Date: September 13, 2006
Author(s): Rob Williams
You want fast memory? Kingston has fast memory! Though not usually touted as the company to offer killer gaming memory, the HyperX PC2-8000 kit proves to be one of the fastest kits we’ve tested!
Since the beginning of Techgages existence, we’ve taken a look at over 30 kits of memory, but this is the first time we’ve had some Kingston on the workbench. As mentioned in the intro, Kingston is not a name that’s usually thrown around all over in the enthusiast overclocking side of things, but I am not sure why after taking a look at these sticks.
The more common “top end” memory speed now is PC2-8000, and that’s what we are taking a look at today. There is PC2-8500 on the market also, but they haven’t proved to be that popular yet. Especially due to the fact that many PC2-8000 kits will easily overclock to those speeds with no effort.
Kingston usually pairs two modules together in a single package, but for whatever reason, I received two individual packages. It’s a small plastic blister pack that requires you to cut through the sticker in order to open it up. Nothing major here… but it does a good job of keeping everything safe. Underneath each of the modules are some simple installation instructions.
Kingston uses blue heatspreaders on all of their modules, which help them stand out. I can’t recall off the top of my head another company that uses this shade of blue on their spreaders..
With a close-up of the sticker, we can see a lot of information… moreso than most other companies will put. We have the model number, production codes, voltage and the density. We also find out that these modules are Assy in the USA. Yeah!
That’s all we can possibly cover about the physical aspects of the modules. Nothing amazing here, but nothing lacking either. No complaints!
These modules are from their HyperX line which tells us their are designed for gamers and overclockers. Of course, PC2-8000 speeds are nothing you would likely throw into your moms PC, so we can expect some great performance here. They are equipped with 5-5-5-15 timings which are not tight, but not surprising either. The majority of DDR2-1000 modules include these timings, although most that I have personally tested will do 4-4-4 reliably with a pinch of extra voltage.
I am not positive of the chips uses in these modules, but it would not be a far stretch to assume that they are Micron D9GMH. Most new enthusiast modules use either GMH or GKX, but considering GKX is much more expensive, it’s shunned. GMH proves to be a huge hit though, and it’s evident since it’s been used in kits ranging from PC2-6400 all the way up to PC2-9000.
Speaking of OC’ing success… here are all of the tested settings I used:
As you can probably tell, I don’t touch tRC until later on in testing.. usually only the final overclock. It doesn’t usually make a huge difference in performance regardless, but for a top overclock it’s nice seeing the smallest possible numbers. Though some DDR2 kits have a rough time with low tRAS at high frequencies, that was not the case here. We had a smooth four of a kind here, although it required some juicy 2.65vdimm.
You may ask if it’s safe to even use that much voltage to begin with. Yes, and no. It depends on the chips really, but these ones had no problem whatsoever. I always keep a 120mm fan pointed straight at the modules, so they did not get as hot as they would otherwise. In fact, I didn’t find it to be an issue at all, and I wouldn’t have a problem pumping even more voltage into them. However, that’s getting a tad ridiculous since I would be focusing on tighter timings as opposed to higher frequencies. It’s just not worth it to pump that much voltage into the modules for minor gains that tighter timings bring.
That said, DDR2-1100 4-4-4 with 2.65v was 100% stable throughout all of the tests, and this includes 3D stable. I played Half-Life 2 for an hour without a single hitch… so obviously these modules have some major overclocking appeal.
Throughout all of our benchmarks regardless of what we are reviewing, testing is done in a clean and stand-alone version of Windows XP Professional with SP2. Prior to testing, these conditions are met:
If you are interested in using the same benchmarks as us, feel free to visit the developers website:
The testing rig used for today’s benchmarking is as follows:
I will be pairing the Kingston against the recently reviewed OCZ 7200 EPP, which has proved to be one of my favorite kits this year so far.
First up is one of my personal favorites… Sandra. As expected, we have very tight scores here… the OCZ won some, the Kingston the rest. We have quite nice top scores though… came close to 11K but no cigar. Perhaps at 2.8GHz it could have happened.
Before taking our unbuffered results too seriously, please take a look at our precise settings. They differ from the way most perform unbuffered, but this takes a lot away from the CPU and relies solely on the memory.
Unlike the buffered tests, we can really see some differences here. Surprisingly enough, Kingstons kit beat out the 7200 EPP in each test! Impressive, but the Kingston does retail for more, so it’s nice to see an actual increase somewhere.
Everest scores are not some I take to heart, but like Sandra, they are good to get a grasp of how well your overclocks scale. Here, the first few sets of results vary quite a bit regarding the Read, although the Write doesn’t move too much. The latter 3 sets are all painstakingly close, however.
The same goes for the latency. I cannot explain why the first three sets varied so much, but the final three were identical!
Are you craving a graph crammed with 36 results? Well here she be!
Only PC Mark had rather similar results between both kits, though the top 3D Mark 01 scores came close to one another. Sciencemark varies a lot from kit to kit, even with the same speeds. That’s evidenced in the results there. I’ve never broke through 10K in Sciencemark, but we came very close this time.
Yet again, the first few sets vary more than the last few. This tells me that GMH, at higher frequencies will not sway much from one another… at all. There’s no other way I can wrap my head around it, as all settings were identical between tests.
Ahh boy. If you want an expensive memory kit, there are many to choose from. Though I don’t have an SRP or availability information, you can bet it will hover around the ~$400 mark. Possibly less if Kingston wants to have a good edge on the competition. So the real question… is this kit worth it?
There are a few different angles to this. Are you an overclocker, or not? If you are not, then it would make more sense to go with the OCZ 7200 EPP that we mentioned in the review, because it clocks just as high as this Kingston. However, the HyperX -may- have GKX chips.. I am not sure. If they did, then they would likely be able to clock higher than I could take them due to my CPU max overclock.
Regardless, these are some fantastic modules… I had a blast playing with them. If you are able to go beyond 2.5vdimm, then these have a lot of potential. DDR2-1100 with 4-4-4-4-13 timings was completely stable… I could run my rig forever on that. That’s rather impressive for a DDR2-1000 kit with 5-5-5-15 timings :-)
If you are not an overclocker and don’t even want to think about giving it a shot, then yes these modules were made for people like you. If you don’t care about spending a lot of money , then go for it. However, if you are a novice OC’er and want to finally jump in, it would be more economical to go for a Kingston HyperX PC2-6400 kit or something similar from a competitor. For what the memory delivers, I am awarding them a 9 out of 10.
If you have a comment you wish to make on this review, feel free to head on into our forums! There is no need to register in order to reply to such threads.
Copyright © 2005-2019 Techgage Networks Inc. - All Rights Reserved.