Assembly And Overclocking

This time around, with an established Core 2 Quad CPU and P45-based motherboard, we have very little to report about the hardware assembly and operating system installation.
All of the components fit inside the Rosewill Wind Ryder case without complaint, including the long Radeon HD 4870 X2.
Our only concern was that the massive Xigmatek HDT-S1283 cooler would pose a problem and prevent the case cover from fitting, but although it appeared to be close, it never touched a thing.
As we mentioned, the case surprised us with its good finish, build quality, and quiet fans, which are somewhat uncommon traits to discover in such a low-priced case.
Once everything was put together, the Gigabyte GA-EP45-DS3L motherboard booted up without a hitch, using proper multipliers and frequencies without a BIOS upgrade. We didn't need to make a single tweak. The 32-bit Vista Ultimate operating system also installed without incident.
We happily continued by benchmarking the system without crashes or any other incidents of note. We then turned our attention to overclocking the happy little system.
Overclocking
As with our experience assembling the system, it wasn't much trouble at all to overclock this month's enthusiast system. Unfortunately, without a more expensive liquid-cooling setup, temperatures were once again our limiting factor.
We're not complaining about the job our Xigmatek HDT-S1283 did because it performed well for an air cooler in the situation in which it was placed, but there's no denying that a Q9550 can produce a lot of heat when pushed.
We tweaked some voltages and settings and were able to boot to a front side bus (FSB) speed of 400 MHz on the first try and a 3.4 GHz CPU speed. Encouraged, we tried to up the bus speed but encountered crashing.
We set the Vcore to 1.35 V and to 1.3 V for the CPU termination, MCH core, and ICH core. Memory was upped slightly to 1.9 V, and we limited the PCI Express (PCIe) bus to 100 MHz. We turned off virtualization and thermal throttling as well. With these settings we were able to set the FSB to 425 MHz, for a 3.61 GHz clock speed.
Without pushing further, we wanted to see how hot and stable the system was at this speed, so we performed some Prime95 testing. While the system was completely stable, the Speedfan monitoring application reported temperatures of 77 degrees Celsius at load. Since we accepted high temperatures in our Core i7 920 testing from our last SBM (and considering that the CPU would probably never be stressed that hard except in a synthetic benchmark like Prime95), we continued with the ~800 MHz overclock and called it a day. A clock speed of 3.62 GHz is an acceptable overclock for a Q9550 on air and we wanted to produce some results that a buyer might reasonably expect from similar components.
Sapphire's 4870 X2 card was the next overclocking focus. In our experience, dual-GPU cards like the 4870 X2 can be a little fussy when overclocking, but we were happy to see that the card would allow us to max out the Catalyst driver's Overdrive option without displaying any artifacts. Unfortunately, the driver maxed out at an overclock of 780 MHz on the core and 950 MHz on the memory, which was a mere 30 MHz core overclock and a 50 MHz memory overclock. While this won't break any overclocking records, it turned out to be almost identical to the overclock we got in our e8500/4870 X2 SBM build, which overclocked at 777 MHz core and 950 MHz memory.
Locked and loaded, we went forth into benchmarking territory...
Latest Build Your Own News
Latest Build Your Own reviews
- 23/12 – System Builder Marathon, Dec. 2011: System Value Compared
- 21/12 – System Builder Marathon, Dec. 2011: $600 Gaming PC
- 20/12 – System Builder Marathon, Dec. 2011: $1200 Enthusiast PC
- 19/12 – System Builder Marathon, Dec. 2011: $2400 Performance PC
- 14/12 – Power Supply 101: A Reference Of Specifications



Forgive me if I'm wrong, but wouldn't the RAM be running at 850MHz as opposed to 800?
Looking at the Crysis scors - This systems are 13ish frames faster than the $625 system. Its a lot of extra cash to make the game playable.
Spelling and the grammars to be added in later!
For UK:
CPU: £235
Graphics: £350
RAM: £75
Motherboard: £90
HDD: £45
CPU cooler: £18
Case: £25
PSU: £65
Optical drive: £22
Total: £925
all this looks very good, but how do you play games on this with "NO MONITOR". A system comprises of a pc and a monitor surely?..
£925 @ exchange rate = $1320 approx..yet again tailored to US prices..
i give up on thw.uk..
Rephy, it's not as simple as an exchange rate comparison, you must also take into account our thieving government which has VAT @ 15% (if your lucky) and import duty costs. Again the government likes to gets it's cut and so you pay import duty on the VAT and also the shipping!
1)First and foremost; this article proves that the E8500 is the winner when it comes to games. How?! It achieved better FPS in crysis where every frame counts, while producing -though lower than the Q9550 and i7 920-perfectly playable FPS in the other games and it is cheaper and produces far less heat. Perhaps I will be wronged if THW showed us the minimum FPS alongside the average FPS, I know the charts will be more complicated but it will add more value to the games benchmarks at no added testing effort..Don Woligroski, please consider this! You know, after years of gaming I saw it's not the average FPS that matters most, it's the steadiness of the frame rate that makes the playing experience smooth and I remember games (like oblivion) that had -on my machine-high FPS coupled with a choppy gameplay!
2)Although the case is said to have a good finish and silent fans, it does look ugly, but I can't complain about that at 30$!
3)A sillier complaint: why picturing the case's interior with a careless cable routing ?!!
i hope that 32bit OS was a typo, that would just be plain stupid
I really don't understand these kinds of articles. If anyone is going to the trouble of building a custom PC, surely they will tailor it to the user's specific needs? Anyone who isn't interested in doing the research will surely just buy a Dell etc. If this is your target audience wouldn't you be better off doing a controlled comparison of a few Dells at different price points? (Which would be pretty useful when recommending something for friends and family.) If your audience is the sort of person who will go to the hassle of a custom build then you would serve them better by reviewing individual components thoroughly and consistently.
I really don't understand these kinds of articles. If anyone is going to the trouble of building a custom PC, surely they will tailor it to the user's specific needs? Anyone who isn't interested in doing the research will surely just buy a Dell etc. If this is your target audience wouldn't you be better off doing a controlled comparison of a few Dells at different price points? (Which would be pretty useful when recommending something for friends and family.) If your audience is the sort of person who will go to the hassle of a custom build then you would serve them better by reviewing individual components thoroughly and consistently.
I think you are missing the whole point of SBMs..they "suggest" compnents that match each other in performance and test what the whole system is capable of. Of course any one is free to finetune the choices to his/her needs butI have to say building 3 systems monthly at different price points and comparing them is no small task and I know many peaple would like to copy these systems completely or partially. Besides, pre-assembled systems are often overpriced and no enthusiast would miss the joy of a DIY assembly and would apppreciate someone who gives general dirctions to him/her.
In the UK Mesh Computers systems compare favourably with buying the kit from Microdirect and DIY. No warranty with DIY.
uk sux