EVGA P55 FTW a closer look

This board looked interesting so I picked up the FTW to take a closer look at it. This is my first EVGA board. Nice sleek monotone coloring of grays and black with a touch of red. The board it self is very lite and compact looking. This board has a separate physics slot which I might just try out with a 9800GTX or GTS250 which ever I can get the best deal on. I'm going to run 2x 285 in SLI also.
The OC BIOS is a little different than I'm used to with different names, organization and incrementing values. Best of all right out of the box, it booted first time absolutely no problems. There's always that first moment with a new motherboard when you checked that everything is plugged in and then you hit power button on and cross your fingers that it will boot up. This EVGA powered up without a glitch.

Where are going to start with the i7 860 and then later with 750(due in on Friday 9/18). Same G.Skill 2k RipJaws (awesome memory) and the Zalman 1156 cooler. Thermaright has a new HSF called the MUX-120 (True version for 1156) for all you TRUE fans.

Core i5/i7 architecture and FTW highlights video.

EVGA FTW product info
http://www.evga.com/products/moreInfo.asp?pn=132-LF-E657-KR&family=Motherboard Family

FTW Driver Links here.
Fleet
http://www.evga.com/eleet/
Drivers & Bios
http://www.evga.com/Support/Drivers/default.asp

pci view:

back:

front:


cpuz: "stock" setting. Notice the "Auto" setting in BIOS for CPU volt is 1.223 and Multiplier is 18 on this board.

Would be interesting to see if it does better than my LE, now open the gas and let it scream

I only see 8 VRM phases on that board, correct?

Spec is 12+2 on the CPU and 3 on the memory.

If you mean those little boxes around CPU socket, yes I count 8. I checked specification and it's says 12+2 analog design. I also see 5 more black boxes on the board.

Good, should yield to a more stable OC at higher voltages. So where are the OC snappies????

Okay starting to figure the EVGA Bios settings. Here first descent stable OC. More to come yet.

Nice going...these 860's seems to run hot, mine is a little fireball as well, much hotter than the 750.

keep it going!

Brolloks said:

Nice going...these 860's seems to run hot, mine is a little fireball as well, much hotter than the 750.

Click to expand...

how much hotter?

This FTW has been more challenging for me to OC than the Maximus III but got it to 4.312 with some work and a few blue screens. It's turbo seems to be tied to speed step. So if you want x22 multiplier then speed step is enabled otherwise you get 21x multiplier(see in pics 2 & 3). In the Maximus III these were separate bios settings. I'm still trying to find the wining combination but overall I really like this board.

4.312 oc


Pic 2 Turbo on x22 multiplier


Pic 3 Turbo off x21 multiplier

Up next are two Vantage runs. I ran these to get a “gaming experience” bench Mark and to compare to same runs done on a i7 920 with same cards using a X58 Rampage Extreme II.

First run is with a single EVGA 285 running at stock clock on the i7 860 with turbo on and clocked so that it will be at 4.0ghz when in Turbo mode.



The second Vantage run is SLI with (2) 285 over-clocked (trying to match a previous run similar on a 920) on same I7 860 and clocked the same.



My thoughts after looking at the results are that the p55/i7 860 combo is very close to the x58 i7 920 combo. So for me it just comes down to which I like better for Gaming as they both perform relatively the same or equal.

Conclusion:

I really like this EVGA FTW board. The sleek black and gray look just grows on you or is the winking red EVGA light at the old NB location. It blinks at the rate of a slow heartbeat. It is nicely laid out with lots of room for your SLI configuration and tons of USB ports. I used the 2 power connections from the PSU (Corsair 1000W) and I did notice a lot more stability. I didn’t get a chance to add in the 9800GTX for physics (no card) but it would just add to the gaming experience which is why I am doing this in the first place.

Anyway being my first EVGA board is was a learning experience to get used to BIOS and its terminology and its clocking tools. Its implementation of the i7 860 turbo feature was road block at first. Seems like with turbo ON it has a mind of its own but really its doing what is was designed to do and that is OC your computer for you when it’s under a load.

I now had to set bios based on the expected end result. I wanted to use my old technique and control it myself but that lead to one blue screen after another. So if you calculate what the GHZ will be based on block * 22 and then set bios less than that and use *auto for voltage, everything goes much better. I always set the memory specifications manually and everything else on *auto except for block (set at 170 – 180) and multiplier (set lower like at 20 or less). The i7 860 is a different technology and it is going to take some time to get used to it doing the OC work for you.

I do recommend a buy for the FTW and I7 860 combination For The Win!


Still to come is a run with the i5 750 on this FTW board.

Looking promising! I just ordered an FTW myself. I liked my EVGA 680i so much i had to go EVGA. Got my i5-750 in today Batch#: L925B606T hoping to hit 4.5GHz with water like another poster did with the ASUS board. Mind if i post my results here next week??

Please do, anyone is welcome to join this FTW thread

alien what batch you got?

Am i correct that you're using the 1600MHz ripjaws? I got the 2000 for my build. Also what was the vantage score difference for the SLi runs from the X58 to the P55.Everyone is talking about the x8/x8 lane bandwidth of the P55.

rubicsphere said:

Also what was the vantage score difference for the SLi runs from the X58 to the P55.Everyone is talking about the x8/x8 lane bandwidth of the P55.

Click to expand...

im still not convinced that x8 on pcie 2.0 is going to be limiting factor in a 2 card sli/cf setup. GPU's starting with NV8800GTS barely started to use more then x8 PCIE 1.0a bandwidth. since PCIE 2.0 doubles the bandwidth at x8 you get PCIE 1.0a x16 bandwidth. though as some of the numbers show, LGA1156 is slower in sli/cf but not by much. to make the point more clear, pcie 1.0a x8 bandwidth is AGP 8x bandwidth.

sources:
Accelerated Graphics Port

AGP 8x
A 32-bit channel operating at 66 MHz, strobing eight times per clock, delivering an effective 533 MHz resulting in a maximum data rate of 2133 MB/s (2 GB/s); 0.8 V signaling.

Click to expand...

PCI Express

In PCIe 1.x , each lane carries 250 MB/s. PCIe 2.0, released in late 2007, adds a Gen2-signalling mode, doubling the rate to 500 MB/s. PCIe 3.0, currently in development (for release around 2010), will add a Gen3-signalling mode, at 1 GB/s.

As a point of reference, a single-lane PCIe 1.x slot has nearly twice as much bandwidth as the most common PCI interface, a 32-bit 33 MHz PCI bus (133 MB/s). A four-lane PCIe 1.x slot has bandwidth comparable to the fastest version of PCI-X 1.0 (64-bit 133 MHz.) An eight-lane PCIe 1.x slot has a transfer rate comparable to the fastest version of AGP.

Click to expand...

rubicsphere said:

Am i correct that you're using the 1600MHz ripjaws? I got the 2000 for my build. Also what was the vantage score difference for the SLi runs from the X58 to the P55.Everyone is talking about the x8/x8 lane bandwidth of the P55.

Click to expand...

Nope, I got the 2000 RipJaws and they are killer !!

SLI Standoff

Asus RE II x58 & Core i7 920 D0




EVGA FTW P55 & core i7 860


Variance:

3d Mark Score 920 vs 860

P27863 vs P27296 2.07 %

GPU Score
23906 vs 23579 1.38 %

CPU Score
55357 vs 51787 6.89 %

GPU Test 1
72 vs 71.02 1.37 %

GPU Test 2
67.98 vs 67.05 1.38 %

CPU Test 1
3556.71 vs 3400.96 4.57 %

CPU Test 2
176.40 vs 163.18 8.10 %

Feature Test 1
1497 vs 1493.72 2.19%

Feature Test 2
14.70 vs 14.85 1.01 %

Feature Test 3
81.28 vs 81.15 .16 %

Feature Test 4
32.79 vs 32.15 1.99 %

Feature Test 5
42.82 vs 42.64 .42 %

Feature Test 6
93.87 vs 93.75 .12%


The biggest area of variance is in the CPU score 6.89% higher and not the GPU score only 1.38 % variance. From the data I'd conclude that the PCIx concern is mute. The bigger issue is the 920 CPU just being overall 6.8% faster.

Overall the 3D Mark scores are really close with just a 2.07% difference. There is a variable that needs to be considered even though I tried to make them the same, it is entirely possible that one combo maybe slightly higher in OC of the CPU and or the 285s. It was not a tightly closed test but a fair test.

Evilsizer said:

alien what batch you got?

Click to expand...

Batch #: L9221B884T

I believe this is a good batch as I hit 4.4+ on air ! and probably higher if I knew what I was doing...

Evilsizer said:

im still not convinced that x8 on pcie 2.0 is going to be limiting factor in a 2 card sli/cf setup. GPU's starting with NV8800GTS barely started to use more then x8 PCIE 1.0a bandwidth. since PCIE 2.0 doubles the bandwidth at x8 you get PCIE 1.0a x16 bandwidth. though as some of the numbers show, LGA1156 is slower in sli/cf but not by much. to make the point more clear, pcie 1.0a x8 bandwidth is AGP 8x bandwidth.

Click to expand...

My feelings exactly maybe when we get our hands on the 5870x2 or whatever nvidia calls their next dual gpu card we will see a difference. Thanks for the info.

And to Alien: We almost have the same build! like i said before i hope i get similar results.

rubicsphere said:

And to Alien: We almost have the same build! like i said before i hope i get similar results.

Click to expand...

What are you using for HSF ?