Value DDR2-800 from Super Talent: Where Can You Go?
by Wesley Fink on November 29, 2006 12:05 AM EST- Posted in
- Memory
Game Performance Comparison
Real-world benchmarks - specifically gaming benchmarks - provide the most useful measure of memory performance if you keep in mind what they represent. Memory is just one small part of overall gaming performance, and the AnandTech benchmarks keep everything the same except memory speed. Even the CPU speed is kept constant (except for the Highest Speed (overclocking) test). As a result performance improvements in FPS are very small because the only factor influencing the test results is memory speed. Many factors affect system performance, and memory speed is just one of those factors.
Results for high-end memory were very close at tested speeds. Therefore, the scale range was reduced to better show the small differences in these memory benchmark results at each speed. Please keep this in mind when viewing the charts. A normal zero scale would make performance differences appear much smaller than these expanded scale charts. Values for the tested memories are included below each chart for reference.
The AnandTech memory test suite uses Far Cry, Half Life 2 and Quake 4 for memory testing because these games are sensitive to memory performance and they are generally not GPU bound.
We have said before that the Core 2 Duo chips are not memory starved, and if you look closely you will see that the real world performance of this value RAM is only a bit slower in games than tested memory that costs twice as much. The Super Talent can't reach DDR2-1066, but with a modest overclock at DDR2-1000 the performance is still competitive with the top DDR2 memory we have tested at speeds to DDR2-1000 or higher.
There is no doubt the best memory gives you a few frames per second better gaming performance at all speeds, and it will also allow your system to reach higher CPU overclocks due to the added latitude of the memory. However, the real world difference between 4-4-4 timings and 3-3-3 timings at DDR2-800 is quite small.
It is interesting that memory is varying from 3-2-2-6 timings at DDR2-400 all the way to 5-5-5-15 timings at DDR2-1000. Despite the rapid drop in memory timings, all three games continue to show improvement in frame rates as memory speed increases. Put another way, performance continues to improve as memory speed increases. The Super Talent T800UX2GC5 performance was pretty much as expected. Results were virtually the same as the best DDR2 we have tested to DDR2-800, with a slight drop off due to slower timings at DDR2-800. The Super Talent could not reach DDR-1067, but it did perform well to DDR2-1000 with an 889 ratio pushed to a 300 bus setting (from the stock 266). Memory speed can definitely improve system performance, but not to the extent of an upgraded video card or a higher speed processor.
Real-world benchmarks - specifically gaming benchmarks - provide the most useful measure of memory performance if you keep in mind what they represent. Memory is just one small part of overall gaming performance, and the AnandTech benchmarks keep everything the same except memory speed. Even the CPU speed is kept constant (except for the Highest Speed (overclocking) test). As a result performance improvements in FPS are very small because the only factor influencing the test results is memory speed. Many factors affect system performance, and memory speed is just one of those factors.
Results for high-end memory were very close at tested speeds. Therefore, the scale range was reduced to better show the small differences in these memory benchmark results at each speed. Please keep this in mind when viewing the charts. A normal zero scale would make performance differences appear much smaller than these expanded scale charts. Values for the tested memories are included below each chart for reference.
The AnandTech memory test suite uses Far Cry, Half Life 2 and Quake 4 for memory testing because these games are sensitive to memory performance and they are generally not GPU bound.
We have said before that the Core 2 Duo chips are not memory starved, and if you look closely you will see that the real world performance of this value RAM is only a bit slower in games than tested memory that costs twice as much. The Super Talent can't reach DDR2-1066, but with a modest overclock at DDR2-1000 the performance is still competitive with the top DDR2 memory we have tested at speeds to DDR2-1000 or higher.
There is no doubt the best memory gives you a few frames per second better gaming performance at all speeds, and it will also allow your system to reach higher CPU overclocks due to the added latitude of the memory. However, the real world difference between 4-4-4 timings and 3-3-3 timings at DDR2-800 is quite small.
It is interesting that memory is varying from 3-2-2-6 timings at DDR2-400 all the way to 5-5-5-15 timings at DDR2-1000. Despite the rapid drop in memory timings, all three games continue to show improvement in frame rates as memory speed increases. Put another way, performance continues to improve as memory speed increases. The Super Talent T800UX2GC5 performance was pretty much as expected. Results were virtually the same as the best DDR2 we have tested to DDR2-800, with a slight drop off due to slower timings at DDR2-800. The Super Talent could not reach DDR-1067, but it did perform well to DDR2-1000 with an 889 ratio pushed to a 300 bus setting (from the stock 266). Memory speed can definitely improve system performance, but not to the extent of an upgraded video card or a higher speed processor.
Facebook
Twitter
RSS
28 Comments
View All Comments
Wesley Fink - Wednesday, November 29, 2006 - link
Super Talent has advised:"This kit is on sale at ewiz for $241.02. You could point readers to
http://www.ewiz.com/detail.php?name=T800UX2GC5">http://www.ewiz.com/detail.php?name=T800UX2GC5&quo...
The kits will also appear at other resellers in the near future.
Frumious1 - Wednesday, November 29, 2006 - link
Try searching for "T800UX2GC5" and you should find it.http://www.ewiz.com/detail.php?name=T800UX2GC5">eWiz
Not in stock anywhere else that I see right now, but Newegg has the T800UX2GC4 at $280 with a $20 mailin rebate, so I bet they'll get the C5 as well, and hopefully closer to $200. In the mean time, try http://froogle.google.com/froogle?hl=en&q=T800...">using a search engine like Froogle/
Postoasted - Wednesday, November 29, 2006 - link
Shouldn't we be suspicious of reviews where the test sample is provided by the product maker?Frumious1 - Wednesday, November 29, 2006 - link
Yes, and we should stop reading hardware websites because that's where all their hardware comes from. We should only pay attention to Newegg.com reviews, because all of those people really purchased the products they're reviewing! (/sarcasm)I've rarely (if ever) been able to match AnandTech performance results with the same RAM chips they use, but then I rarely have the same CPU and motherboard that they've got either. If they push everything to the same limit, you can at least figure the relative differences are there. Truthfully, I don't think more than a small fraction of people that worry about having the biggest epenis need more than DDR2-800 memory. That will get you just about everywhere you need to go with overclocking (except perhaps with the E6300/E6400 on extreme overclocks), so unless you care about the extra 3% potential performance there's not much reason to buy $500 RAM kits.
Wesley Fink - Wednesday, November 29, 2006 - link
Product on AnandTech is not a one-shot deal. If memory or motherboard manufacturers supply hand-picked hardware and users can't duplicate what we find the RMAs go through the roof. This is very expensive for the manufacturer. They quickly learn it is in their best interest to supply a sample with typical performance. The supply issue becomes self-policing.As Editors and enthusiasts we are also not idiots. We do buy samples on a regular basis and compare them to what we find with manufacturer samples. If results are out of line we scream loudly - to the manufacturer and in these pages. Accepting samples from manufacturers for review gets you information MUCH faster, but a review at AT is a privilege - not an obligation. Manufacturers who abuse the "typical sample" rule get moved down in queues or out of our review cycle.
The performance of the Super Talent is nothing spectacular; it is good performance from a fairly rated DDR2-800 memory. There is nothing in our results to raise any concerns. The point of the review was that value DDR2 is almost as good in performance as the best DDR2, and if you are on a tight budget you can save money with value DDR2, within reason, and get more performance by putting the difference in a video card upgrade or a CPU upgrade.
We have asked Super Talent to provide info on where this memory can be purchased. We will pass that along as soon as we receive an answer.
lopri - Wednesday, November 29, 2006 - link
What if the manufacturers are big, I mean HUGE, or if they command a monopoly/duo-poly status in the market? Namely, Intel, AMD, NV, ASUS, et al. Do they consider it an honor to be reviewed @AT? I remember Anand's E6600/E6700/X6800 all hitting ~4.0GHz when they debuted. Retail samples still can't achieve such clocks even months after the initial review, let alone at that time they were merely achieving 3.30~3.60GHz. But it'd be hard to ignore Intel's new products, I'd assume?Wesley Fink - Wednesday, November 29, 2006 - link
There is variation in overclocking among Core 2 Duo samples, but almost all of our retail chips - those we bought - will do in the 3.6GHz to 4.0GHz range. My last retail E6800 does 3.6+ at stock voltage and right at 4GHz on good air cooling. The retail runs 2100 FSB while the Intle sample will not do 1MHz over 1800 FSB.Intel supplied the pre-launch chips, but we have bought everything since. There is definitely variation, but our retail purchases do not vary significantly from the Intel supplied chips, except in maximum FSB which we commented on in the nVidia 680i launch review.
Hippiekiller - Wednesday, November 29, 2006 - link
Poop comes from butts teehee