NVIDIA 680i: The Best Core 2 Chipset?
by Gary Key & Wesley Fink on November 8, 2006 4:45 AM EST- Posted in
- CPUs
Memory Performance
The AMD AM2 has essentially removed concern about the Memory performance of chipsets by providing a low-latency memory controller directly on the processor. Intel continues to integrate the memory controller in their chipsets, and the Intel Core 2 chipsets achieve competitive memory performance and low latency with caching schemes and read-ahead algorithms. Because Intel has done such an excellent job of providing stellar memory performance in the 975x/P95 chipsets, any chipset that hopes to compete with Intel would have to perform similarly in memory controller performance.
To assess the effectiveness of the NVIDIA 680i chipset we measured memory performance with Everest 3.50 from Lavalys and compared results to the top-end Intel 975X chipset.
The results of memory performance on the NVIDIA 680i are truly impressive compared to the 975X. NVIDIA has clearly produced a competitive memory controller for Intel Socket 775, a task which has eluded other chip makers who have tried to compete on this platform. Results can only be called equivalent in any of the Everest benchmarks.
Another widely used measure of latency and memory bandwidth is ScienceMark 2.0.
ScienceMark confirms the results with Everest 3.5. NVIDIA has built a memory controller in the new 680i chipset that is every bit as good as the outstanding Intel memory controller in the 975X. This is not a minor achievement.
The final confirmation of memory controller performance with the 680i chipset comes with SiSoft Sandra 2007. The most common measurement is buffered or Standard Memory Bandwidth.
The AMD AM2 has essentially removed concern about the Memory performance of chipsets by providing a low-latency memory controller directly on the processor. Intel continues to integrate the memory controller in their chipsets, and the Intel Core 2 chipsets achieve competitive memory performance and low latency with caching schemes and read-ahead algorithms. Because Intel has done such an excellent job of providing stellar memory performance in the 975x/P95 chipsets, any chipset that hopes to compete with Intel would have to perform similarly in memory controller performance.
To assess the effectiveness of the NVIDIA 680i chipset we measured memory performance with Everest 3.50 from Lavalys and compared results to the top-end Intel 975X chipset.
| Everest 3.5 Memory Performance | ||||
| Chipset | Read | Write | Copy | Latency |
| NVIDIA 680i | 8045 | 4865 | 5506 | 54.8ns |
| Intel 975x | 7751 | 4868 | 5512 | 52.3ns |
The results of memory performance on the NVIDIA 680i are truly impressive compared to the 975X. NVIDIA has clearly produced a competitive memory controller for Intel Socket 775, a task which has eluded other chip makers who have tried to compete on this platform. Results can only be called equivalent in any of the Everest benchmarks.
Another widely used measure of latency and memory bandwidth is ScienceMark 2.0.
| ScienceMark 2.0 Memory Performance | ||
| Chipset | Latency (512 byte stride) | Memory Bandwidth |
| NVIDIA 680i | 37.12ns | 5449.37 MB/s |
| Intel 975x | 37.81ns | 5430.69 MB/s |
ScienceMark confirms the results with Everest 3.5. NVIDIA has built a memory controller in the new 680i chipset that is every bit as good as the outstanding Intel memory controller in the 975X. This is not a minor achievement.
The final confirmation of memory controller performance with the 680i chipset comes with SiSoft Sandra 2007. The most common measurement is buffered or Standard Memory Bandwidth.
| Sandra 2007 Standard Memory Performance | ||
| Chipset | INT (Integer) | FLT (Float) |
| NVIDIA 680i | 5969 | 5944 |
| Intel 975x | 5949 | 5953 |
Facebook
Twitter
RSS
60 Comments
View All Comments
Wesley Fink - Thursday, November 9, 2006 - link
The other time you might need a fan on the northbrdige is when using water cooling or phase-change cooling. There is no air-flow spillover from water-cooling the CPU like there is with the usual fan heatsink on the CPU, so the auxillary fan might be needed in that situation.Wesley Fink - Thursday, November 9, 2006 - link
The 680i Does NOT require active notrthbridge cooling and is shipped as a passive heatpipe design. At 80nm it is much cooler than the 130nm nVdia chipsets. The fan you see in the pictures is an included accessory for massive overclocking, much like Asus includes auxillary fans in their top boards.In our testing we really did not find the stock fanless board much of a limitation in overclocking as the northbridge did not get particularly hot at any time. We installed the fan when we were trying to set the OC record and left it on for our 3 days at 2100 FSB. Since it is a clip and 3 screws to install we left it on.
IntelUser2000 - Monday, November 13, 2006 - link
That's funny. A cooler running one consuming more power. Must be the die size is much larger :D.
yacoub - Thursday, November 9, 2006 - link
ah okay thanks for that clarification! =)yacoub - Thursday, November 9, 2006 - link
NTune would be a lot more interesting if it wasn't so slow to respond to page changes, cumbersome, and a gigantic UI realestate hog.The same functionality in a slimmer, more configurable, and efficient UI design would be highly desireable.
yacoub - Thursday, November 9, 2006 - link
and actually, that goes for the entire NVidia display/GPU settings configuration panel.Khato - Wednesday, November 8, 2006 - link
Each CPU is going to have a max FSB clock that it'll run stably at for the same reason that it has a max core logic frequency. The main difference here is that you have two possible barriers: signal degredation due to the analog buffers not being designed for such high speed and then whatever buffer logic there is in the CPU to clock cross from FSB to core not liking the higher frequency. I'm kinda leaning towards the buffer logic being the limiting factor, since I'd expect the manufacturing variance in the analog buffers to be minimal. That and the described 75MHz variance in top FSB frequency between various processors sounds reasonable for non-optimized logic.Staples - Wednesday, November 8, 2006 - link
I have no need for SLI. Makes the board more expensive and an SLI setup is just not worth it to me. I was about to buy a P965 chipset but now I am interested in a the 650i Ultra. Will we see a review of this chipset in the future? Most of it seems to be exactly the same as the 680i however it does lack some features and I am afraid that those missing features may affect performance. As it stands now, do you expect the performance of the 650i Ultra to perform identical to the 680i SLI?Gary Key - Wednesday, November 8, 2006 - link
We do not, we do expect the 650i SLI to perform closely to it. We will have 650i boards in early December for review. :)
Pirks - Wednesday, November 8, 2006 - link
is this functionality where you can overclock your CPU and FSB and memory on the fly without rebooting Windows available only on nForce mobos? I'm a stability freak and I want to be able to raise and lower my clocks and voltage on the fly, similar to the way Macs do this - they spin their fans under load and become totally quiet when idle - I wanna do the same so that my rig is dead quiet when idle/doing word/inet/email/etc and becomes noisy and fast OCed beast when firing up Crysis or something. and I want this Mac-style WITHOUT rebooting Windowsso do I have to buy nVidia mobo for that?
600i series only or earlier nForce 4 or 5 series will do as well?
I still can't dig what's up with these "dynamic BIOS updates that _require_ reboot to work" - so can you OC without rebooting or not? if yes - what are these BIOS options that nTune changes that DOES require reboot?
could you happy nTune owners enlighten me on that stuff? thanks ;)