First Look: AM2 DDR2 vs. 939 DDR Performance
by Wesley Fink on April 17, 2006 12:05 AM EST- Posted in
- CPUs
Socket 939 Fast DDR-400 vs. AM2 Fast DDR-2
DDR has a somewhat narrow operating range. Mainstream DDR might extend from DDR400 to perhaps DDR433. The very best, and most expensive, DDR might extend from DDR400 to a bit over DDR600. The useful range of fast DDR is even less because as the timings increase for higher speed the bandwidth starts to fall. Therefore top bandwidth with many of the top DDR products peaks somewhere in the DDR500 to DDR550 range and bandwidth normally drops as we move higher in frequency at poorer timings.
While ratios can be used with lower performing memory, the main purpose of ratios is to offset the performance limitations of cheaper memory so the user can enjoy the increased performance of overclocked processor without having to similarly invest in expensive memory capable of 1:1 memory overclocking. The point is that we will not be investigating memory ratios on DDR memory, since they are not generally used to increase memory bandwidth and performance.
DDR2 is a different animal when it comes to ratios. As DDR2 matured, memory makers and processor manufacturers have sanctioned higher and higher speed grades of DDR2 that are achieved with memory ratios. Users select and use higher rated DDR2 products to increase memory performance. On the Intel platform there was theoretically a price to pay since the memory controller was on the chipset.. This meant 1:1 DDR2 memory is theoretically top performance without compromise, while other speeds had some degrading from maximum performance due to the overhead of the FSB ratios.
AMD uses HyperTransport, which means there is no FSB. All memory speeds should theoretically perform as if they were 1:1 on HT. This simply means additional DDR2 speeds should carry no penalty at all. A user can choose DDR2-400, DDR2-533, DDR2-667, DDR2-800, DDR2-1066, and other DDR2 speeds based on performance needs and their budget, as there is no one speed which is best in a HyperTransport design.
For all of these reasons, performance of AM2 was benchmarked at DDR2-400, DDR2-533, DDR2-667, and DDR2-800. Due to limitations of current AM2 motherboards in that no timings faster than 3 are available for memory, all 4 speeds were evaluated at 3-3-3 memory timings. This is a fast memory timing for DDR2 at any of those speeds, as the fastest DDR timings we have tested at lower DDR2 speeds is 3-2-2. At DDR2-800, 3-3-3 are the fastest timings ever tested at this speed.
In addition, performance at DDR2-800 3-3-3, the fastest stock speed we could set on current AM2 boards with this memory, was compared to DDR performance at DDR400 2-2-2, which is generally the fastest stock speed on current Socket 939 motherboards. The performance difference is reported in %.
Benchmark results were very interesting. From a broad perspective, memory bandwidth and latency of DDR400 was matched by DDR2 at a speed of just over DDR2-533. Put another way, DDR2-533 provides equivalent bandwidth and latency to DDR400 on this 4th spin of AM2. DDR2-667 provided a bit better performance, and DDR2-800 3-3-3 provided a 12% improvement (lower number) in Latency over DDR400 and a 13.3% to 28.6% improvement in bandwidth depending on the memory benchmark used. It is interesting that memory writes enjoyed the biggest bandwidth boost.
Unfortunately, the added memory bandwidth did not translate into the kinds of increases in real-world gaming performance that many might expect. Gaming Performance at DDR2-800 3-3-3 at the same CPU speed increased 1% to 6.7% in our tests - with all except Call of Duty 2 in the 1% to 4% range. Call of Duty 1.2 was used in these tests, which supports dual processors, and test results did not follow the same expected scaling of other benchmarks. For this reason, the Call of Duty 2 results are considered suspect until we have more experience with Revision 1.2.
Since increased memory bandwidth did not translate into similar increases in gaming performance we can only conclude that this iteration of AM2 is not particularly memory starved - a result that was really expected. This does open the door for future upgrades and revisions of AM2 that might make better use of the available memory bandwidth with fast DDR2 memory.
DDR has a somewhat narrow operating range. Mainstream DDR might extend from DDR400 to perhaps DDR433. The very best, and most expensive, DDR might extend from DDR400 to a bit over DDR600. The useful range of fast DDR is even less because as the timings increase for higher speed the bandwidth starts to fall. Therefore top bandwidth with many of the top DDR products peaks somewhere in the DDR500 to DDR550 range and bandwidth normally drops as we move higher in frequency at poorer timings.
While ratios can be used with lower performing memory, the main purpose of ratios is to offset the performance limitations of cheaper memory so the user can enjoy the increased performance of overclocked processor without having to similarly invest in expensive memory capable of 1:1 memory overclocking. The point is that we will not be investigating memory ratios on DDR memory, since they are not generally used to increase memory bandwidth and performance.
DDR2 is a different animal when it comes to ratios. As DDR2 matured, memory makers and processor manufacturers have sanctioned higher and higher speed grades of DDR2 that are achieved with memory ratios. Users select and use higher rated DDR2 products to increase memory performance. On the Intel platform there was theoretically a price to pay since the memory controller was on the chipset.. This meant 1:1 DDR2 memory is theoretically top performance without compromise, while other speeds had some degrading from maximum performance due to the overhead of the FSB ratios.
AMD uses HyperTransport, which means there is no FSB. All memory speeds should theoretically perform as if they were 1:1 on HT. This simply means additional DDR2 speeds should carry no penalty at all. A user can choose DDR2-400, DDR2-533, DDR2-667, DDR2-800, DDR2-1066, and other DDR2 speeds based on performance needs and their budget, as there is no one speed which is best in a HyperTransport design.
For all of these reasons, performance of AM2 was benchmarked at DDR2-400, DDR2-533, DDR2-667, and DDR2-800. Due to limitations of current AM2 motherboards in that no timings faster than 3 are available for memory, all 4 speeds were evaluated at 3-3-3 memory timings. This is a fast memory timing for DDR2 at any of those speeds, as the fastest DDR timings we have tested at lower DDR2 speeds is 3-2-2. At DDR2-800, 3-3-3 are the fastest timings ever tested at this speed.
In addition, performance at DDR2-800 3-3-3, the fastest stock speed we could set on current AM2 boards with this memory, was compared to DDR performance at DDR400 2-2-2, which is generally the fastest stock speed on current Socket 939 motherboards. The performance difference is reported in %.
Benchmark results were very interesting. From a broad perspective, memory bandwidth and latency of DDR400 was matched by DDR2 at a speed of just over DDR2-533. Put another way, DDR2-533 provides equivalent bandwidth and latency to DDR400 on this 4th spin of AM2. DDR2-667 provided a bit better performance, and DDR2-800 3-3-3 provided a 12% improvement (lower number) in Latency over DDR400 and a 13.3% to 28.6% improvement in bandwidth depending on the memory benchmark used. It is interesting that memory writes enjoyed the biggest bandwidth boost.
Unfortunately, the added memory bandwidth did not translate into the kinds of increases in real-world gaming performance that many might expect. Gaming Performance at DDR2-800 3-3-3 at the same CPU speed increased 1% to 6.7% in our tests - with all except Call of Duty 2 in the 1% to 4% range. Call of Duty 1.2 was used in these tests, which supports dual processors, and test results did not follow the same expected scaling of other benchmarks. For this reason, the Call of Duty 2 results are considered suspect until we have more experience with Revision 1.2.
Since increased memory bandwidth did not translate into similar increases in gaming performance we can only conclude that this iteration of AM2 is not particularly memory starved - a result that was really expected. This does open the door for future upgrades and revisions of AM2 that might make better use of the available memory bandwidth with fast DDR2 memory.
Facebook
Twitter
RSS
37 Comments
View All Comments
phial - Thursday, July 5, 2007 - link
"Comments are often made that any evaluation of AM2 should include overclocking. Those theorists argue that the faster processor clock will move AM2 to the higher "bus" speeds that make for better efficiencies of the DDR2 memory controller - or something along those lines."OK who hired this guy? He knows NOT what he speaks of. Its apparent through the entire review.
phial - Wednesday, July 4, 2007 - link
"This will likely shift to patterns similar to those seen in bandwidth positioning as the AM2 memory controller is further refined and game patches make better use of AM2 capabilities."Wow... and your posting hardware reviews on a high traffic website.. amazing...
Zebo - Tuesday, April 18, 2006 - link
I get everest of 42ns 2-2-2 @ 200 with FX60 speeds on DFI NF4. I get 36ns when cranking up to 250mhz 2-2-2. And that reminds me - I guess you did'nt want to embarres AMD's new chip too much by putting old, supported, DDR 550 @ 2-2-2 in there huh?What a dog this change is.
Wesley Fink - Wednesday, April 19, 2006 - link
Actually our Beta motherboard and CPU were not 100% stable at 275. Also I don't have any memory that will do DDR550 (275 clock speed) at 2-2-2. Perhaps you mean another speed or other timings.Our Crucial 2GB kit can do DDR500 at 3-2-3 timings, but our DDR2-800 memory is also capable of doing 3-2-3 at some speeds if the beta motherboard had supported those timings. That's the reason we compared 3-3-3 500, which should be very very close to each memory's top performance and a fair comparison.
Our latency testing was at a slower speed than your FX60, which is why you got lower latency numbers. We did, however, set both the DDR and DDR2 platforms to the exact same speed and ratios for all the benchmarking. The results comparison was as fair as we could make it.
spinportal - Tuesday, April 18, 2006 - link
Let's see.. the 1GB PC-4400 DDR-550 kit is roughly 200$US for 2.5-3-3 timing. The DDR2-667 PC-5300 1GB kit is ~180$US for 3-3-3 timing for use to OC to DDR2-833. Unless I'm missing something, I don't understand your beef about spending $20 more for equal performance and using more juice.puffpio - Monday, April 17, 2006 - link
I can see gaming benchmarks being VERY graphics and cpu calculation intensive...but in terms of the bus between the cpu and main memory..i dont see it as that intentive
I wonder if a multitasking benchmark w/ lots of datasets would show off the increased bandwidth? (giant spreadsheets, databases, image processing, video encoding, etc etc)
shady28 - Tuesday, April 18, 2006 - link
"Real world" single user and light multitasking apps aren't going to use up the bandwidth. A heavy duty multitasking environment is what's needed to really test the benefits of dual core + DDR2. The thing is, this doesn't really describe what most PCs spend 99% of their time doing.
That said, Conroe does look promising, and I'm interested to see what AMD can come up with to counter it. The initial performance advantages, according to *Intel* benchmarks, are pretty significant. Hopefully AM2 is just laying the groundwork for faster processors from AMD. Still, I can't help but think the performance boosts we've been getting since about 2001 (3.06Ghz P4 HT) are marginal, and that this is yet another marginal boost.
NullSubroutine - Monday, April 17, 2006 - link
I am still convinced that the transition from DDR to DDR2 is primiarily because it unifies the memory production industry, lowers power for server and mobile areans, increases memory bandwidth for upcoming memory hog (ie vista), and servers as a platform of increased memory capability. As for the last point I think most of us believe that quad core CPU's may use more memory bandwidth and the original DDR may act as a bottleneck for the system. I also think it is clear that memory bandwidth is not a bottleneck for AMD, but I applaud them taking care of a potential bottleneck before it exists, unlike Intel who has continuelly bottlenecked their CPUs by insuffient memory bandwidth.NullSubroutine - Monday, April 17, 2006 - link
should say serves obviously.
eRacer - Sunday, April 16, 2006 - link
Was the DDR2 using a 2T command rate and was the DDR-400 tested using 1T? Is there an option to change the command rate on the AM2 motherboards? I believe at least some Pentium DDR2 motherboards with NVIDIA chipsets have the option to use a 1T command rate.I've seen some Sandra 2005 standard memory tests gain 500MB/s or more just by using 1T instead of 2T. A gain like that would allow DDR2-533 1T to score higher than DDR2-800 2T in the Sandra 2005 standard memory test. If DDR2-800 could use a 1T command rate the performance gain in the non-synthetic benchmarks might be a few percent higher on average. While that doesn't sound like much it would be roughly double the DDR2-800 performance advantage over DDR-400 in this review.