For nearly two months we have struggled with the 1.0 version of this mainboard. Luckely MSI found out it needed to fix some issues with a new revision, so they send us version 1.1. While things now run smoothly, some BIOS issues still has to be fixed.
Realtek ALC888T Revision 1.0, Realtek ALC888 Revision 1.1
Realtek RTL8111B PCIe network MAC
Marvel 88SE6111 IDE/SATA controller
Fintek F71882FG Super I/O and VRM-controller
passive cooling of chipsets/VRM
8Mb BIOS, version 1.00, 1.10, 1.20, P08 with Revision 1.0, 1.32 with Revision 1.1
Mainboard Revision: 1.0, 1.1
1x PCIe x16, 1x PCIe x16 @ x4 speed
2x PCIe x1
2x Dual-Channel DDR2-slots for PC2-5400U/PC2-6400U memory up to 8GB
4-Port SATA II featuring RAID 0, 1, 5, 0+1, JBOD
2-Port eSATA II
1-Port SATA II
1x PS/2 keyboard
1x PS/2 mouse
2x eSATA II
1x Gb LAN
6x USB 2.0
7.1 Audio analog
1x SP/DIF digital out
1x 1-port Firewire bracket
4x SATA cable
2x 1x HDD-power to 1x SATA power cable
1x Floppy cable
1x IDE cable
FSB Options: Auto, Manual
Bus Speeds: Auto, 200 to 799 in 1MHz increments
Memory Ratios: Auto, 1:1, 1:1.2, 1:1.25, 1:1.5, 1:1.66, 1:2
Memory Speed: Auto, Manual
PCIe Speeds: 100MHz~200MHz in 1MHz Increments
Core Voltage: Auto, 1.275V to 2.0625 in -0.0125V increments
CPU Clock Multiplier: Auto, manual if EIST disabled
Core 2 Duo: 6x-11x in 1X increments - Core 2 Duo, downwards unlocked
Core 2 Extreme: 6x-16X, downwards unlocked
DRAM Voltage: Auto, 1.80V to 2.10V in 0.05V increments, 2.10V to 3.30V in 0.10V increments
DRAM Timing Control: Auto, Manual
NB Voltage: 1.25V to 1.60V in 0.025V increments, 1.65V
SB I/O Voltage: 1.50V, 1.60V, 1.70V, 1.80V
SB Core Voltage: 1.05V, 1.15V
FSB VTT Voltage: 1.20V to 1.55V in 0.025V increments
The colour scheme is quite pleasing, a black PCB with yellow and black memory slots and red and black SATA connectors.
The MSI P35 Platinum is of course an "all-solid" design, and was expected initially at a €150,- price tag.
The voltage regulator is an 4-phase analogue part, although we would have prefered a digital VRM. The power connector is positioned under the memory slots.
The 8-pin CPU power connector is odd placed between the heatpipes of the chipset coolers. This will give you trouble during install, when you have already mounted the cpu-cooler. With revision 1.1 MSI has installed an additional adapter to the CPU-power-connector so installation is more easy now. But we can't understand why MSI did not place the connector on the right edge of the board.
Revision 1.0 Revision 1.1
The heatpipes stick close to the CPU keepout area, so removal of bigger CPU-coolers is a mess.
The ICH9R supports six SATA II ports, but lacks support for IDE. The purple SATA ports are SATA masters. We would have liked to remain all chipset ports for internal use, because any SIL/Marvell controller can do eSATA II too, but MSI decided to take the other way around. They implemented a Marvell IDE/SATA combo, so you get an an single internal SATA port and one IDE. Two SATA ports will interfere with long two-slot graphic cards such 8800 GTX and MSI did not include 90° angled SATA cables.
Network connectivity is provided by one Realtek RTL8111B via PCIe. An VIA VT6308P provides two ports of FireWire 400. MSI uses a standard Realtek audio codec, this time the ALC888T which has special support for VOIP, but only with revision 1.0 which was not on any shelves. With Revision 1.1 MSI dumped the VOIP support and selected the more cheaper ALC888, we guess that saves a buck or two.
Revision 1.0 (sorry, bad picture :)
The case connectors are colour coded, but you need the manual if you forget the codes, because they are not described on the board itself.
The configuration of the slots is not perfect, as using dual slot graphic cards means you loose one of the PCI slots. It would have been nice to install a 3rd PCI Slot, because PCIe cards are still not common. Between Revision 1.0 and 1.1 no differences to speak of.
The heatpipe-construction, we call them "Looping", keeps the temperature of the P35 under 60°C even at 1.55V. But during testing while it got 30°C inside our lab, we noticed, it may not keep the temperatures in safe limits. So we recommand a fan to support the Northbridge cooler. The colour of the construction indicates it was made from copper, but MSI just uses cheaper aluminium and coloured it. MSI did not change anything between revisions.
During our initial tests, we discovered some problems with overclocking the memory. It seems the problems where severe, so MSI decided to make a new revision. Check out the differences for yourself:
MSI P35 Platinum 1.0 (provided by MSI)
MSI P35 Platinum 1.1 (provided by MSI)
Intel Core 2 Duo E6700 (provided by Intel)
Scythe Andy Samurai Master (provided by Scythe-Europe)
Kingston 2GB PC2-9600U Kit KHX9600D2K2/1G (provided by Kingston)
CL5-5-5-15-CR2T at 1.90V@ 800MHz
CL5-5-5-16-CR2T at 2.30V@ 1066MHz/1200MHz
Corsair Dominator 2GB PC2-10000U Kit TWIN2X2048-10000C5D (provided by Corsair)
CL5-5-5-15-CR2T at 1.90V @ 800MHz
CL5-5-5-16-CR2T at 2.40V @1066MHz/1200MHz
AMD ATI Radeon X1950XTX (provided by AMD)
Silverstone Element SF50EF-Plus (provided by Silverstone)
Western Digital WD4000KD (provided by Ditech)
SilenX iXtrema Pro 14dB(A) (provided by PC-Cooling.at)
The successor of the P965, is may be slightly faster, with the only difference being support for 45nm CPUs and DDR3 memory. With prices sky-high, DDR3 is not an issue as yet.
MSI has choosen an AMI BIOS. The overclocking settings are set in the two menus. The memory timing submenu is still the most user unfriedly so far. While any other vendor will give you the choice to set it manually or auto, MSI doesn't. If you disable SPD you have to set every timing by yourself. Its also annoying that some OC modules do support tWTR timings of 3 but the lowest setting is 4.
After fideling around we noticed the BIOS sets values as it pleased.
The first row is the SPD readout from Everest, the second row is what the BIOS sets when using SPD, the third shows the manual entered values and the forth what the BIOS did with them.
We are working with MSI to fix this issue but it does take longer than we expected. We guess that extra latencies happens because of the chipset, but no explanation was given, only "other vendors" do the same, which is not satisfying at all.
We did of course try 500MHz, and we saw the light - at least a BIOS screen. Stable operation was not possible, some P965 boards can do this without a problem. With 490MHz we could boot up windows, but applications will crash. Reducing the FSB to 485MHz, was the maximum stable setting to run some test. This is not that bad, but we expected more, especially with the Revision 1.1 board.
The board supports 1066MHz due overclocking, with the Revision 1.0 we could not run it stable with even 1000Mhz in certain situations. Revision 1.1 changed that. Memory clock with 1067MHz does run without any problems. With 1200MHz the situation changed, Corsair did not like to boot up into Windows, Kingston worked, but crashed with GordianKnot. Other boards can do better. The interesting part is, we do use a BIOS 1.32 which does still not show on the MSI website.
When setting VCore in the BIOS do not expect that is the VCore you will receive. Using default it will set to 1.275V with our CPU, but of course checking with cpu-z and Everest we saw it is at least 0.025V higher. Overclocking went well up to 3.66GHz, higher overclocks did not work stable, but 3.66GHz is what we expected, nothing changed with the board revisions.
Before benchmarking we checked the reference frequency. At 266.66MHz, 320MHz and 333.33MHz FSB the board was slightly overclocked, so we re-calculated any benchmarks results to "should be" frequencies.
Also note we did the tests with the Revision 1.0 board, which is slightly faster than the Revision 1.1 with the 1.32 BIOS. But the difference is marginal.
We didn't bother with synthetic benchmarks, because they won't tell you anything interesting, we concentrate on real world applications, especially multimedia encoding. Everest will report above 9,300MB/s and SiSoft above 7,200MB/s when benchmarking memory-reads.
Gordian Knot/XVID 1.1.2:
For our Gordian Knot testing we took an PAL epsiode from "Babylon 5" with a length of 41 minutes, 57 seconds and 8 frames.
We tried to "emulate" the most common usage of Gordian Knot:
1st: We have a perfect master, so we only de-interlace the content and resize it, without any other manupulations, we marked this as "fast".
2nd: You get bad mastering on many DVDs, especially "old" stuff or when the studios are in a hurry for the release. In this case you like to improve the picture quality, which is done by filtering the content. You can choose from tons of filters for any purposes you can think of, we only used the most common "undot", "FluxSmooth" and "MSharpen". Of course we also de-interlaced, filters were done before any resizing took place (which is slower). We marked this as "slow".
If you need more infos about filters, we recommand reading the doom9.org forum.
UPDATE 2007-08-03: Sorry, we have displayed cut benchmark-results.
The same episode we encoded, we used for our MP3-testing. We don't recommand using MP3 for encoding, because AC3 can do the job better, but nearly 42 minutes gives us approximately the length of any given album.
A measurement in seconds, as many sites do, is useless, because the differences are too small. So we used the built-in play/CPU ratio, this means the CPU is encoding x-times faster then the track-length. Fast memory does not play an important role here. For your convenience we show you also the single-threaded benches, they will be produced with any other L.A.M.E. version, because only lameMT can do more than one thread and take advantage of a second dice.
We used this setting: lamemt --vbr-new -q 2 -V 2 -m j --strictly-enforce-ISO --resample 48
This time we checked both idle and full power consumption.
The P35 will take a bit more power than the P965, but still less then any nForce chipset.
Compared to the ASUS P5K Deluxe power consumption is really low, it seems the VRM is very effective, even it is only a 4-phase design. ASUS does better with load, but still some more consumption. Of course the board does not have fancy features such as WLAN or second LAN, but we have disabled them on the ASUS boards anyways.
This board is a fine addition to everyone, as long as you can live with some BIOS issues. The operation is stable and it does consume much less power compared to the ASUS P5K, especially in idle mode. We liked the fact that it overclocks well and works really stable.
The board has a quite good layout, with the only issue of the cpu power connector, where the power cord will hinder the airflow of case-fans on the back. The sata connectors should have placed more on the edge of the board, because big two-slot graphic-cards like 8800 GTX will get in the way and there is plenty of space on the board, so relocating some other stuff would have been wise.
To be a top overclocker board it lacks 1200MHz memory support at least with our test modules and the current BIOS, this may change in the future. For an average user the board works quite well and if you can live with 485 MHz FSB this should be the board for you. MSI could go better with accessories.
The MSI P35 Platinum retails between €140,- and €150,- this is about €30,- to €40,- cheaper than ASUS P5K Deluxe and it will also save some cash on your power-bill.