EVGA 680i LT SLI: NVIDIA's 680i Cost Reduced
by Gary Key on March 28, 2007 4:00 AM EST- Posted in
- CPUs
Overclocking - QX6700
Needless to say, quad core overclocking on the 680i platforms has been a huge disappointment until recently. There is a new hardware revision of the EVGA 680i board and the latest BIOS on the ASUS Striker now allows 400+ FSB operation. Our board reached a final 362FSB at the stock multiplier and this was a setting that we found to be very stable during testing. We were able to POST and enter XP at 368FSB but could not complete our benchmark testing. Once again, the lack of voltage options hindered our results but they are still very good.
We dropped the multiplier on our QX6700 and were able to hit 8x450. The board completed the majority of our benchmarks at this setting but could not pass quad Prime95 or our four application multitasking test without the number three core waving the white flag. We finally reached a benchmark stable setting of 432FSB and probably could have gone higher but we had one of the dreaded FSB holes in the 440FSB range.
The one problem noted during testing is that we had to set our processor voltage to 1.4750V in order to maintain system stability. Our processor normally will run in the 3600MHz range at 1.4125V, and this was the highest voltage we have required on a board to reach this range. The reason for this is that Vdroop is terrible on this board with the quad core processor. We would see drops as high as .07V under load testing with the average drop being around .04V during game testing. The board also liked to "sing" to us while overclocked with the quad core installed. It was the one noise emitting from our case that we found to be louder than the chipset fans.
While we sound a bit pessimistic, this board did deliver results beyond our expectations in quad core testing. We just feel like a couple of the "corner cutting" activities inside the BIOS limited us once again from reaching the board's true capabilities.
Test Setup
A 2GB memory configuration is standard in our XP test beds as most enthusiasts are currently purchasing this amount of RAM. Our choice of mid-range OCZ Flex XLC PC-6400 memory represents an excellent balance of price and performance that offers a very wide range of memory settings during our stock and overclocked test runs. We also utilized our Corsair XMS2 Dominator (Twin2x2048-9136C5D) memory on this board to verify DDR2-1066 to DDR2-1200 compatibility with another memory type.
We are currently completing testing of several other memory modules ranging from Transcend, TwinMOS, and WINTEC DDR2-800 down to A-DATA DDR2-533 for compatibility and performance benchmarks in our roundup article. Our memory timings are set based upon determining the best memory bandwidth via MemTest86 and test application results for each board. We only optimize the four main memory settings with sub-timings remaining at Auto settings.
We are utilizing the MSI 8800GTX video card to ensure our 1280x1024 resolutions are not completely GPU bound for our motherboard test results. We did find in testing that applying a 4xAA/8xAF setting in most of today's latest games created a situation where the performance of the system at our standard resolution did not really change. Our video tests are run at 1280x1024 resolutions for this article at High Quality settings. We also tested at 1600x1200 and 1920x1200 4xAA/8xAF for our NVIDIA SLI results that will be presented in our next article.
All of our tests are run in an enclosed case with a dual optical/hard drive setup to reflect a moderately loaded system platform. Windows XP SP2 is fully updated and we load a clean drive image for each system to ensure driver conflicts are kept to a minimum. We were able to run our memory test modules at 3-4-3-9 1T for our benchmark results and will be updating the results of our other boards with 1T operation shortly.
| EVGA 680i LT SLI Quad Core Overclocking Testbed |
|
| Processor: | Intel Core 2 Extreme QX6700 Quad Core, 2.66GHz, 8MB Unified Cache 1066FSB, 10x Multiplier |
| CPU Voltage: | 1.4500V (default 1.3500V) |
| NB Voltage: | 1.40V |
| FSB Voltage: | 1.40V |
| Cooling: | Tuniq 120 Air Cooling |
| Power Supply: | OCZ ProXStream 1000W |
| Memory: | OCZ Flex XLC PC2-6400 (2x1GB) (ProMOS Memory Chips) |
| Video Cards: | 1 x MSI 8800GTX |
| Hard Drive: | Western Digital 150GB 10,000RPM SATA 16MB Buffer |
| Case: | Cooler Master CM Stacker 830 |
| Maximum OC: | 362x10 (3-4-3-9 2T, 797MHz, 2.20V), CPU 1.4500V 3625MHz (+36%) |
| Maximum FSB OC: | 432x8 (3-4-3-9 2T, 798MHz, 2.20V), CPU 1.4500V 3460MHz (+62% FSB) |
| . | |
 |
| Click to enlarge |
Needless to say, quad core overclocking on the 680i platforms has been a huge disappointment until recently. There is a new hardware revision of the EVGA 680i board and the latest BIOS on the ASUS Striker now allows 400+ FSB operation. Our board reached a final 362FSB at the stock multiplier and this was a setting that we found to be very stable during testing. We were able to POST and enter XP at 368FSB but could not complete our benchmark testing. Once again, the lack of voltage options hindered our results but they are still very good.
 |
| Click to enlarge |
 |
| Click to enlarge |
We dropped the multiplier on our QX6700 and were able to hit 8x450. The board completed the majority of our benchmarks at this setting but could not pass quad Prime95 or our four application multitasking test without the number three core waving the white flag. We finally reached a benchmark stable setting of 432FSB and probably could have gone higher but we had one of the dreaded FSB holes in the 440FSB range.
The one problem noted during testing is that we had to set our processor voltage to 1.4750V in order to maintain system stability. Our processor normally will run in the 3600MHz range at 1.4125V, and this was the highest voltage we have required on a board to reach this range. The reason for this is that Vdroop is terrible on this board with the quad core processor. We would see drops as high as .07V under load testing with the average drop being around .04V during game testing. The board also liked to "sing" to us while overclocked with the quad core installed. It was the one noise emitting from our case that we found to be louder than the chipset fans.
While we sound a bit pessimistic, this board did deliver results beyond our expectations in quad core testing. We just feel like a couple of the "corner cutting" activities inside the BIOS limited us once again from reaching the board's true capabilities.
Test Setup
| Standard Test Bed Performance Test Configuration |
|
| Processor: | Intel Core 2 Duo E6600 (2.4GHz, 4MB Unified Cache) |
| RAM: | OCZ Flex XLC PC2-6400 (2x1GB) 2.20V, 3-4-3-9 (1T where applicable) (ProMOS Memory Chips) |
| Hard Drive: | Western Digital 150GB 10,000RPM SATA 16MB Buffer |
| System Platform Drivers: | Intel - 8.1.1.1010 NVIDIA - 9.35, 8.43 ATI - 6.10 |
| Video Cards: | 1 x MSI 8800GTX |
| Video Drivers: | NVIDIA 97.92 |
| CPU Cooling: | Tuniq 120 |
| Power Supply: | OCZ ProXStream 1000W |
| Optical Drive: | Plextor PX-760A |
| Case: | Cooler Master CM Stacker 830 |
| Motherboards: | ASUS Striker Extreme (NVIDIA 680i) - BIOS 1002 ASUS P5N-E SLI (NVIDIA 650i) - BIOS 0401 ASUS P5B-E (Intel P965) - BIOS 0402 DFI LANParty UT ICFX3200-T2R/G (AMD RD600) - BIOS 12/22 EVGA 680i LT SLI (680i LT) - BIOS Award 721N0P01 GIGABYTE GA-N680SLI-DQ6 - BIOS F4B Intel D975XBX2 (Intel 975X) - BIOS 2333 MSI P6N SLI Platinum (nForce 650i) - BIOS 1.22 |
| Operating System: | Windows XP Professional SP2 |
| . | |
A 2GB memory configuration is standard in our XP test beds as most enthusiasts are currently purchasing this amount of RAM. Our choice of mid-range OCZ Flex XLC PC-6400 memory represents an excellent balance of price and performance that offers a very wide range of memory settings during our stock and overclocked test runs. We also utilized our Corsair XMS2 Dominator (Twin2x2048-9136C5D) memory on this board to verify DDR2-1066 to DDR2-1200 compatibility with another memory type.
We are currently completing testing of several other memory modules ranging from Transcend, TwinMOS, and WINTEC DDR2-800 down to A-DATA DDR2-533 for compatibility and performance benchmarks in our roundup article. Our memory timings are set based upon determining the best memory bandwidth via MemTest86 and test application results for each board. We only optimize the four main memory settings with sub-timings remaining at Auto settings.
We are utilizing the MSI 8800GTX video card to ensure our 1280x1024 resolutions are not completely GPU bound for our motherboard test results. We did find in testing that applying a 4xAA/8xAF setting in most of today's latest games created a situation where the performance of the system at our standard resolution did not really change. Our video tests are run at 1280x1024 resolutions for this article at High Quality settings. We also tested at 1600x1200 and 1920x1200 4xAA/8xAF for our NVIDIA SLI results that will be presented in our next article.
All of our tests are run in an enclosed case with a dual optical/hard drive setup to reflect a moderately loaded system platform. Windows XP SP2 is fully updated and we load a clean drive image for each system to ensure driver conflicts are kept to a minimum. We were able to run our memory test modules at 3-4-3-9 1T for our benchmark results and will be updating the results of our other boards with 1T operation shortly.
Facebook
Twitter
RSS
16 Comments
View All Comments
MrWizard6600 - Thursday, March 29, 2007 - link
dont know the full details, but the newest P26 Bios from EVGA Removed LinkBoost support“LinkBoost was removed from nForce 680i SLI because it did not show significant demonstrable benefit in games. We had hoped newer games would take advantage of this additional bandwidth but this has not been the case. Please note that future BIOS upgrades will only remove the automatic overclocking component of LinkBoost. Users can still manually overclock the PCI-Express and HyperTransport buses in the BIOS."
SniperWulf - Wednesday, March 28, 2007 - link
Any chance I can get that registry key that you deleted after the nv4_disp crash?yacoub - Wednesday, March 28, 2007 - link
It's time to start recommending Supreme Commander to be the RTS game test choice. Now that there's a new RTS game that can crush systems (SupCom) the way FEAR did for FPS games for the last couple years, a game that makes use of multiple cores at that, we should start to see it used in the suite of games in future motherboard, cpu, and gpu tests. Perhaps Anandtech have already started planning for this?yyrkoon - Wednesday, March 28, 2007 - link
Well, it is not an RTS game, but one I play frequently should also be used to test gaming rigs in my humble opinon. Titan Quest (Immortal Throne if deemed nessisary).My gaming Rig is not top of the line, but it is no slouch, and this game constantly drags in areas that are low light areas, or are outside, and night time. Anyhow, and game that brings my opteron 1210 / ABIT NF-M2 nView / 2 GB DDR2 6400 memory, eVGA 7600GT KO to a crawl, SHOULD be used as a benchmark, at least once in a while. F.E.A.R does not lag like this game does on my system . . .
Stele - Wednesday, March 28, 2007 - link
I agree to an extent, as the current test suite's figures are starting to show their age a wee bit... 100+ FPS, as Quake ended up with newer CPUs/GPUs. Having said that, Anandtech was probably trying to ensure consistency in its tests, so that at the very least, past test results for other products can always be quite easily compared over time. Perhaps in the near future we might see it yet! :)JarredWalton - Wednesday, March 28, 2007 - link
I believe Gary is looking at benching with Supreme Commander. That's one of the games that will be added in future reviews, as far as I know.yacoub - Wednesday, March 28, 2007 - link
Awesome, glad to hear that! One can see the benefit to maintaining benchmark titles for as long as possible to allow the greatest comparison back to past reviews. On the up side, SupCom looks to easily task systems well into the future (much like FEAR has thus far since its release) and would be an excellent choice for the RTS benchmark game. :)yyrkoon - Wednesday, March 28, 2007 - link
Firstly, I have to get this out. It is a completely LAME excuse for nVidia, eVGA, Foxconn, or whomever, to 'claim' the reason this board does not have passive cooling (OTES or equivelent), is to help cut costs. 20 bux less, and they rip out passive cooling ? Lame. Take ABITs NF-M2 nView $85usd , including passive cooling, and moderate other features. Granted, it is based on the NF4 chipset, but who cares, it has passive cooling, and loads of other features for the price.I have to agree somewhat with the first poster, this board is dead, before it even reaches market. Crippled features, missing other features, all for to save $20-$50 usd. Sorry nVidia, and whomever else, but you have not only completely wasted your time, but ours as well, for having to read a review on such a shitty product.
yacoub - Wednesday, March 28, 2007 - link
very much agree. the passive cooling is (and should be) an integral feature of current generation motherboards. there's no reason ANY motherboard released today should require active northbridge (or southbridge) chipset cooling (unless you're overclocking, obviously).yyrkoon - Wednesday, March 28, 2007 - link
One thing I did keep wondering the whole time I was reading the article, and you were mentioning how the BIOS was basically a crippled version of the full blown i680 SLI version was: 'I wonder how soon someone would mod a BIOS version based on the i680 SLI chipset BIOS, to add removed 'features' such as certain voltage settings, etc.'Also, Jarred, or whomever else would like to answer, how hot does the PWM section of this motherboard run ? Seems lately, dual core on up motherboards tend to run hot in the PWM.