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
fenacv - Tuesday, January 8, 2008 - link
http://www.pricebat.ca/EVGA-122-CK-NF67-T1-LGA-775...">http://www.pricebat.ca/EVGA-122-CK-NF67...-SLI-ATX...If you don't really care the prefermace, I found it's onsale just buy one only 138 bucks. It's cheap.
TheBeagle - Friday, December 8, 2006 - link
I'm wondering if these touted new 680i boards are vaporware, especially the Gigabyte GA-N680SLI-DQ6 board. Ever since you first alerted us to the fact that the 680i chipset was replacing the 590 version, I've been waiting to see this whole new array of motherboards. However, aside from a few boards (ASUS and a few others) the major board manufacturers haven't been forthcoming with these products. Maybe this is just going to be some sort of a big Christmas present that Santa delivers on the holiday. If you guys at AnandTech have some info on this, I'd sure like to hear about. Thanksmbf - Wednesday, November 29, 2006 - link
I, for one, am going to seriously miss the native hardware firewall of the nForce3 and nForce4 chipsets, so I'm anything but "thankful" for seeing it "jettisoned into deep space." Actually, this was one of the coolest features of the nForce chipsets and truly innovative.nVidia's stance as to removing it because the functionality is built into Windows Vista doesn't ring true. A software solution can never work as efficiently and transparently as a hardware solution. And what of the people having no intention to switch to Windows Vista, and there are many reasons for not wanting to. They're practically left out in the cold.
I second the opinion that nVidia probably botched the hardware in some form or other, although the hardware firewall works quite well on my nForce3 250gb based system, once you get familiar with its quirks. This actually doesn't bode well for nVidia's "inventiveness" and "forward-thinking" (think DualNet), since chances are nVidia will drop support completely rather than work out the bugs that inevitably will be there. Removing the hardware firewall is the best example of this.
Also, and this is a bit off-topic in regard to the rest of this topic, wasn't there supposed to be ECC memory support in the new northbridge for the 680i chipset? I remember reading about the northbridge also being used in the new nForce Pro series chipsets. Another feature that has been removed in the mean time?
skrewler2 - Monday, November 13, 2006 - link
How was the Tuniq Tower 120 on the board? I've heard lots of people complaining about backplates not fitting right on this board because the back of the mobo has lots of capacitors... Did you need to do any modding or did it just work?Gary Key - Monday, November 13, 2006 - link
I used the Scythe Infinity in my testing, Wes used the Tuniq. I did try the Tuniq and it was okay with an extra pad on the backplate that negated any damage to the capacitors.mlau - Monday, November 13, 2006 - link
Did you test a recent Linux kernel on this board?Which components are supported (I don't care about "raid"),
and how buggy are the HPET, (IO-)APIC and ACPI implementations?
Governator - Saturday, November 11, 2006 - link
First off, very well done article guys, but I've a question on the layouts with regards to PCI slots so far with the Asus and Evga; are we to expect similar layouts with upcoming boards from other manufacturers? I ask because I'm planning on a water cooled SLI setup upon a 680i and am planning on an X-Fi card but not sure if I'll be able to use the middle PCI slot, TIA...Gov
Gary Key - Sunday, November 12, 2006 - link
Most of the 680i boards have the same basic layout. On the Asus Striker board you should be able to use the X-FI with most watercooled SLI setups as an example. It will all depend on your setup but you can kiss the middle PCIe slot good-bye. ;)Governator - Saturday, November 18, 2006 - link
Hi Gary, sorry I meant to reply sooner but thanks for this. I'm hoping I'll be in good shape with the fact that I'll be using the new 8800GTX water block codeveloped by BFG Tech from Danger Den which appears that it'll only take up one slot allowing for that bottom PCI slot to go to the X-Fi card, thoughts? TIA ;)deathwalker - Friday, November 10, 2006 - link
I wonder if there are Matx mobo's in the future for the 600 series chipsets.