Navigate: Advice Beginners BIOS Guide CPUs Links Mainboards Memory Network Storage Video/Sound Cards Contact Forum SiteMap Sponsors WebNews Home	.	.
CPU Intel P4 840 D P4 820 D P4 630 P4 640 P4 650 P4 660 P4 670 AMD Athlon64 3500+ 3700+ 3800+ 4000+ X2-3800+ X2-4200+ X2-4400+ X2-4600+ X2-4800+ 1-Way Opteron Opteron 144 Opteron 146 Opteron 148 Opteron 150 Opteron 152 2-Way Opteron Opteron 240 Opteron 242 Opteron 244 Opteron 246 Opteron 248 Opteron 250 Opteron 252 2-Way Dual Core Opteron Opteron 270 Opteron 275 nVidia GF 7800GT GF 6800GT GF 6600GT ATI R X850 XT PE R X850 XT R X800 XT PE R X800 XT R X800 XL Memory Corsair Crucial Kingston Mushkin OCZ What are you shopping for?

Summary

The first big wave of DDR has come and gone, most manufacturers have learned the tricks of the trade and the performance of new chip revisions is skyrocketing. Additional players are entering the DDR scene, starting with lesser know companies like Nanya and including the SDRAM Giants Infineon as well as some smaller, high performance oriented companies like Mosel-Vitelic. With memory prices at an all time low, the additional required investment for DDR memory is no longer an obstacle in the move towards the DDR platform but there are still quite a few questions that need to be answered. Amongst the issues at hand are the different brands, that is, what is hot, and what isn't. In addition, working with DDR requires re-thinking some of the BIOS setting strategies, particularly with respect to stability, performance and overclocking. .

Since my employment is related to Mushkin, I have lately been accused of being biased towards Mushkin products. This concern may be valid to some, however, personally I don't really care about it. On the other hand, for the sake of peace, the easiest solution for a round-up like this is to leave Mushkin products out for the time being and post a separate review in the near future. In other words, this one here is a Mushkin-less DDR roundup, DUH.

Lets Move On

Despite some rumors and forecasts predicting the end of the PC world and DDR in particular in favor of some Rambus powered handheld devices, there is no doubt that the PC platform will survive the next few years and fully embrace DDR technology. In this case, the term PC encompasses both desktop and, most importantly, laptop computers which currently are showing the highest growth rate of all computer devices. Speaking of laptops, one of the reasons why the industry wholeheartedly embraces the DDR format as opposed to the standard SDRAM are that there is basically no price overhead, and, further, DDR is operating at 2.5V. The voltage delta compared to the 3.3V at which SDRAM is running can account for as much as 40% power saving. This is important since energy conserving measures are becoming one of the critical design factors in personal computing.

If one branch of the PC market (laptop) moves over to DDR, there is little incentive to maintain the other branch using a somewhat outdated (SDRAM) technology with lower performance and higher power demands, therefore, it is almost a push-pull scenario where, on opposite ends of the scale, the laptop and desktop formats mutually endorse each other's move towards DDR.

The beginning of the DDR era was somewhat hampered by all different kinds of glitches stemming from redesigned SDRAM chips based on over-aged die and it took awhile to get things under control. Only six months ago, everybody was getting excited about running PC2100 and possibly even at latency settings of 2:2:2 which still appeared to be reserved for a few of the lucky ones who got their hands on a Golden DIMM.

R&D in the memory industry hasn't stopped, though. Die technology moved from .25 to .18 µm and is heading towards .14 µm with internal chip voltage at 1.8 V. Needless to say that each subsequent revision showed further improvement in terms of performance.

What was lagging behind, though, was the mainboard and chipset technology that would provide a platform to really push the newer DDR DIMMs and show where the limits are. Keep in mind that PC2700 or DDR 1.5 as intermediate step towards DDRII has been on the roadmap for Q4 2001 or Q1 2002, postulating 333MHz data rate equivalent to a memory bus operating at 166 MHz. In view of the tight timings required by DDR, 166 MHz clock rate appeared to be somewhat difficult to achieve, leading to the proposal to move from the current TSOP (tiny small outline packaging) to a TBGA (tiny ball grid array) factor for the purpose of reducing trace length and pin impedance. However, because of the clock forwarding or source-synchronous timing of the DDR bus using a strobe signal, running at 166 MHz clock rate appears to be actually easier than in an SDRAM scenario.

next page:    => the Data Input / Output Strobe (DQS) and what it does for performance =>