PCSTATS tests
DDR memory on two major CPU platforms - the AMD Athlon64 and the Intel Pentium 4
- because RAM behaves differently with each processor architecture. For
instance, Intel users tend to run their memory at high speeds with high latency
settings, and in general that works quite well for Intel-based computers. It
would be detrimental to system performance if you were to run the same DDR
memory settings on an AMD Athlon64 system, since latency plays a much larger
role in system performance.
Because much of our audience is made up of enthusiasts, PCSTATS also runs
overclocking tests so you can get an idea on how far the memory will
go.
On both Intel
and AMD test PCs, we're only interested in seeing how high we can go with the
memory running 1:1, as running with other dividers puts the overclocking
bottleneck elsewhere, and not with the system memory. On the Pentium 4 test
system we sometimes use higher latencies in our tests, as latencies are not as
important. With the AMD Athlon64 test system, the DDR RAM latency must run at
2-2-2-5, or the memory's tightest possible timings, as quick access is more
important to the CPU design.
Patriotic overclocking!
One thing
I like about overclocking PDP Systems' memory is that they state right from the
beginning what to expect. With the PDC1G3200+XBLK, they state on the label that
the memory can run at 200 MHz while using the most aggressive 2-2-2-5 memory
timings, while loosening things up to 3-4-4-8 should allow you to run up to 266
MHz.
Starting
with our Athlon64 test system, we raised the motherboard clock speed slowly
while maintaining 2-2-2-5 memory timings. At 213 MHz our system showed some
signs of instability and would often crash back to desktop when running 3D
applications. Raising the memory voltage to 2.8V stopped this. We easily passed
220 MHz and even 230 MHz, but seemed to hit a wall getting past 231 MHz with
tight timings. Loosening the timings allowed us overclock the memory a few MHz
higher, but we didn't really try too hard at this, since Athlon64 systems should
always be paired with low latency memory.
With the Intel rig, things didn't start off too
well. To get the memory to function properly at 200 MHz with 2-2-2-5 memory timings
we had to raise the voltage to 2.6V right off the bat. After that,
things went pretty smoothly and the memory overclocked nicely. Using the aggressive timings, we were
limited to a maximum of 223 MHz; anything higher and we'd get a BSOD
during bootup. Raising the voltage to 2.8V didn't help, so we decided to
lower memory timings to 3-4-4-8 to see how fast the memory could run at
higher latencies.
We were certainly not
disappointed and the memory hit a very respectable 264 MHz with only 2.8V! If we
tried to go past 264 MHz, the system would sometimes crash for no apparent
reason. PDP specifies that the 266MHz (533MHz) speed should be accompanied by a
voltage of 2.85 to 2.95V, so we were satisfied that it would reach its
advertised rate.
|
PCStats Test System Specs: |
|
Athlon64 test system |
Pentium 4 test system |
Processor: |
AMD Athlon64 3200+ |
Intel Pentium 4 2.4C |
Clock
Speed: |
10 x 200 MHz = 2 GHz 9 x
231 MHz = 2.08 GHz |
12 x 200 MHz = 2.4 GHz 12
x 264 MHz = 3.17 GHz |
Motherboards: |
Asus K8V DLX, K8T800 |
ABIT IC7-MAX3, i875P |
Videocard: |
Asus Radeon X800XT PE |
Asus Radeon X800XT PE |
Memory: |
2x 512MB PDP Systems Patriot
PDC1G3200+XBLK 2x 512MB Corsair TwinX3200XL PRO 2x 512MB OCZ PC3200 Platinum Limited Ed. 2x 512MB Corsair TwinX1024-4400C25PT 2x 512MB TwinMOS Speed Premium PC3200 |
Hard Drive:
|
40GB WD Special Ed |
20GB WD HDD |
CDROM: |
MSI X48 CD-RW/DVD-ROM |
NEC 52x CD-ROM |
PowerSupply: |
Vantec Stealth 470W |
PC Power & Cooling
510W |
Software Setup |
WindowsXP Build 2600 VIA 4in1 4.55V Catalyst
4.11 |
WindowsXP Build 2600 Intel INF 5.02.1012 Catalyst 4.11 |
Benchmarks |
Winstone 2004 SiSoft
Sandra 2004 PCMark04 3DMark2001SE X2: The
Threat UT2003 | |