Kingston KVR400X64C25/512

Kingston is the world's largest independent Memory manufacturer with annual sales over $2.4 Billion. Kingston is a very well-known memory brand, and it can be found at almost any large computer dealer. Kingston has a very complete line of memory products from the cheapest Value RAM to their enthusiast HyperX products. Kingston is also a brand that you will often see on sale at the big computer chains.

Kingston KVR400X64C25/512 was supplied as two individual DIMMs, rather than a 1GB kit. The RAM is rated at CAS 2.5 at DDR400.

Heatspreaders are not as often seen in Value products, since manufacturers are usually doing whatever is necessary to bring in the lowest price. This means eliminating extras like heatspreaders. Memory chips on the CAS 2.5 are blanks that are labeled with the Kingston logo.

Specifications

The KVR400X64C25/512 is rated at CAS 2.5. Kingston supplied two Value RAM samples for testing - this CAS 2.5 and a lower cost Value RAM rated at CAS 3.

 Kingston KVR400X64C25/512 (DDR400) Memory Specifications
Number of DIMMs & Banks 2 DS
DIMM Size
Total Memory
512 MB
1GB
Rated Timings 2.5-3-3 at DDR400
SPD (Auto) Timings 2.5-3-3-8
Rated Voltage 2.6V

Test Results

Kingston KVR400X64C25/512 (DDR400) - 2 x 512Mb Double-Bank
CPU Ratio at 2.4GHz Memory
Speed
Memory Timings
& Voltage
Quake3
fps
Sandra UNBuffered Sandra Standard
Buffered
Super PI 2M places
(time in sec)
Wolfenstein - Radar - Enemy Territory fps
12x200 400 DDR 2.5-3-2-6
2.6V 1T
533.0 INT 2470
FLT 2642
INT 6016
FLT 5987
83 112.0
12x204
(2.45GHz)
Highest CPU/Mem Performance 2.5-3-3-6
2.7V 1T
533.2 INT 2636
FLT 2797
INT 6143
FLT 6112
81 112.4
To be considered stable for test purposes, Quake3 benchmark, UT2003 Demo, Super PI, Aquamark 3, and Comanche 4 had to complete without incident. Any of these, and in particular Super PI, will crash a less-than stable memory configuration.

We haven't made a mistake here. As you can see, there is really no overclocking headroom at all with this Kingston 2.5 memory. While it actually outperforms specs with 2.5-3-2-6- timings at DDR400 at stock 2.6V, we could not complete our memory benchmarks above a 204 CPU setting. Super Pi and Unbuffered Memory tests worked fine to almost DDR480, but at any setting above 204 (DDR408), games would freeze or run in spurts and freezes. This points to a likely conflict with the AMD on-chip memory controller or the DFI nF4 motherboard, and in fact, the Kingston 2.5 may reach much high memory speeds on an Intel platform.

We have found Aida 32 to be very useful in examining read/write performance and memory latency. Aida 32 is now available as Everest Home Edition, a free download from www.lavalys.com.

Kingston KVR400X64C25/512 (DDR400) 2x512Mb Double-Bank
Everest 1.51
CPU Ratio at 2.4GHz Memory Speed Memory Timings
& Voltage
Everest READ
MB/s
Everest WRITE
MB/s
Everest Latency
ns
12x200 400 DDR 2.5-3-2-6
2.6V 1T
5800 1918 47.3
12x204
(2.45GHz)
Highest CPU/Mem Performance 2.5-3-3-6
2.6V 1T
5855 1991 46.9

The Memory Speed range was so limited with the Kingston CAS 2.5 that there is really little on which to comment. However, despite the problems above 204 on the Athlon 64, at stock DDR400, the Kingston KVR400X64C25/512 performed without any problems.

Performance Test Configuration Kingston KVR400X64C3AK2/1G
Comments Locked

102 Comments

View All Comments

  • CanadianDoc - Sunday, April 17, 2005 - link

    1 GB (2 x 512 MB) of Crucial Ballistix PC 3200 now lists for $192 U.S. at crucial.com.

    On any mobo, it runs at 200 MHz (DDR 400) at 2-2-2-6 timings at 2.8 V, outperforming any other RAM in this review.

    As shown by AnandTech (www.anandtech.com/memory/showdoc.aspx?i=2386), on a DFI nF4 mobo, it overclocks to 280 MHz (DDR 560) at 2.5-3-3-6 timings at only 2.9 V, where it closely matches the top performance of any RAM available at any price, bar none. IMHO, this is real value.

    Adding a Venice 3200+ overclocked to 10 x 280 MHz = 2.80 GHz (www.xbitlabs.com/articles/cpu/display/athlon64-venice.html), a Thermalright XP-90 heatsink with a SilenX 92mm 14dBA fan, a Seagate 7200.8 SATA NCQ hard drive, and a Gigabyte X800 XL video card with SilentPipe cooling (GV-RX80L256V), in an Antec Performance One (P160) case (www.anandtech.com/casecooling/showdoc.aspx?i=2346) with an XG Magnum 500W heatpipe PSU (available later this month, according to www.xgbox.com), you have the makings of an ultra-quiet gaming rig with near state-of-the-art performance at a great bang-for-the-buck price.
  • ozzimark - Thursday, April 14, 2005 - link

    you can. i'm a big advocate of crucial ballistix

    speaking of that company, why were they missing from the testing too?
  • BaronVonAwesome - Thursday, April 14, 2005 - link

    195 seems a bit steep for value RAM. I'd like to point out that you can get Ballistix from Crucial for less than 210 I believe.
  • Den - Thursday, April 14, 2005 - link

    Thanks for pointing out that the part numbers are on page two. As many others have said, it would have been really nice to see the best the different rams would do at 2.9v in addition to what they would do with DFI voltage. Also it would have been nice to see a greater range of memory tested but I understand you are limited by not being able to afford to buy and test what you want. Beggars can't be choosy. Thanks for doing this though, the general concept was good and there was some interesting information.
  • NXIL - Thursday, April 14, 2005 - link

    Paying $100 to buy some Corsair Valueram (Newegg--had an $88 special yesterday) would have been the fair and reasonable course of action.

    Consumer Reports has been testing products fairly for decades--they don't accept advertising, and, they buy the cars, electronics, and other items they test anonymously. Of note is that "Sharper Image" has sued them recently (SI lost) when they tested one of their bogus "air purifiers"--seems to make the air less clean, actually, by releasing ozone.

    Anandtech should get some Corsair RAM and test it--and, they should buy samples of the other brands tested, and compare the "off the street" samples with the ones provided by the manufacturers. Unless I am mistaken, video card makers were found to be rigging their drivers to test better with certain benchmarks. It would not be too surprising to find that memory manufacturers had taken some of their special high cost "binned" chips and sent them out with a "value" label on them.

  • MadAd - Wednesday, April 13, 2005 - link

    Oh dear ... I left this review for 2 days till I had time to read it properly, I'm sorry to say I wish I had not bothered.

    1) Should have been titled value ram from non value companies. Wheres the real value ram?

    Since the price began to drop we seem to be up to our ears in stuff ive never heard of purporting to be cas3@3200 - Stupid me for thinking that thats what a value roundup should include, noname oem kit, not some hand selected bunch of good-but-value-priced ram from the majors.

    2) Why is the following question being avoided and ignored? Namely why wasnt any corsair memory got from another source and included in the test?

    A reader posted some assumptions however further to that it could be i) they use pretty much the same chips all over and dont want to give the game away that the extra $100 doesnt get you much increase and ii) anandtech dont want to bite the hand that feeds so bowed down to the request rather than doing whats right and finding the truth come hell or high water.

    Im not usually this negative but I do feel quite let down, sorry.
  • classy - Wednesday, April 13, 2005 - link

    I think a lot of folks are missing it. I think some people need to look at this way. You would have upgrade your cpu 1-2 speed grades to equal the performance increase that using the VX or BH5 memory would bring. Even at stock speeds. Now yes you need more voltage, but DFI I believe produces the best A64 borads as well as the best athlon XP board, Ultra B, which all are capable of supplying the necessary voltage. You can also use the ddr booster on some boards. Its a great alternative to increase system performance without spending a lot of money. And this is memory that you can grow with for at least a couple of years. I believe AMD said it won't go DDR2 until 2007. Great article, good stuff, and nice to see other ways to bring a performance increase. Nice job Wesley
  • unclebud - Wednesday, April 13, 2005 - link

    84 comments in 2 days!!! haven't seen that in a while...
    thanks for the article! good reading as usual
  • XRaider - Wednesday, April 13, 2005 - link

    The cheap OCZ is faster then the Plat. Rev2 !!! Ain't dat a Bitch!!! >:-(
    WTF is up with OCZ. shooot.
  • srstudios - Wednesday, April 13, 2005 - link

    Wesley, I hate to be a nag, but now you have both the OCZ PC3200 Gold, and the OCZ PC3200 Premier with the same model number. PC3200 Gold P/NOCZ4001024ELDCGE-K
    is correct.

    The Premier should be PN- OCZ4001024PDC-K, at least I think that's the correct part. http://www.ocztechnology.com/products/memory/ocz_d...

    I have a gig of the 'old' BH-5, Mushkin Black L2 PC3500, do you think they would play nice with this new value BH-5? I don't think I would combine them, just wonder if you guys have any thoughts about it.

Log in

Don't have an account? Sign up now