Range : 0.9ns / 3.0clocks
4kB Range : 0.9ns / 3.0clocks
16kB Range : 0.9ns / 3.0clocks
64kB Range : 0.9ns / 3.0clocks
256kB Range : 4.9ns / 15.8clocks
1MB Range : 15.5ns / 49.7clocks
4MB Range : 21.4ns / 68.9clocks
16MB Range : 67.6ns / 217.3clocks
64MB Range : 72.6ns / 233.4clocks

Benchmark Results
Cache/Memory Bandwidth : 55.2GB/s
Detailed Benchmark Results
2kB Blocks : 167.11GB/s
4kB Blocks : 210.38GB/s
8kB Blocks : 215.14GB/s
16kB Blocks : 218GB/s
32kB Blocks : 232.2GB/s
64kB Blocks : 202.23GB/s
128kB Blocks : 185.85GB/s
256kB Blocks : 166.4GB/s
512kB Blocks : 108.7GB/s
1MB Blocks : 92.84GB/s
4MB Blocks : 43.72GB/s
16MB Blocks : 11.52GB/s
64MB Blocks : 11GB/s
256MB Blocks : 11GB/s
1GB Blocks : 11GB/s
4GB Blocks : 11GB/s

-------------------------
CPU-Z version 1.51
-------------------------
Processor 1 (ID = 0)
Number of cores 4 (max 4)
Number of threads 4 (max 4)
Name AMD Phenom II X4 955 Black Edition
Codename Deneb
Specification AMD Phenom(tm) II X4 955 Processor
Package Socket AM2+ (940)
CPUID F.4.2
Extended CPUID 10.4
Brand ID 13
Core Stepping RB-C2
Technology 45 nm
Core Speed 3215.4 MHz (16.0 x 201.0 MHz)
HT Link speed 2009.6 MHz
Instructions sets MMX (+), 3DNow! (+), SSE, SSE2, SSE3, SSE4A, x86-64
L1 Data cache 4 x 64 KBytes, 2-way set associative, 64-byte line size
L1 Instruction cache 4 x 64 KBytes, 2-way set associative, 64-byte line size
L2 cache 4 x 512 KBytes, 16-way set associative, 64-byte line size
L3 cache 6 MBytes, 48-way set associative, 64-byte line size
FID/VID Control yes
min FID 5.0x
Features XD, VT
Package Type 0x1
Model 55
String 1 0x3
String 2 0x6
Page 0x0
Attached device PCI device at bus 0, device 24, function 0
Attached device PCI device at bus 0, device 24, function 1
Attached device PCI device at bus 0, device 24, function 2
Attached device PCI device at bus 0, device 24, function 3
Attached device PCI device at bus 0, device 24, function 4
-------------------------
CPU-Z version 1.51
-------------------------
Chipset
------------------------------------------------------------------------------

Northbridge AMD 790FX rev. 00
Southbridge ATI SB750 rev. 00
Graphic Interface PCI-Express
PCI-E Link Width x16
PCI-E Max Link Width x16
Memory Type DDR3
Memory Size 8192 MBytes
Channels Dual (Unganged)
Memory Frequency 803.9 MHz (1:4)
CAS# 8.0
RAS# to CAS# 8
RAS# Precharge 8
Cycle Time (tRAS) 24
Bank Cycle Time (tRC) 32
Command Rate 2T
NB Frequency 2411.6 MHz

Sorry about my incomplete results. I reran all the benchmarks you have below Here is a more complete version like you have listed.

Here are my SiSoftware Lite 2011c results (Windows 7x64):

Benchmark Results
Aggregate Memory Performance : 12.81GB/s
Int Buff'd iSSE2 Memory Bandwidth : 12.8GB/s
Float Buff'd iSSE2 Memory Bandwidth : 12.82GB/s

Benchmark Results
Aggregate Arithmetic Performance : 71.55GOPS
Dhrystone ALU : 79GIPS
Whetstone iSSE3 : 64.88GFLOPS

Benchmark Results
Aggregate Multi-Media Performance : 124.77MPixel/s
Multi-Media Int x16 aSSE2 : 97MPixel/s
Multi-Media Float x8 iSSE2 : 160.37MPixel/s
Multi-Media Double x4 iSSE2 : 87.28MPixel/s

Benchmark Results
Memory Latency : 77ns
Detailed Benchmark Results
1kB Range : 0.8ns / 3.0clocks
4kB Range : 0.8ns / 3.0clocks
16kB Range : 0.8ns / 3.0clocks
64kB Range : 0.8ns / 3.0clocks
256kB Range : 4.2ns / 15.7clocks
1MB Range : 14.8ns / 55.4clocks
4MB Range : 19.8ns / 74.4clocks
16MB Range : 71.1ns / 266.5clocks
64MB Range : 77.0ns / 288.6clocks

Benchmark Results
Cache/Memory Bandwidth : 93.11GB/s
Detailed Benchmark Results - It gives the peak number and a range on mine so I put both
2kB Blocks : 459GB/s (203.88GB/s - 459GB/s)
4kB Blocks : 463.44GB/s (187.66GB/s - 463.44GB/s)
8kB Blocks : 463GB/s (162.13GB/s- 463GB/s)
16kB Blocks : 461.42GB/s (183.36GB/s - 461.42GB/s)
32kB Blocks : 464.5GB/s (175GB/s - 464.5GB/s)
64kB Blocks : 413.16GB/s (186.89GB/s - 413.16GB/s)
128kB Blocks : 388GB/s (197.19GB/s - 388GB/s)
256kB Blocks : 368.7GB/s (170GB/s - 368.7GB/s)
512kB Blocks : 318.2GB/s (152.86GB/s - 318.2GB/s)
1MB Blocks : 182.55GB/s (97.88GB/s - 182.55GB/s)
4MB Blocks : 89.5GB/s (59.33GB/s - 89.5GB/s)
16MB Blocks : 34.46GB/s (27.53GB/s - 34.46GB/s)
64MB Blocks : 12.1GB/s (11.7GB/s - 12.1GB/s)
256MB Blocks : 12GB/s (11.54GB/s - 12GB/s)
1GB Blocks : 11.37GB/s (11GB/s - 11.37GB/s)
4GB Blocks : 9GB/s (8.8GB/s - 9GB/s)

-------------------------
CPU-Z version 1.56
-------------------------
Processor 1 (ID = 0)
Number of cores 6 (max 6)
Number of threads 6 (max 6)
Name AMD Phenom II X4 955 Black Edition
Codename Thuban
Specification AMD Phenom(tm) II X6 1055T Processor
Package Socket AM3 (938)
CPUID F.A
Extended CPUID 10.A
Revision PH-E0
Core Stepping 0
Technology 45 nm
Core Speed 3750.4 MHz (12.5 x 300.0 MHz)
HT Link speed 2100.0 MHz
Instructions sets MMX (+), 3DNow! (+), SSE, SSE2, SSE3, SSE4A, x86-64, AMD-V
L1 Data cache 4 x 64 KBytes, 2-way set associative, 64-byte line size
L1 Instruction cache 4 x 64 KBytes, 2-way set associative, 64-byte line size
L2 cache 4 x 512 KBytes, 16-way set associative, 64-byte line size
L3 cache 6 MBytes, 48-way set associative, 64-byte line size
-------------------------
CPU-Z version 1.56
-------------------------
Chipset
------------------------------------------------------------------------------

Northbridge AMD 870 rev. 00
Southbridge ATI SB850 rev. 00
Graphic Interface PCI-Express
PCI-E Link Width x16
PCI-E Max Link Width x16
Memory Type DDR3
Memory Size 16384 MBytes
Channels Dual (Unganged)
Memory Frequency 800.0 MHz (3:8)
CAS# 9.0
RAS# to CAS# 9
RAS# Precharge 9
Cycle Time (tRAS) 24
Bank Cycle Time (tRC) 33
Command Rate 2T
NB Frequency 2400.1 MHz

Mr Bill wrote:

mmmmmdonuts21 wrote:

Mr Bill, I ran the Sandra benchmark for memory on my system you listed above and got 13.67 GB/s. I forgot what my latency bench was. I can probably look when I get home. Any ways my memory is G. Skill RipJaws 1600 (4x4gb) @ 9-9-9-24-2T and NB@2400 HT@2100. I don't know if that sways you in anyway but I figure I post it anyways.

This system? So what version of Sandra did you use?
MB: Gigabyte 870a-ud3
RAM: 4x4GB GSkill RipJaws 1600

CPU: 275*14 (3850) @ 1.5V
NB: 2750 @1.35V
RAM: 1466 (9-9-9-24-2T)
HT: 2000

I could never fully get this system to run more than a day without rebooting while running F@H so I went back to the settings I have listed above in the previous post. Additionally, I have the Masters Version of CS3, so if there are some tests you would like me to run with Lightroom I can give it a try so you can compare systems. (I would boot @3.2 so you could see what the additional cores would do for you)

Additionally in the previous post my complete RAM timings were 9-9-9-24-4-33-9-5 as viewed by Sandra.

I am rerunning benchmarks with San2011-1730 and will post for comparison. Its good to see that you can run 4x4GB of DDR3 and still keep the 800MHz DDR3 and 2400MHz NB. I had wondered if an X6 could do it.