DFI UT P35-T2R: Tweakers Rejoice!
by Rajinder Gill on October 18, 2007 2:00 PM EST- Posted in
- Motherboards
Quad-Core, Part 3
Finding stable VTT/GTLs @ 3.6GHz with G0 stepping quad-core CPUs
Assessing a motherboard's quad-core overclocking capabilities is best done at settings which are not at the extreme limits of attainable performance. The primary reason for this is because GTL settings become ever more sensitive to one digit (0.01v) changes as we near the maximum chipset capabilities. Our experience shows that 440-450FSB on this board is a good reference/starting point, as all chipsets should be capable of attaining this level of FSB with current G0 stepping CPUs, using a "relatively minor" (minor in terms of GTL tuning) amount of tweaking. Pushing high FSBs too soon is likely to bring mixed results, without attaining sufficient user understanding to make improvements when failures are experienced.
Although there is no adherent logic that can be applied to making GTL adjustments (without a scope), having a scale does give us the chance of some success. A safe point for starting out is around 400-440FSB using around 1.4V-1.5V VTT and appropriate GTLs. This will allow the user to get a feel for VTT/GTL change requirements as the FSB speed is scaled upwards. Below is the approach we used to get our foot in the door;
Recommended Overclocking Base Settings and Components
After entering target timings and voltages in the BIOS, we begin with some Prime95 testing in order to determine the effects of applied VTT/GTL adjustments. At this point we should already have determined sensible timings for RAM, which should not induce failures while we are checking for FSB stability. Sticking below default memory timing specifications while using specification voltage is a very wise choice at this point (use enough voltage for the RAM to run its specified speeds). We probed actual VDimm levels with a DMM to ensure that BIOS set values are not adrift from user set BIOS values. We could not find any significant level of voltage droop/undervolting/overvolting under idle or load conditions.
Our recommendations in terms of RAM are that you choose a good kit that can run at least DDR2-1066 @ CAS5. (2GB or 4GB of OCZ CAS5 PC9200 RAM has proven very successful with the 913 BIOS.) This will push towards the best performance possible from this board. At this time the P35 Chipset seems to favor Micron D9GMH based modules over the highly regarded D9GKX bin. A reason for the preference of the lower bin is unknown at this time. Our D9GKX modules required higher levels of voltage than we were accustomed to providing them, and we were not able to match the D9GMH based module results in terms of speed, stability, or timings ranges. The advantage to the user is of course good performance for less money, as D9GKX based modules are typically more expensive than their lower binned counterparts.
Lower speed-binned RAM is usable by adjusting strap/divider/FSB combinations in BIOS, to stay within RAM capabilities at the expense of overall performance. Tests between FSB and chipset strap ratios show the performance penalty in real world applications and games benchmarks is very small.
Of paramount importance is a level of patience and perseverance. GTL tuning takes time but can be very rewarding in terms of performance advantages. Again, perseverance is the key! Now, this is what we did to help determine which settings to change.
Remember up to the 913 BIOS the actual GTL core functions are mislabeled. The actual core each function relates to is shown below:
CPU GTL 1/3 REF Volt = GTL ref 0/1 (Cores 0 and 1)
CPU GTL 0/2 REF Volt = GTL ref 2/3 (Cores 2 and 3)
Finding stable VTT/GTLs @ 3.6GHz with G0 stepping quad-core CPUs
Assessing a motherboard's quad-core overclocking capabilities is best done at settings which are not at the extreme limits of attainable performance. The primary reason for this is because GTL settings become ever more sensitive to one digit (0.01v) changes as we near the maximum chipset capabilities. Our experience shows that 440-450FSB on this board is a good reference/starting point, as all chipsets should be capable of attaining this level of FSB with current G0 stepping CPUs, using a "relatively minor" (minor in terms of GTL tuning) amount of tweaking. Pushing high FSBs too soon is likely to bring mixed results, without attaining sufficient user understanding to make improvements when failures are experienced.
Although there is no adherent logic that can be applied to making GTL adjustments (without a scope), having a scale does give us the chance of some success. A safe point for starting out is around 400-440FSB using around 1.4V-1.5V VTT and appropriate GTLs. This will allow the user to get a feel for VTT/GTL change requirements as the FSB speed is scaled upwards. Below is the approach we used to get our foot in the door;
Recommended Overclocking Base Settings and Components
- Intel Quad-core CPU - G0 stepping
- A sturdy and proven PSU: we used the OCZ ProXStream 1000W for the majority of our testing
- High quality DDR2: you generally want RAM that is capable of CAS4 operation at DDR2-1100 with speeds up to DDR2-1280+ at CAS5
- High performance GPU: one that is capable of withstanding high PCI Express speeds such as the NVIDIA 8800 GTX Ultra
- Good Cooling: Thermalright Ultra-120 (with high CFM fans in push/pull) or very good water-cooling (i.e. Swiftech Apogee GTX/D-Tek Fuzion block)
- High Speed 80-120mm fans: used to cool NB/PWM and RAM
- Microsoft XP SP2 for 2GB or Vista 64-bit for 4GB Memory configurations
- Target CPU VCore is 1.37V-1.43V for 450FSB @ 3.6GHz. (Good cooling is required!)
- Target VNB is 1.55V-1.65V
After entering target timings and voltages in the BIOS, we begin with some Prime95 testing in order to determine the effects of applied VTT/GTL adjustments. At this point we should already have determined sensible timings for RAM, which should not induce failures while we are checking for FSB stability. Sticking below default memory timing specifications while using specification voltage is a very wise choice at this point (use enough voltage for the RAM to run its specified speeds). We probed actual VDimm levels with a DMM to ensure that BIOS set values are not adrift from user set BIOS values. We could not find any significant level of voltage droop/undervolting/overvolting under idle or load conditions.
Our recommendations in terms of RAM are that you choose a good kit that can run at least DDR2-1066 @ CAS5. (2GB or 4GB of OCZ CAS5 PC9200 RAM has proven very successful with the 913 BIOS.) This will push towards the best performance possible from this board. At this time the P35 Chipset seems to favor Micron D9GMH based modules over the highly regarded D9GKX bin. A reason for the preference of the lower bin is unknown at this time. Our D9GKX modules required higher levels of voltage than we were accustomed to providing them, and we were not able to match the D9GMH based module results in terms of speed, stability, or timings ranges. The advantage to the user is of course good performance for less money, as D9GKX based modules are typically more expensive than their lower binned counterparts.
Lower speed-binned RAM is usable by adjusting strap/divider/FSB combinations in BIOS, to stay within RAM capabilities at the expense of overall performance. Tests between FSB and chipset strap ratios show the performance penalty in real world applications and games benchmarks is very small.
Of paramount importance is a level of patience and perseverance. GTL tuning takes time but can be very rewarding in terms of performance advantages. Again, perseverance is the key! Now, this is what we did to help determine which settings to change.
- Run Prime95 in XP or Vista and note which cores fail. The reason the Prime95 Torture Test is recommended for this situation is that we can study which cores are failing the test consistently and then attempt to set the corresponding CPU core GTL values to improve failure time frames or eliminate failures altogether.
- Repeat the Prime95 Torture Test several times using the same settings to pick up patterns of time and core related failures (if you are experiencing them). At this point, as stated earlier, it is important not to be overzealous with tight RAM timing settings.
- If there is no pattern (i.e.; all cores fail at random), then consider tuning NBGTL by 1-2 digits in either direction, then monitor the effects again to look for changes from previous values. VCore, VNB and VDimm are all likely candidates if not within range.
- If GTL CORE 0/1 and GTL 2/3 values are close enough to optimum for your CPU, it is likely that tuning NBGTL will create a more predictable and consistent failure on one or two of the four cores. For most CPUs you will not have to stray too far from the NBGTL scale value to find this point, though we have heard of reports where some CPU/motherboards prefer the NBGTL at around a 66% mark of VTT.
- 1-2% variance between GTL setting adjustments sounds remarkably close between systems in theory
- Move the corresponding GTL scale for the failing core by adding or subtracting one digit from the recommended values in the VTT/GTL table, do not stray over 1-2 digits at a time, it is imperative to check for Prime95 failure times after EACH change in GTL value.
Remember up to the 913 BIOS the actual GTL core functions are mislabeled. The actual core each function relates to is shown below:
CPU GTL 1/3 REF Volt = GTL ref 0/1 (Cores 0 and 1)
CPU GTL 0/2 REF Volt = GTL ref 2/3 (Cores 2 and 3)
Facebook
Twitter
RSS
30 Comments
View All Comments
retrospooty - Sunday, October 21, 2007 - link
Not really. Check this memory test here at AT. http://www.anandtech.com/memory/showdoc.aspx?i=312...">http://www.anandtech.com/memory/showdoc.aspx?i=312...At DDR 2000 at 6X500 CPU the best sisoft standard memory score for DDR3 is 9138 at DDR3 2000 8-7-6-18 ... I get 8871 with my DDR2 1000mhz at 4-4-4-10 with a similar setup on the DFI P35 mobo. DDR3 even at a highly overclocked 2000mhz is hardly faster than DDR2 at 1000 now. I am not going to go to DDR3 until Nehalem is cheap, and DDR3 latency is lower and cheap, 2 more years out.
And PCI-E 2.0 is not going to help for a long time. Like with PCI to AGP, and AGP to PCI-E there is no improvement at all right off the bat. The graphic bus standard is always WAY ahead of the actual cards need for it, 1-2 years at least. We are not even utilizing PCI-E 8x, much less 16x, not even with dual 8800GTX's. Also , PCI-E 2.0 is backward compatible, so I I get the latest greatest Graphic card next year (whatever it is) it will be PCI-E 2.0, but will work fine on PCI-E 1 motherboards.
Raja Gill - Thursday, October 18, 2007 - link
It's one of those things in life, despite having no real logic to it, 'tweaking' actually has a market. DFI took a risk, while this segment does not guarantee large sale volumes, some of the return is made up in a higher profit margin. Without DFI around, some of the more pioneering BIOS options (that actually are of use to Joe blow due to Intel's binning strategy), would never filter thru to other boards. The asking price is high, both Gary and myself recognised this. It has to be said that those who have paid the premium have generally not been dissapointed. In terms of the cascade results, yes, boardss like this and Asus's ROG line are used by the benchmarking community to chase big records, again this is a growing community who pays high dollar for every last drop of performance. We are aiming to satisfy a slightly wider audience with the performance board reviews, by looking at how they will be used, and help out with settings by spending long hours exploring the BIOS functions ourselves, still trying to give a balanced view for JB and some insight to help the benchmarkers decide if the board fits their needs... The bios's on so many of these boards are getting more complex, we hope those who buy the boards can use some of the settings we provide to get a base to work from, (if they need it)..If there's anything else you would like to see added, please let us know....
regards
Raja
Jodiuh - Friday, October 19, 2007 - link
SCREW THEM ALL! I've been drinking and still managed this in under 2 hours.http://img.photobucket.com/albums/v212/jodiuh/DFI%...">http://img.photobucket.com/albums/v212/...0UT%20P3...
Sure I've been able to achieve good results w/ lowly air on the Giga DS3, Asus P5B, DFI Dark, and Abit IP35, but this board:
1. makes oc fun
2. great layout (usb ports towards the front where they should be, shame Abit)
3. great stock cooling, especially for air
4. excellent bios
5. did I mention speed drunk oc? (beer, then 50mhz on the bus)
It's truly a great game. And for about $260, it's a steal. There's no other board that offers this much pleasure. Sure it's $, but then didn't most of you pay $320 for your 6600's? :D
Avalon - Friday, October 19, 2007 - link
Any drunk monkey could OC a decent C2D to 3.5Ghz. I've had plenty of fun OC'ing on my $90 Gigabyte P35 DS3L, which also has a great layout and decent BIOS. It's no DFI, but it gets the job done for $160 less, which is enough money for me to buy a decent video card, 4GB of RAM, 750GB of hard drive space, etc...Don't get me wrong, DFI is by far my favorite motherboard company, and their BIOS tweaking is the best available. However, I will NEVER consider a $260 board a "steal" when the only thing I'll get out of it over a significantly cheaper board is that I have more settings to play with if I feel bored. Fortunately, when I get bored, I load up a game, instead.
cmdrdredd - Saturday, October 20, 2007 - link
Avalon hit it right on the head. Any decent C2D can do over 3.2Ghz and some up close to 4Ghz with proper cooling on $90 boards. There is absolutely NO reason to buy a $260 board this late to market. Not to mention that it is also slower than boards which retail for less money in this review at the settings used.The number of people actually willing to and wanting to use 550Mhz+ FSB is too small to matter.
/end
Jodiuh - Tuesday, October 23, 2007 - link
Most of the 965 board's are a crap shoot for 500fsb, but the 2 p35's I've used had no qualms about this. It's not just getting the OC for me tho. I needed a board that:-doesn't look like rainbow brite exploded on it (Gigabyte DS3)
-doesn't use an AMI bios, hits 500+ FSB easily (Asus P5B Deluxe)
-doesn't crackle (DFI P965-S, Abit IP35 Pro)
-isn't NV (EVGA 680i)
-has usb ports, front panel header, and SATA ports in the far right of the board
-and of course the great stock cooling (uses nuts/bolts), excellent bios, and stable voltages add to the package
There's just no other non-Asus board that did all that right now. Paying another $70 for the extras isn't an issue for a hobbyist that plays w/ his rig.
Acanthus - Friday, October 19, 2007 - link
Yes, we did.And on quad cores, this board does no better than other boards much cheaper than itself.
I am not here to bash DFI, they make great stuff.
The price of admission is too high for this board for quad overclocking from the results i've seen. The FSB limitation appears to be the G0 quads themselves, they all seem to top out at 470-510fsb anyway. The Asus P5K and Gigabyte P35 DQ6 both easily do those speeds on unmodified boards.
customcoms - Thursday, October 18, 2007 - link
Thanks anandtech for providing an overclocking article TRULY worth reading! We needed a motherboard review like this when DFI's original nForce4 boards came out-trying to figure out every bios option, while fun and rewarding, is a very very time consuming process.Its one thing for a reviewer to claim "this board is capable of 500+ fsb speeds" and another COMPLETELY for that reviewer to document virtually every step and every bios option they used to get to that speed. This review is along the lines of the user reviews I normally turn to to verify the components I am interested work as advertised. While the cascade cooling results and extreme price of this board could have been left out, it is good to see a reviewer actually using the hardware in the environment a $300 (or $260, or any board over $200 imo) motherboard is going to be used in.
Jodiuh - Thursday, October 18, 2007 - link
It can be found for $260 from motherboardpro. Comes w/ a nice little LAN carrying harness too.cmdrdredd - Thursday, October 18, 2007 - link
Still too much money. As was said before, it still does not beat some of the boards which have been out for a long time. Sure it looks nice, but for the money spent here, you can buy more memory or the next step up in CPU if you go with a less expensive board.It's late, expensive, and does not have any magic that makes me say "it's better than anything else out there". Even the $90 IP35-E is a good match against this DFI board unless you have to or like to spend hours setting up the bios just so.
It's nice, but like I said too little too late.