PQI & G. Skill: New Choices in 2-2-2 Memory
by Wesley Fink on October 9, 2004 11:35 AM EST- Posted in
- Memory
Test Results: PQI 3200 Turbo
To be considered stable for test purposes, Quake3 benchmark, UT2003 Demo, Super PI, Aquamark 3, and RTCW had to complete without incident. Any of these, and in particular Super PI and Return to Castle Wolfenstein, will crash a less-than stable memory configuration.Intel Test Results
PQI 3200 Turbo (DDR400) - 2 x 512Mb Double-Bank | ||||||
Speed | Memory Timings & Voltage |
Quake3 fps |
Sandra UNBuffered | Sandra Standard Buffered |
Super PI 2M places (time in sec) |
Wolfenstein - Radar - Enemy Territory fps |
400DDR800FSB | 2-2-2-5 2.5V |
337.5 | INT 2845 FLT 2935 |
INT 4523 FLT 4515 |
129 | 71.7 |
433DDR866FSB | 2-2-2-5 2.75V |
364.8 | INT 3172 FLT 3205 |
INT 4907 FLT 4902 |
120 | 77.6 |
466DDR933FSB | 2-3-3-5 2.75V |
381.5 | INT 3236 FLT 3253 |
INT 5167 FLT 5165 |
113 | 81.5 |
500DDR1000FSB | 2.5-3-3-5 2.75V |
405.8 | INT 3362 FLT 3395 |
INT 5557 FLT 5562 |
106 | 87.3 |
533DDR1066FSB | 2.4-3-4-5 2.75V |
423.0 | INT 3599 FLT 3563 |
INT 5944 FLT 5903 |
100 | 91.1 |
556DR1122FSB | 3-3-4-6 2.75V |
443.0 | INT 3643 FLT 3699 |
INT 6188 FLT 6183 |
96 | 95.3 |
PQI reached DDR556 on the Intel platform, which places it among the best DDR400 performers we have tested. The memory reached even higher overclocks on the AMD test bed.
AMD Results
PQI 3200 Turbo - 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 | 2-2-2-10 2.6V 1T |
512.9 | INT 2605 FLT 2796 |
INT 6091 FLT 6039 |
81 | 110.4 |
11x218 | 436 DDR | 2-3-2-10 2.7V 1T |
512.1 | INT 2725 FLT 2906 |
INT 6446 FLT 6368 |
81 | 110.1 |
10x240 | 480 DDR | 3-3-3-10 2.7V 1T |
513.0 | INT 2849 FLT 2977 |
INT 6627 FLT 6545 |
80 | 110.6 |
9x267 | 533 DDR | 3-4-3-10 2.75V 1T |
517.2 | INT 2930 FLT 3159 |
INT 6910 FLT 6821 |
80 | 111.5 |
8x298(2.38GHz) | Highest Mem Speed 596 DDR |
3-4-3-10 2.8V 2T |
514.7 | INT 2916 FLT 3099 |
INT 6765 FLT 6685 |
80 | 110.7 |
9x285(2.57GHz) | HIGHEST Performance 570 DDR |
3-4-3-10 2.75V 1T |
551.4 | INT 3172 FLT 3413 |
INT 7402 FLT 7282 |
74 | 119.1 |
The PQI 3200 Turbo is also based on the latest Samsung TCCD chips. The latest TCCD chips appear to perform much better on Athlon 64 than the first TCCD samples tested in =F-A-S-T= DDR Memory: 2-2-2 Roars on the Scene.
PQI reaches DDR596 at the highest memory speed, very close to DDR600. Highest performance was achieved at a Command Rate of 1T at DDR570.
10 Comments
View All Comments
adamofwales - Thursday, November 11, 2004 - link
I am considering purchasing a matched pair 1024MB total, of the PQI Turbo 2700, 2-2-2-5 timings, (PQI2700-1024DAL) and I was wondering, do you think that it will overclock as well as the 3200 with the same timings? I read somewhere that the PQI 2700 Turbo 2-2-2-5 512x2 will run at 2-3-2-5 at 3200 speeds.What do you think?
Wesley Fink - Saturday, October 16, 2004 - link
#3 - We have NEVER implied you need an FX53 to review memory. We have , however, stated the need for a standardized memory test bed and the FX53 is the CPU we have chosen. The trends over spped would apply to any Athlon CPU since they are all unlocked below the stock speed.Others - We are planning a Value RAM roundup in the near future - after the huge number of new equipment launches for the rest of October. Since every memory vendor now has a Samsung TCCD memory it should be clear that TCCD is now at the top in almost everyone's mind. Samsung TCCD chips are also expensive, which is why we have reviewed alternative brands based on those chips.
MadAd - Thursday, October 14, 2004 - link
yup, i have to agreeIts difficult to complain at the tremendous quality of memory reviews here at AT but I too believe it would be useful to have a catchup on how the other half of the memory market is shaping up.
If it was a case of 'this months exculsive is next months mainstream is a 6 months time bargain' like gpu/cpu/etc then it wouldnt matter so much, but its not, leaving a gap in the product review landscape.
Infact, what is value ram at all these days? Lower speed binned chips from a recent stepping silicon (like gpu) or seperatly RND'ed low cost engineering or even lower purity processes?
You see, theres an article in the making already :)
CalvinHobbes - Tuesday, October 12, 2004 - link
I'd love to see a comparison of cheaper memory as well. I'm in the market for some new ram and I just want to know if I can spend $170 for 1GB or is it really worth while to spend the $245+ for the 2-2-2 stuff.Zebo - Monday, October 11, 2004 - link
Sup Concillian;):PIt's almost like AT only reviewing FX's and EE's on the processor side.
I really feel AT is doing a diservice to the community by continually pimping this overpriced RAM in every review. Even for overclcokers this holds true, since much budget ram scales the same as the boutique ram when pushed..albeit with mybe a little looser timings and a little slower.
But sure as heck ain't 100% slower to justify boutiques ram 100% price premium. Especially when most users are on fixed budget and thier money is better spent on a better video card, more HD space or something else.
But comming to AT, as a builder, you'd think this overpriced RAM is your only choice since that's all they like present and are getting your budget jammed on the front end for almost nothing in return.:(
Zebo - Monday, October 11, 2004 - link
Would you guys *PLEASE* test some budget ram like crucial 8T to show what a ripoff this boutique stuff is price/performance wise?Concillian - Monday, October 11, 2004 - link
It is very interesting to see the very high performance results of the best memory out there. However, I feel it would be useful to compare this to some of the common forms of value memory.As a consumer about to go spend hard earned dollars on a new motherboard/CPU/RAM, the question I ask myself is:
Is it worth it to spend the bucks on super fast memory, or do I spend about HALF and get decent PC3200 CAS 2.5 value memory from the likes of OCZ, Mushkin, or Corsair and use a memory divider when overclocking an A64.
In reality, the typical memory showcased here on Anandtech is very expensive, roughly twice the price of typical value memory.
When you can get an A64 2800+ and motherboard for around $200, I can't be the only one questioning whether $250-$300 just on a gig of memory to overclock a $200 mobo/CPU combo, when closer to $150 may work almost as well. I can't help but wonder if the extra ~$150 (or a nearly 40% increase in cost of the total package mobo + CPU + 1gig RAM in this case) is really worth the system performance.
Uff - Monday, October 11, 2004 - link
This is the second memory test that claims that you need an FX-53 to test memory speed because it's completely multiplier unlocked. I don't see you going above the multiplier 12 anywhere in these tests, thus you could do the exact same thing with 3400+ (2.4GHz version), 3700+ or 3800+, as all the AMD CPUs are multiplier unlocked downwards.Secondly, do you have any further information on the Corsair 2-2-2 sticks? My own tests have shown they can barely run at 3-4-4-10@218MHz fsb and fail to reach 240MHz at any timings :(
Wesley Fink - Sunday, October 10, 2004 - link
The timings used at each speed are included in the Test Results tables on pages 6 and 7.There is only so much information you can include in a chart before it gets too confusing, but we always include timings and voltages for each speed in the Test Results tables.
AkumaX - Sunday, October 10, 2004 - link
nice article! were all different memory speeds (actual: 200, 233, 250, 275) at 2-2-2-5|10 also?