Hi,
I recently bought C2D 4300 and mainboard Gigabyte GA-945P-S3 Rev. 1.0 to set up my own C2D rig. I'm a budget system builder and have a little idea of overclocking the CPU, not to the extreme but at least, hitting 2.7 GHz,( I know E4300 can give more than that). With the rest of the components as stated in my signature, I set up to overclock. The mobo is Rev. 1.0 that comes with F2 bios. I downloaded BIOS F4 as it is the latest and the flashed the mobo.
As usual, this is what I do to the BIOS prior to overclocking:
- Disable CPUID Max to 3
- Disable No-Execute Memory Protection
- Disable C1E
- Disable TM2
- Disable EIST
- Disable Smart Fan Control Method
The CPU is installed with the stock Intel HSF, and the memory is Kingston KVR667, at 2GB (2 x 1 GB DDR2 667).
As an indication, this is a review of the overclocking potential of this mobo, which is competitive with the ever popular 956P-DS3.
Hardcoreware Review on 956P-DS3 and 945P-S3After enabling hidden options in the BIOS, I enabled CPU Host Frequency at 201, mem multiplier at 3, PCIE freq. at 100 MHz, no overvoltage and memory still at Auto). This stage is just for testing purposes only, and what did I get? After saving BIOS, the system is like halting for a little longer than a typical restart, and then it booted up without the litlle overclocking than I did. I entered BIOS and I saw that BIOS automatically disabled CPU Host Frequency. After a while, I found out that by
setting PCIE Frequency back to Auto, than this problem is solved. I tested this scenario with 210, 215 FSB and this scenario stays the same. I call this
the first anomaly!!

I wonder how many of you experience the same.
After fixing PCIE Frequency to Auto, I went on to overclock the CPU further. I put 230 FSB, and the system passed without problems. Then I select
233 FSB (which means 233 x 9 = 2097 MHz), and the system also passed this setting. Easy Tune 5 reported the
CPU temp at 47C and
+12V reading at 11.92V. So, I said hmmm...pretty good so far.....
However unfortunately that is the last setting that the mobo allowed me to go. Hitting 234 and 235 FSB and so on only resulted in BIOS disabling the CPU Host Frequency, falling back to default CPU setting. Bear in mind that:
- typical E4300 and even E6300 overclockers only start upping core voltage after exceeding 2.7 GHz at least.
- which means, rarely overclockers will start upping FSB voltage before it nearing 280 or 300 FSB (depend on CPU multiplier)
- I don't have to increase DIMM voltage as at 233 FSB x 2.66 mult resulted in 620 MHz which is way below my DDR2 667 rating.
- CPU temperature still doesn't exceed 50 C.
- at 233 FSB, the +12V rail is still reporting 11.92V, no matter what program that I used (Everest included)
Therefore, I'm stuck at 233 FSB where I dont have to overvoltage the CPU, FSB, DIMM, PCIE and not even tweaking the memory timings. My question is: am I hitting what they call ' FSB Wall', or am I having insufficient power, or my mobo just had its manufacturing defect encountered???
I know many will question my PSU, CM Extreme Power 380W being the culprit here, furthermore I also have 7800GT installed in the system. Well, let me give a hint : under my previous AMD64 3000+, Abit KN8 NF4, 1GB DDR400 (and the rest of the components stays the same), I manage to overclock to the Venice core to 250x9=2250 MHz, which indicates how much power I can squeeze from the same PSU. Now, I am hoping that under 65 nm process, 65W Allendale-CPU powerhouse, I can free more power requirement for my rig, which hopefully translates to better overclocking.
However, I need some inputs and ideas from you guys to help solve this problem. The approach is like this: first I am hoping for those who has the same problem as indicated as the first anomaly above to share their experience, and then maybe I need to reconsider if it is true that I need to change the PSU. My thought is that I don't want to buy a new PSU but still having the same problem due to FSB Wall or whatever.
Please, experts needed.