MSI P35 Neo combo is the first Intel P35 in our lab. This board is intended for the budget market, so it's a conventional board, not an "all-solid capacitors" design. It lacks RAID because this would've increased the cost and it only comes with a few accessories.
VIA VT6308P Firewire controller
Realtec RTL8111B network MAC
Fintek F71882FG Super I/O and VRM-controller
Passive cooling of the chipset
2Gb BIOS, version 1.00B11
1x PCIe x16
3x PCIe x1
1x Dual-Channel DDR2-slots for PC2-4300U/PC2-5400U/PC2-6400U memory up to 4GB
1x Dual-Channel DDR3-slots for PC3-6400U/PC3-8500U memory up to 4GB
4-Port SATA II
1-Port Marvell 88SE6111 SATA II
1-Port Marvell 88SE6111 IDE
Rear I/O Ports:
1x PS/2 keyboard
1x PS/2 mouse
1x parallel port
1x serial port
1x Gb LAN
4x USB 2.0
7.1-channel analog audio
1x SATA cable
1x HDD-power to SATA cable
1x Floppy cable
1x IDE cable
Bus Speeds: Auto, 400 to 2000 (200~500) in 1MHz increments
Memory Ratios: Auto, 1:1, 1:1.2, 1:1.25, 1:1.50, 1:1.67, 1:2.00
Memory Speed: Auto
Memory Voltage: 1.50V/1.80V to 3.00V/3.30V in various increments for DDR3/DDR2
PCIe Speeds: 100MHz~200MHz in 1MHz Increments
Core Voltage: Auto, +0.125V to +0.7875V in 0.0125V increments
FSB Options: Auto, Manual
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 2.80 in 0.05V increments
DRAM Timing Control: Auto, Manual
NB Voltage: 1.25V to 1.65V in 0.025V increments
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 1.60V in 0.025V increments
The power-voltage-regulator is a low-cost 3-phase analog part, we did not expect a 4-phase VRM, due to increased manufacturing costs. Of course this cheap implementation has drawbacks when it comes to overclocking, because higher VCore is required to achieve stable operation. The power-connector is positioned above the IDE connector and that may cause problems when using IDE devices, because the cables will intersect. On the other hand, the 4-pin CPU power connector is well placed on the outer edge next to the CPU socket. Due the use of some very tall conventional capacitors, you may have trouble to install high-end coolers, because some heatpipes might interfere. You should install the CPU power connector before attaching any high-end coolers.
The colour scheme of the memory slots is quite useless, two colours would have been much better. The DDR3 slots are on top with the DDR2 slots below.
iCH9 supports six SATA II connectors, but only four are fitted. An extra SATA connector is provided by the Marvell 88SE6111 controller, which also includes one IDE UDMA133 port. We hope it works better than the JMicro JMB363 controller.
Network connectivity is provided by the Realtek RTL8111B PCIe controller, the internal iCH9 MAC is not used. Although this is a budget board, a VIA VT6308P Firewire controller is integrated, but you can use only one port, because the bracket for the second port isn't supplied, but many cases do have internal connectors.
The panel connectors are colour coded and labelled, this is standard on all MSI boards. You'll also find four extra USB 2.0 headers, but you can't use them, because MSI does not provide any USB 2.0 brackets
MSI P35 Neo Combo
Intel Core 2 Duo E6700 (provided by Intel)
Scythe Andy Samurai Master (provided by Scythe-Europe)
Patriot 2GB PC2-6400U Kit PDC22G6400LLK (provided by Patriot)
CL5-5-5-15-CR2T at 1.80V
AMD ATI Radeon X1950XTX (provided by AMD)
OCZ ProXStream 1000W (provided by OCZ)
Seagate Barracuda 7200.9 500GB SATA (provided by Seagate)
SilenX iXtrema Pro 14dB(A) (provided by PC-Cooling.at)
The successor of the P965 seems a little bit faster compared to its older brother, conclusive results will be posted when we get our hands on some high-end motherboards and put them through our lab. The only new feature here is support for DDR3 memory, which is insanely expensive right now and will not result in faster operation due the added latency penalty. Some of the iCH9 derivates will have a hardware firewall feature, well known from Nvidia.
MSI has once again choosen an AMI BIOS, which is still in Beta. The recognition of SPD has some faults, the CPU-fan can't slow down when using 3-pin CPU-fans. There is also no setting for the command rate. To let anyone set the CPU-VCore at +0.7875V is quite insane and of course useless. MSI has a lot work ahead, we will keep you updated when a final BIOS arrives.
We could reach 320MHz without a problem, when setting the bus to 333MHz, the auto DRAM setting of the BIOS forced the memory to work at 1000MHz, which is hardly appropriate. Setting the memory ratio manual results in a bad setting of 800MHz with CL5-5-5-13. Any standard memory will fail, so booting is not possible.
The board does work at 1000MHz and maybe more, but this is a value board, so we don't expect to run the memory at 1200MHz, further tests will follow when we get the final BIOS.
Due the limitations and bugs of the BIOS we could only reach 3200MHz, but we needed to increase the VCore by +0.1500V which is way too high and much higher than any other board. This seems to be a result of the poor 3-phase VRM, final verdict will be done when we receive the final BIOS.
We didn't bother with synthetic benchmarks, because they won't tell you anything
interesting, we concentrate on real world applications, especially multimedia encoding.
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 manupulation, 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, but we only used the most common "undot", "FluxSmooth" and "MSharpen". Of course we also de-interlaced, filters were applied before any resizing took place (which is slower). We marked this as "slow".
If you need more info about filters, we recommend reading the doom9.org forum.
Due to the bug in the BIOS we needed to run our 3200MHz test with CL5-5-5-13-CR2T, which only overclocking friendy memory can provide, but shouldn't influence the results measurable.
The same episode we encoded, was used for our MP3-tests. 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 than the track-length. Fast memory does not play an important role here.
For your convenience we also show you the single-threaded benchmarks, 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 core.
We used these settings:
lamemt --vbr-new -q 2 -V 2 -m j --strictly-enforce-ISO --resample 48
This is a budget mainboard, we guess it's going to cost around €110 / $100 street price, so don't expect anything exciting. Due the BIOS bugs final testing could not complete, so we will give you an update when we can. We also heard that this is a prototype board and that the new BIOS will solve a lot of issues that we experienced.
It is not significantly faster than the 965, but it brings some cool features. We expect a new BIOS before we continue with the testing.