ASRock 775Dual-VSTA: Does DDR2 matter?
by Gary Key on August 8, 2006 6:35 AM EST- Posted in
- Motherboards
Memory Performance
Our VIA based ASRock motherboard provides two DDR2 and three DDR memory ratios. The majority of end-users will select the memory ratio that matches their memory speed. We are testing four ratios at the fastest stable timings we can achieve and still pass our benchmark test suite. With these set ratios, CPU speed remains the same at 1.86GHz in our test platform with memory speed being varied by selecting the different ratios.
There are some downsides to this approach. With the memory controller in the chipset, instead of part of the processor as in AMD Athlon 64 systems, there is a small performance penalty for speeds other than a 1:1 ratio (DDR2-533 in this case). However, the penalty is in reality very small, though the performance between various chipset designs can vary a great deal as we will see in a future article.
Due to performance reasons we did not test the DDR266 setting as we believe most people will not need this setting. Although the BIOS offers a 1T Command Rate, we never could get this setting absolutely stable without drastically raising the memory latency settings. We even tried our more expensive memory modules with the same results. Our memory settings were derived from extensive stress testing with a variety of applications. While certain settings that allowed lower latencies worked well with certain applications, the final settings we arrived at had to work with all applications.
The memory performance result at DDR2-533 (1:1 ratio) leads the other memory speeds in all cases. This particular setting delivered the best raw performance although we will see in our application and game benchmarks that this advantage is negated by other platform components.
Normally memory bandwidth improves with increases in memory speed and reductions in memory timings. To evaluate memory bandwidth SiSoft Sandra 2007 Professional was used to provide a closer look at scaling. As we have been saying for years, however, the Buffered benchmark usually does not correlate well with real performance in applications on the same computer. For that reason, our memory bandwidth tests have always included an UnBuffered Sandra memory score. The UnBuffered result turns off the buffering schemes, and we have found the results correlate better with real-world performance.
In this case, we find the combination of our memory settings at DDR2-667 and the VIA memory controller generates Sandra results that are up 32% lower than the DDR2-533 settings. The DDR2-533 results are up to 48% better than the DDR-333 scores with the DDR2-667 and DDR-400 scores being comparable. The Sandra memory score is really made up of both read and write operations. It is also a synthetic benchmark that does not always reflect real world performance.
To provide more detail on the impact of memory performance, we also compared pure number crunching with Version 1.5 of Super Pi, using the time to calculate 2 million places of Pi at the different memory speeds. Our other memory test is the latency measurement from the latest version of Everest. The results are interesting as the DDR2-667 ties DDR2-533 in Super Pi while DDR-400 trails slightly and DDR-333 brings up the rear. Looking at the latency figures, DDR2-667 is almost equal to DDR2-533 and DDR-400 trails slightly, while the results for DDR-333 are terrible and follow the Sandra bandwidth figures when compared to our DDR2-533 scores.
Our VIA based ASRock motherboard provides two DDR2 and three DDR memory ratios. The majority of end-users will select the memory ratio that matches their memory speed. We are testing four ratios at the fastest stable timings we can achieve and still pass our benchmark test suite. With these set ratios, CPU speed remains the same at 1.86GHz in our test platform with memory speed being varied by selecting the different ratios.
There are some downsides to this approach. With the memory controller in the chipset, instead of part of the processor as in AMD Athlon 64 systems, there is a small performance penalty for speeds other than a 1:1 ratio (DDR2-533 in this case). However, the penalty is in reality very small, though the performance between various chipset designs can vary a great deal as we will see in a future article.
Due to performance reasons we did not test the DDR266 setting as we believe most people will not need this setting. Although the BIOS offers a 1T Command Rate, we never could get this setting absolutely stable without drastically raising the memory latency settings. We even tried our more expensive memory modules with the same results. Our memory settings were derived from extensive stress testing with a variety of applications. While certain settings that allowed lower latencies worked well with certain applications, the final settings we arrived at had to work with all applications.
 |
| Click to enlarge |
The memory performance result at DDR2-533 (1:1 ratio) leads the other memory speeds in all cases. This particular setting delivered the best raw performance although we will see in our application and game benchmarks that this advantage is negated by other platform components.
Normally memory bandwidth improves with increases in memory speed and reductions in memory timings. To evaluate memory bandwidth SiSoft Sandra 2007 Professional was used to provide a closer look at scaling. As we have been saying for years, however, the Buffered benchmark usually does not correlate well with real performance in applications on the same computer. For that reason, our memory bandwidth tests have always included an UnBuffered Sandra memory score. The UnBuffered result turns off the buffering schemes, and we have found the results correlate better with real-world performance.
In this case, we find the combination of our memory settings at DDR2-667 and the VIA memory controller generates Sandra results that are up 32% lower than the DDR2-533 settings. The DDR2-533 results are up to 48% better than the DDR-333 scores with the DDR2-667 and DDR-400 scores being comparable. The Sandra memory score is really made up of both read and write operations. It is also a synthetic benchmark that does not always reflect real world performance.
To provide more detail on the impact of memory performance, we also compared pure number crunching with Version 1.5 of Super Pi, using the time to calculate 2 million places of Pi at the different memory speeds. Our other memory test is the latency measurement from the latest version of Everest. The results are interesting as the DDR2-667 ties DDR2-533 in Super Pi while DDR-400 trails slightly and DDR-333 brings up the rear. Looking at the latency figures, DDR2-667 is almost equal to DDR2-533 and DDR-400 trails slightly, while the results for DDR-333 are terrible and follow the Sandra bandwidth figures when compared to our DDR2-533 scores.
Facebook
Twitter
RSS
55 Comments
View All Comments
shambf00 - Monday, December 4, 2006 - link
Gary, please, how were you able to change the memory ratio?I have DDR 400mhz on this motherboard with the E6300, however, the motherboard sees it as DDR 333mhz even after I set the speed manually.
Can you (or anyone else who knows about this) tell me where to change the memory ratio so my ram can run at it's normal speed?
Thanks!
lumbergeek - Friday, February 9, 2007 - link
The only way I was able to change the memory ratio (this board with E6300, 2GB Mushkin low-latency DDR400 and an AGP ATI 1650pro) was to play around with the manual setting for RAM and FSB - the board seems to select the multiplier on it's own. If anyone knows how to adjust it manually, I'd love to know about it myself!tomppi - Thursday, August 24, 2006 - link
..have had some 2-3-2-6-1 kingston memory for 2 years now (KHX3200AK2_1G)used to be on a ASUS A7N8X-E Deluxe. only did 2.5-3-3-12-1
now with the asrockdual-vsta i can't keep it stable unless i run at ~3-3-3-12-2 (400mhz)
why can't I even come close to the speeds shown in this article :(
GoatMonkey - Wednesday, August 9, 2006 - link
I have a possibly stupid question. Will this board work with only one stick of memory or do they need to be added in pairs. I have a single 1GB DDR400 stick in my Athlon XP3200 that would be nice to be able to use, but if I have to buy new memory anyway I might as well get the DDR2.Gary Key - Wednesday, August 9, 2006 - link
No Question is Stupid. :)However, sometimes my answers might appear that way. ;-)
The board works fine with a single stick of memory but you will be regulated to single channel operation. This places a 3%~7% performance penalty on most applications although you will need benchmarks to tell the difference in most cases.
johnsonx - Wednesday, August 9, 2006 - link
Perhaps it might be worthwhile to show performance numbers for the single-channel situation in one of the upcoming articles? I'm sure many people are in similar situations. My 754 machine just has a single 1Gb DDR400 stick as well, so if I were to consider a value Core 2 cpu & board combo (of the sort Fry's may offer in the coming months) I might want to continue using that memory if it will work and not completely cripple the performance.veryevilmike - Thursday, August 10, 2006 - link
Is the asrock 775-HDTV rev2 (using the ati xpress200 chipset) one of the budget boards on the review list? It is limited to single channel but was contemplating this as a stopgap for a few months until rd600 & co arrive, when the board would become a cheap htpc. its also one of the only uATX conroe boards available.All up, congrats on doing such a useful 'realistic' review early on in the piece - makes a nice change from all the super-expensive stuff that is not in the picture for 99% of people.
GoatMonkey - Wednesday, August 9, 2006 - link
Sounds good to me. I could go with that for a while and later upgrade to the DDR2 and new video card, and a better motherboard at some point after that. An e6600 should be available for around $350 (one of these days anyway), and then this board for under $75 or so. Roughly $400 to $425 for a significant upgrade and an easy path for other components sounds good to me.Sunrise089 - Wednesday, August 9, 2006 - link
Now after the most obvious subject line of all time, here's what I'm thinking:I continue to be impressed with Conroe's performance, but grow more and more dissapointed with Conroe's performance scaled to price. With s939 basically any price paid could get you very near FX-57 speeds. Cheap DFI Infinity board + Cheap Opteron 144 or a lucky 3000+ = 2.6+ghz on air with a good cooler. An FX-57 might hit 3.0ghz on air if you were very lucky.
Conroe doesn't seem to work that way. Sure the top 2 or 3 processors perform the same, give or take, but the lower cache processors loose a lot of performance (much more than the above mentioed Athlons). Then, getting a cheaper motherboard really lowers maximum overclock. Then high-performance memory costs a lot more, and probably adds more performance as well. Finally really cheap parts like this totally cripple performance.
I think as more people come to relize this we will see a reality check from the community, where many on the high end will end up with awesome Conroe systems, but many others will realize with their budgets a nicely upgraded s939 X2 system with their existing servicable DDR memory and a newly discounted X1900XT will be a better overall platform dollar for dollar. This user is certainly beginning to lean that way. And heck, by the time I have to upgrade for a second time Conroe should be available with a 1066mhz FSB anyways.
Paladin165 - Wednesday, August 9, 2006 - link
"the lower cache processors loose a lot of performance"Wheres your evidence for this? Hasn't it been established that the performance hit is around 3% on average? (http://www.anandtech.com/cpuchipsets/showdoc.aspx?...">http://www.anandtech.com/cpuchipsets/showdoc.aspx?...