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
MikeyC - Friday, November 10, 2006 - link
I'm looking forward to this. Any idea on when you guys will have the bin numbers for the different rates of OC-ability? I'm planning on OCing my e6600 on this board this weekend; I'll post up my numbers if that'll help.Gary Key - Sunday, November 12, 2006 - link
We have not figured it out yet. Two CPUs from the same week and they both act differently during overclocking. We are still working with NVIDIA on this matter.Joepublic2 - Friday, November 10, 2006 - link
965 and 975 boards to my knowledge don't support a FSB/mem ratio smaller than 1:1. Does this chipset have the right multiplier to use DDR2-400 while retaining a 1066Mhz FSB?Gary Key - Friday, November 10, 2006 - link
The memory settings are sync or async capable on this chipset if you unlink the FSB and Memory in the BIOS. So the answer is yes to your question but believe me this chipset needs good DDR2-800 to get the most out of it. A 1T command rate can make a significant difference in several applications and games. We already found a 4FPS difference in Q4 at 1280x1024 with DDR2-800 at 1T instead of 2T as an example. We will have more on this in our actual board review.Joepublic2 - Monday, November 13, 2006 - link
I was asking mainly because a conroe board that could run a 4:3 FSB/mem multiplier could be an even better overclocker than the 965. One would only need RAM that could hit DDR 752 for a 500Mhz FSB for example.A great review as always!
VooDooAddict - Friday, November 10, 2006 - link
Any idea if there are mATX boards using any of thse new chipsets on the way?Gary Key - Friday, November 10, 2006 - link
There is the possibility of the 650i Ultra being on a mATX board in late January. However, the suppliers might wait for the new NV Intel IGP chipset coming in Q1. We should have more information in December.
BadThad - Thursday, November 9, 2006 - link
Should be CPU, the X6800 is not a GPU, lol.
Wesley Fink - Friday, November 10, 2006 - link
Correctedyacoub - Thursday, November 9, 2006 - link
Do all of the 680i SLi boards require active cooling on the northbridge? That's actually a deal-killer for me, as motherboard fans are about the worst ones out there anymore since they're small, fast-spinning (and due to those two characteristics they are noisy), usually short-lived, and I've yet to see one that is dynamically controlled by the temp of the northbridge.I'm guessng 650i boards don't require active cooling, but are any of the 680i boards using a non-reference design sporting completely silent cooling?