All that glitters isn't necessarily Gold. I recall a review criticizing OCZ's GOLD memory because the heat-spreaders weren't actually made of GOLD. There are certainly times to question advertising fodder and then there is the painfully obvious. For the record there's no Au (Atomic weight = 196.966) in OCZ GOLD memory. Perhaps the reviewer thought it was open season on OCZ since they'd had a somewhat rocky start when people accused them of not manufacturing their own IC's. If you’re reading this and your interested in OCZ you probably don't need a lesson on how the semi-conductor industry works. Briefly there are a limited number of IC Fabs, as well as a limited number of PCB makers for DDR/DDR2. This just one reason most memory makers out-source and this practice is actually beneficial to you as the consumer. Even if memory makers could afford to build multi-billion dollar Fabs the quality (speed) of the product would be always be determined by that Fab's yield. As much as the Fabrication element of the semi-conductor world is a science the results can and do vary widely.



Take for example Winbond BH5, these ICs were developed to meet JDEC DDR400 specifications at such and such CL-timings and at such and such voltage, the end result; however, produced a PCB/IC formula capable of DDR600 speed, CL2-2-2-5 latencies, accepting 3.4V. Seeking out such products made OCZ Technology a pioneer as they began to package and sell a product guaranteed to perform well beyond JEDEC specifications. Of course as overclocking became mainstream many Overclockers became complacent, expecting overclocks' beyond what they had purchased. There became a sense of entitlement as people purchased PC3500 then expected PC4000 performance based on what they read in forums. In the end, if you can please this crowd as long as OCZ Technology has, you’re doing something right.



With their PC2-6400 GOLD laid out on their tee-shirt you can begin to see the company has become an aggressive entity, but without the huge advertising budget of other companies OCZ has had to earn respect the old fashioned way, one satisfied (or dissatisfied) customer at a time. Their memory is now some of the most coveted out there including such products as their Platinum Series, high voltage solutions found in their VX Series, and an idea so simple it's brilliant such as their XTC heat-spreaders seen below.



What really surprised Overclockers, myself included, was OCZ's venture into the PSU market. They researched and found what is one of the best Power Supply designs on the market; Tagan. They licensed it and actually improved upon it, the end result was their Power Stream line. We were fortunate to get our paws on the very first sample in our OCZ-420ADJ Review and to fan my feathers I may have been the first to recognize Power Stream used Tagan innards.


Well enough of my semi-sycophant, OCZ apple polishing, let's get down to brass tacks. It's always nerve racking when you may be sacrificing a perfectly good piece of hardware by tearing it open. At least I've learned from a truly utterly stupid experience to test the memory first, and then tear it open. I must confess I'd done this in reverse once which made for an uncomfortable letter to the manufacturer explaining how I'd rendered a $300 kit of memory useless to show my readers something they may not want them or the competition to see anyway. With the advent of the heat-spreader (in some cases "IC Comforter") identifying IC's has required removing heat-spreaders. To be honest one would think as far as we've come with thermodynamics memory IC's would either be thermal-epoxied, or at least use thermal paste with better clamps for the heat-spreaders. A few manufacturers do use an epoxy for example Corsair. Below my painstaking effort to remove the heat spreader only revealed blanks?



A brief word about UTT (UnTesTed) or ETT (Effective TesTed) DRAM chips. These are ICs which may not have passed specific tests at the wafer Fab, or at various sub-contractors wafer Fabs often employed to test their ICs. It’s often the case Fabs can save money forgoing costly test stages for these chips, which can then be sold on the market for various purposes. Many of these chips will end up in non-PC specific products such as DVD, or MP3 players or various other electronics. When left to the various manufacturers to test such as a “Fabless” memory manufacturer, these chips can be had for much less then other chips which may meet certain requirements but may still be inferior in performance to the UTT chips. While this scenario would be the exception given the large numbers of blanks which hit the market the odds are still in favor of the memory maker. With motherboard designed to accommodate these chips such as DFI with its extensive BIOS memory settings and exceptionally high voltage options these IC can offer superior performance to memory costing much more. Recently UTT chips based on the Windbond BH5 die flooded the market offering outstanding performance with the right motherboard. There are many legitimate uses for throughout the industry and while there are those batches which are inferior, the opposite is also true.