I just recently did the same thing...MB supports a newer cpu but wouldnt give me the correct core speed til I updated the bios to the latest available for my MB.(installing the AMD64 3400+ instead of the 3100+Sempron actually lowered my clock speed from 1.8 to 1ghz!)

Same with any newer cpu I have ever installed.

This time around I was even lucky enough to get a modded bios flash that allowed overclocking.(the original emachines bios did not allow this)I flashed my bios from a bootable flash drive because I had no floppy drive.

HewlittPackard has a utility that will make a flash drive bootable...just point it at the correct files in your bios flash download(if your MB supports flash drive booting...it should)

If your not up on bios flashing...do some research,its potentially very dangerous to the MB.

Thanks guys. I have a motherboard update for the DP35DP, now I just have to painstakingly record all the current BIOS settings for when the flashing resets them. Ugggh. Never done this before. Ugggh.

I hope that works.

Pursuant to all this, I noticed in DXDiag the cpu is not using several instruction sets (SSE4 among others) which the 9650 should have available. Also, I must say I will be really pissed if this cpu turns out to indeed be onlky a Pentium III Xeon; I don`t think so, but you never know. 

Just thought to post this if you dont know how to set-up a usb drive so you can boot off it go here,

http://www.hardwarecanucks.com/forum/guides-how-tos/1457-how-making-usb-thumbdrive-bootable.html

If you have any problems dont be afraid to ask.

Luck to u.

Results from cpuz 1.47, although bear in mind I had "an error occurred while opening the driver":

Processor

Intel Core 2 Quad

Yorkfield

“Brand ID” was BLANK

“Package” was BLANK

45nm

“Core Voltage” was BLANK

Specification Intel Core 2 Quad

Family 6   Model 7   Stepping A

Ext. Family 6   Ext. Model 17   Revision E0

MMX, SSE, SSE2, SSE3, SSSE3, EM64T

Clocks (Core #0)

2999.5 MHz

“Multiplier” was BLANK

“Bus Speed” was BLANK

“Rated FSB” was BLANK

Cache

4x32 KBytes

4x32 KBytes

2x6144 KBytes

Selection – Processor #1 (unchangeable…)

Cores 4

Threads 4

Cache

L1 D-Cache   Size 32 KBytes   x4

8-way set associative, 64-byte line size

L1 I-Cache   Size 32 KBytes   x4

8-way set associative, 64-byte line size

L2 Cache   Size 6144 KBytes   x2

24-way set associative, 64-byte line size

“L3 Cache” was BLANK

Mainboard

Motherboard

Intel Corporation

Model DP35DP   AAD81073-206

“Chipset” was BLANK

“Rev. ” was BLANK

“Southbridge” was BLANK

“LPCIO” was BLANK

BIOS

Intel Corp.

DPP3510J.86A.0216.2007.0502.1916

05/02/2007

Graphic Interface

all entries here were BLANK

Memory

“Type” was BLANK

2030 MBytes

“Channels #” was BLANK

“DC Mode” was BLANK

“NB Frequency” was BLANK

Timings

all entries here were BLANK

Memory Slot Selection

J6H1 and a J6J1

1024 MBytes each

all other entries were BLANK

Look proper? To me this seems to indicate I did not get screwed, although that cpuz error prevented a complete report. I will do nothing else at this frightening time...

(grin BIOS-flash-apprehensively)

Alright, what that tells me is that your cpu is indeed a genuine Q9650 (The Yorkfield tag is for the Core2 family though... strange... and it is listing Quad_4 units further below - identifying it truly) and that its condition looks fair & solid. Such a verification is necessary before you attempt flashing a Bios to make sure the device is actually the right thing. Forget the Blanks, some of these are to be expected in that situation.

1)

Model DP35DP   AAD81073-206

by Intel, that is the most important info and from which you will find the proper flash_toolkit for that specific mobo.

2)

DPP3510J.86A.0216.2007.0502.1916

for the Bios tag is another key item since that is what flashing will alter to detect the new chipset. You'll possibly even have a new date once the process is completed.

- Flashing must be done **AS IT IS** explained by your mobo guide... i've seen hundreds of boards going beserk after a flash-upgrade simply because people had skipped an important step in manufacturer' instructions.

- Key here is to follow the process to the letter and make sure every steps are performed as they should.

A BIOS is an important system asset as it is refered to by the OS upon starting whatever hardware is to be detected **BEFORE** the declaration routines kick in. This happens so fast, i doubt anyone notices anyway... only problem is IF a bad value is entered or even changed later on (overclocking gimmicks can destroy a bus rail in 10 minutes, for example).

CPU switching is as easy as setting the recommended values (during the flashing process), firing the whole PC back up, and enjoying the 'increased' power.

The most common problems occur when compatibility prevents a Mobo from using a cpu at its optimal settings. Nothing can blast a system so bad it would stop responding... but performance issues may creep in once the Bios-Flashing has failed or (accidently or not, btw) was flawed.

In the flashing guide, there are a number of 'care steps' one can follow to create emergency backups - *just in case*.

In fact... it's not as simple as taking a chipset out and snapping another in its place. So, by following instructions *properly* you should be all set.

Good luck!

They need a utility or at least the capacity to record BIOS settings in a simple text file, so one could simply print them... this is going to take forever and a day...

Thanks again for the info. It looks like I`ll use the Windows flash util, since this sounds like the simplest method (important for me!) and the most commonly used (couldn`t be that bad). Supposedly I just run the .exe and reboot.

Regarding a Q9650 specifically, you folks think it reasonable for me to indeed expect a 15% to 20% performance increase?

Pursuant to BIOS settings, I`ve heard from other quarters that Intel defaults are actually quite superb, so I probably do not need to do any fiddling.

No! Maybe, yes from some *very* recent gaming stuff.

Not everyone taps into quad adressing yet.

Would you risk a 5% overclock to obtain as low as 2_FPS more?

I blow 1.35 volt into the E4600 rail and it stands at mostly 46C on average - they recommend 1.325!

I'll only start worry once that extremely solid default BIOS setting self-diagnoses me an automatic warning at the maximum 60C that i purposely saved into it instead of the 75C default. Gee, the NV8600GS (without a fan - i should add, btw) is on fire at 59C is it? - Yup but it is a DX10_Shader4 at 600+clockhypes- so, i have yet to detect any 'troubles' with quite efficient rendering and much higher FPS than what i had in the earlier PC.

Manufacturers know better and more about fair settings than JoeSchmo from web site X (including simple yours truly!), believe me.

Within reasonable limits... most (if not all) chipsets outlive a CMOS battery, many PCIe devices, four Satas, one OS & too many cranky users - anytime.

Okay - sounds like it is wise to just be grateful the BIOS procedure worked (!). Thanks folks.