DFI LT X48-T2R Initial Thoughts - Page 5 - Overclock.net

Axis01 · 04-16-08

The TMPIN1 is the system temp which is located in the PWM. Not sure what the ACPI temp is reading, but it follows the CPU temp in smartguardian pretty closely.

Axis

MasterKromm · 04-21-08

Clunk updated his review thread with some (much needed) E8500 OCing...

Quote:

Wolfdale E8500 Testing & Overclocks.

Firstly, a few quick notes.

I was planning on testing with a Q9450, but as most people are aware, they are in short supply and some of the retailers that have had stock, have been price gouging so much that a Q9450 is almost double the price that it was meant to be - needless to say, I didn't bother

The alternatives were, a reasonably priced E8200 with it's low multi, I didn't really fancy that much, (although Supershanks has hit 550FSB with his here) a hugely price gouged E8400 for £190+, or an E8500 for £197 - Still overpriced, but a no brainer really, and I liked the idea of the 9.5 multi.

Setup went flawlessly, although it was only after I had done my first OCCT run that I realised I was still on the 9x multi, so I set the CPU N/2 Ratio to Enabled - As soon as this was enabled, the half multis appeared.

I am going to test with the LittleDevil single stage phase unit, which holds the E8500 at around -44c (-51c BIOS), and it is around this point in the review that I should point out that the clutter free CPU socket area on the X48-T2R makes it incredibly easy to insulate aginst condensation, which is a real plus point for a board of this nature.

Anyway, enough waffling from me, lets have a look at what this price gouged Wolfdale can do!

Starting with a relatively standard 3.6Ghz 1:1 overclock...

400x9 1.150v 3600Mhz 1:1 5-5-5-15 400Mhz RAM 333/667 Strap.

Click Image For Full Size.

400x9.5 1.28125v 3800Mhz 1:1 5-5-5-15 400Mhz RAM 333/667 Strap.

Click Image For Full Size.

422x9.5 1.28125v 4012Mhz 1:1 5-5-5-15 844Mhz RAM 333/667 Strap.

Click Image For Full Size.

As you can see, the only things that need tweaking are the CPU voltage and NB voltage. At this speed, we do not need any extra VTT, PLL or GTL tweakage

Voltage Settings
CPU VID Control: 1.28125v
CPU VID Special Add: AUTO
DRAM Voltage Control: 2.19v
SB Core/CPU PLL Voltage: 1.51v
NB Core Voltage: 1.356v
CPU VTT Voltage: 1.1v
Vcore Droop Control: Enabled
Clockgen Voltage Control: 3.45v
GTL+ Buffers Strength: Strong
Host Slew Rate: Weak
GTL REF Voltage Control: Disabled - Not needed yet.
x CPU GTL1/3 REF Volt: 110
x CPU GTL 0/2 REF Volt: 110
x North Bridge GTL REF Volt: 110

Quote:

5Ghz Milestone!

With my previous Wolfdale, the FSB suddenly stopped around 520 and regardless of voltage, wouldn't go much higher. I decided to feed the E8500 some serious voltage (1.6v+), and see how high I could go.

Much to my amazement, I was able to boot at 5Ghz (527x9.5) and run both Wprime tests, however, Superpi caused an instant crash, as did OCCT. CPU-Z was showing 1.77v which I don't believe to be correct, but I will measure with a DMM and confirm this later.

527x9.5 1.66v 5010Mhz 1:1 5-5-5-15 1054Mhz RAM 333/667 Strap.

Click Image For Full Size.

After upping the NB, VTT and Vcore again, I was able to boot at 540FSB

- Needless to say that this wasn't too stable, and I could only get as far as a CPUZ verification before the system started to become unstable.

540x9.5 1.68v 5130Mhz 1:1 5-5-5-15 1080Mhz RAM 333/667 Strap.

At this moment in time, I am going to say that I can't go any further, however, I haven't done any (or needed to) tweak the GTL REF settings, and I will also give SetFSB a run in the near future and add the results here, if there is anything of note.

No tweaking of GTL REF required... Hmmm, @ 500+ FSB the e8200's x8 multi appears to be less of a hindrance.

kennymester · 04-22-08

Cool man!

I just oc'ed my q6700 to 3.6ghz at 400x9. One thing I don't understand though is the strap. Wouldn't you want it to be the correct one for your memory? Like I have mine at 400/800 for my ram. Also what is the top voltage for the NB? I have mine at 1.5, is that too high?

Thanks

MasterKromm · 04-22-08

Quote:

Originally Posted by kennymester

Cool man!

I just oc'ed my q6700 to 3.6ghz at 400x9. One thing I don't understand though is the strap. Wouldn't you want it to be the correct one for your memory? Like I have mine at 400/800 for my ram. Also what is the top voltage for the NB? I have mine at 1.5, is that too high?

Thanks

As for the straps, try sticking to 333/xxx for anything 400FSB+, anything below 400FSB stay on the 266/xxx strap... Higher strap = higher NB latencies and to some extent decreased performance. Another effect of using a higher strap is increased stability @ high FSB and less strain on the NB which may lead to less voltage for stability...

1.5 isn't too high in and of itself, but do you need that much V for stability? There are a lot of factors to consider, but really 1.5v is overkill for only 400FSB... Here check out clunks settings for a Q6600 @ 400x9(1.35V was used for the NB):

Quote:

Test system:

Q6600 G0
2x1GB Ballistix Tracer D9 (Green slots - these proved to have better stability for 2x1GB DIMMS on the X38 for me).
8800GT
Thermalright Ultra Extreme 120
Corsair HX620
Raptor 74GB 16MB Cache.

Voltage Settings
CPU VID Control: 1.31v
CPU VID Special Add: AUTO
DRAM Voltage Control: 2.19v
SB Core/CPU PLL Voltage: 1.51v
NB Core Voltage: 1.35v
CPU VTT Voltage: 1.211v - Note that this is slightly higher than the X38's 1.205v
Vcore Droop Control: Enabled
Clockgen Voltage Control: 3.45v
GTL+ Buffers Strength: Strong
Host Slew Rate: Weak
GTL REF Voltage Control: Disabled - Not needed yet.

MasterKromm · 04-26-08

Some more experimentation -> thoughts...

CPU VID Special Add is an interesting feature, by taking it off auto and setting CPU VID control to auto you can keep c1e and EIST features enabled while maintaining your OC...

I've downloaded and read(only a little) through these two PDFs(archived so I could upload):

C2DInfo.zip

Core 2 duo Desktop processors Sequences (data sheet) PDF:

2.3 Voltage Identification Paragraph 4

Quote:

The processor provides the ability to operate while transitioning to an adjacent VID and its associated processor core voltage (VCC). This will represent a DC shift in the load line. It should be noted that a low-to-high or high-to-low voltage state change may result in as many VID transitions as necessary to reach the target core voltage. Transitions above the specified VID are not permitted. Table 5 includes VID step sizes and DC shift ranges. Minimum and maximum voltages must be maintained as shown in Table 6 and Figure 1 as measured across the VCC_SENSE and VSS_SENSE lands. The VRM or VRD used must be capable of regulating its output to the value defined by the new VID. DC specifications for dynamic VID transitions are included in Table 5 and Table 6. Refer to the Voltage Regulator-Down (VRD) 11.0 Processor Power Delivery Design Guidelines For Desktop LGA775 Socket for further details.

VRD 11.0 guidlines PDF:

2.6.1 Dynamic-Voltage Identification functionality

Quote:

VRD11 architecture includes the Dynamic Voltage Identification (D-VID) feature set, which enables the processor to reduce power consumption and processor temperature. Reference VID codes are dynamically updated by the processor to the VRD controller via the VID bus when a low power state is initiated. VID codes are updated sequentially in 12.5 mV steps and are transmitted every 5 microseconds until the final voltage code is encountered. Processors are capable of transitioning from
standard operational VID levels to either the VR11 or extended VR10 VID table minimum values. They are also capable of returning to a higher VID code in a similar manner. The low voltage code will be held for a minimum of 50 microseconds prior to sequentially transitioning through the VID table to a new voltage reference which can be any higher VID code, but is generally the original reference VID.

Basically what all that means is that the VID changes when power savings features (EIST C1E ect) are implemented... With prior motherboards where an OC required a bump in voltage to remain stable that meant(typically) disabling the power savings features. But since we can add the voltage we need to remain stable while maintaining stock(or auto) VID, the predetermined VID tables can still be used because the % is applied to both the load VID and low VID values. I think if I spent time reading through the documents I might even be able to determine the various VID ranges but I am not that ambitious or motivated.

Clarification:

While under load the power savings features(EIST or C1E) are not in effect thus the voltage is Stock(Auto) VID X the percentage chosen. But when the CPU isn't in use and the multi drops to x6 the VID value is automatically adjusted by the processor.

CPU VID Control: AUTO
CPU VID Special Add: 102.83%

Since my E4300's VID is 1.325 that alone isn't high enough to run 3.0ghz stable so I use the special add set to 102.83% to increase the core voltage to 1.3625. Now then I'll pick an arbitrary value(since I don't know the real value) that represents my adjusted VID when EIST or C1E kick in... Let's say 1.2v just to keep the math simple.

So that VID value is actually 1.2v x 1.0283 -> 1.234v for 334FSBx6 or 2.0ghz.

This is important to keep in mind too,

Quote:

Originally Posted by Clunk

I just spoke to Jimmy at DFI and he said that the reason they set the Special VID to work >1.2v is that when the Vcore is under 1.2v, there can be additional ripple caused if the Special VID is used, so to avoid any power issues, they decided to only allow the Special VID after 1.2v.

That's all the info I have, hope it fills in some blanks

chumley · 04-27-08

thanks 4 the vid info. ive been following clunks info and managed to get my 8500 stable at 4ghz using his settings but i couldnt go any higher. i think its my memory tho.

i can get stable at 4100 but i get errors in prime. im going to have a go at the gtlref thing once i can get my head round how it works. lots of numbers

MasterKromm · 05-20-08

I'm pretty much done with the "heavy" OCn/experimentation... It's just getting to hot here in FL for the use of any real voltage.

I was able to OC my q6700 to 405FSB x10 and 450FSB x9, both of which provided practically identical benchmark figures(duh). But in my quest for higher FSB the magical 500 mark eluded me. Has anyone else with a QUAD been able to run 500FSB? If so what voltages/settings did you use?

Axis01 · 05-20-08

Quote:

Originally Posted by MasterKromm

I'm pretty much done with the "heavy" OCn/experimentation... It's just getting to hot here in FL for the use of any real voltage.

I was able to OC my q6700 to 405FSB x10 and 450FSB x9, both of which provided practically identical benchmark figures(duh). But in my quest for higher FSB the magical 500 mark eluded me. Has anyone else with a QUAD been able to run 500FSB? If so what voltages/settings did you use?

You're doing better than I am. My B3 QX6700 is just too hot to get any higher than 3.5ghz without hitting high 70s and i'm using 350x10. I've backed it down to 3.4ghz till I can get this chip under water. What are your voltages and GTL settings to get your 3.9 or 3.6 for that matter. I've seen a couple 450FSB quads on XS but most are in the 400-430 range for max FSB and all of them are on water as well. I'm thinking i've seen a 500FSB quad but IIRC, it was a QX9650 or 97XX and it was chilled. I can't seem to find it right now though.

Axis

MasterKromm · 05-20-08

You might not believe this... But I had better results leaving my GTL refs on auto.

Temps got too hot for my liking(ambient 31°C load 63°C) when I stress tested @ 450FSB x8 so I only primed for an hour. I'm pretty sure someone else here on OCN has a quad @ 450FSB on the X48-T2R, I can't remember who though.

I only took 1 SS @ 450FSB, but here it is:

**grunion** · 05-20-08

450*8 right here, haven't tried any higher.
All voltages are set to auto, except vcore, vdimm, and NB.

However, when I get time I will tweak the GTL settings.
It's taking a bit more vcore than it took on my P5K for 450*8.

		Overclock.net - Overclocking.net > Intel > Intel Motherboards
DFI LT X48-T2R Initial Thoughts

04-27-08	#46 (permalink)
chumley New to Overclock.net Join Date: Mar 2008 Posts: 6 Rep: 1 Unique Rep: 1 Trader Rating: 0	thanks 4 the vid info. ive been following clunks info and managed to get my 8500 stable at 4ghz using his settings but i couldnt go any higher. i think its my memory tho. i can get stable at 4100 but i get errors in prime. im going to have a go at the gtlref thing once i can get my head round how it works. lots of numbers

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)