Value RAM Roundup: Computing On a Budget
by Wesley Fink on April 11, 2005 4:26 PM EST- Posted in
- Memory
Performance Test Configuration
All seven Value RAM memories were tested on the DFI LANParty nF4 SLI-DR used in our most recent memory reviews. nForce4 is PCI Express, so we used the PCIe version of our standard nVidia 6800 Ultra for testing. Other components remain the same as used in the memory setup in Athlon 64 Memory: Rewriting the Rules. Performance was compared to the 5 top performing memories tested on this same platform in OCZ VX Revisited: DDR Updates on DFI nForce4 and Patriot DDR400 2-2-2/DDR533 3-4-4: Performance AND Value.The A64 test bed includes components that have been proven in Socket 939 Athlon 64 benchmarking, such as the Socket 939 4000+ (same specifications as FX53), and the OCZ Power Stream 520 Power Supply. Since the Athlon 64 tests represent a new series of DDR testing, we are using the current generation nVidia 6800 Ultra video card for benchmarking. We have found the 6800 Ultra to be a particularly good performance match to nVidia motherboards.
All other basic test conditions attempted to mirror those used in our earlier Intel memory reviews. However, test results are not directly comparable to tests performed on the Intel test bed.
| AMD nForce4 Performance Test Configuration | |
| Processor(s): | AMD 4000+ (FX53) Athlon 64 (2.4GHz, Socket 939, 1 MB cache, Dual Channel, 1000HT) |
| RAM: | Kingston KVR400X64C25/512 (DS) 2X512MB Kingston KVR400X64C3AK2/1G (DS) 2X512MB Mushkin PC3200 EM (DS) 2X512MB OCZ PC3200 Value Series (VX) (DS) 2X512MB OCZ PC3200 Gold (BH5) (DS) 2X512MB OCZ PC3200 Premier (DS) 2X512MB Transcend JM366D643A-50 (DS) 2X512MB Patriot PC3200+XLBT (DS) 2X512MB OCZ EL PC4000 VX Gold (DS) 2X512MB Corsair TwinX1024-4400C25 (DS) 2X512MB Crucial Ballistix (DS) 2X512MB OCZ PC3200 Platinum Rev 2 (DS) 2X512MB |
| Hard Drives: | Seagate 120GB SATA 7200RPM 8MB Cache |
| PCI/AGP Speed: | Fixed at 33/66 |
| Bus Master Drivers: | nVidia nForce Platform Driver 6.39 |
| Video Card(s): | nVidia 6800 Ultra 256MB PCIe, 256MB aperture, 1024x768x32 |
| Video Drivers: | nVidia Forceware 71.84 Release |
| Power Supply: | OCZ Power Stream 520W |
| Operating System(s): | Windows XP Professional SP1 |
| Motherboard: | DFI LANParty nF4 SLI-DR |
| BIOS: | 3/10/2005 Release |
In past benchmarking, we have found performance of the nForce4 and nForce3 chipsets to be virtually identical, and we have found AGP and PCIe performance to be virtually the same in the benchmarks that we use for memory testing. Therefore, you can also compare these results to TCCD benchmarks in recent memory reviews. The differences in results are that the nVidia 71.84 driver is a bit faster than the 61.77 used in earlier memory review. The DFI nForce4 platform is also a bit better at memory overclocking than the MSI K8N Neo2 used in past memory testing. The performance differences, however, are not large enough to negate performance comparisons.
With nForce3 motherboards, we achieved the fastest performance on AMD Athlon 64 chipsets (nForce3, VIA K8T800 PRO) at Cycle Time or tRAS of 10. However, the nForce4 appears to behave a bit differently with memory. Therefore, we ran a complete set of Memtest86 benchmarks with only tRAS varied to determine the best tRAS setting for these memories. We achieved the best bandwidth at tRAS settings ranging from 5 to 8, so a tRAS setting of 6 was used for testing wherever possible.
Test Settings
All AMD Athlon 64 processors are unlocked downward, and the FX CPUs are unlocked up and down. This feature allows a different approach to memory testing, which truly measures performance differences in memory speed alone. All tests were run with CPU speed as close to the specified 2.4GHz of the 4000+/FX53 as possible, with CPU speed/Memory Speed increased at lower multipliers to achieve 2.4Ghz. This approach allows the true measurement of the impact of higher memory speed and timings on performance, since CPU speed is fixed, removing CPU speed as a factor in memory performance.The following settings were tested with the seven Value RAM memories on the DFI nF4 test bed:
- 2.4GHz-12x200/DDR400 - the highest stock memory speed supported on VIA 939/nF3-4/SiS755-FX motherboards.
- 2.4GHz-11x218/DDR436 - a ratio near the standard DDR433 speed
- 2.4GHz-10x240/DDR480 - a ratio near the standard rating of DDR466. Several of the Value Memories achieved stable performance at this speed and complete benchmarks were run at this speed with those memories.
- Highest Memory Performance - the highest memory bandwidth and game performance that we could achieve with the memory being tested. This is rarely the highest memory speed that we could achieve. It is normally a lower speed with 1T Command Rate and tighter memory timings.
We ran our standard suite of memory performance benchmarks - Quake 3, Return to Castle Wolfenstein-Enemy Territory-Radar, Super Pi 2M, and Sandra 2004 Standard and UnBuffered. We also included Everest Home Edition memory tests, free at www.lavalys.com, for read speed, write speed, and Latency.
Facebook
Twitter
RSS
102 Comments
View All Comments
ChineseDemocracyGNR - Tuesday, April 12, 2005 - link
#39,"2) RAM multipliers are usually limited. If you have a standard set of 400, 320, and 266 speeds, you could only achieve DDR400 speed at a CPU frequency of 250. Anything lower than that would be running the RAM at less than 400. Most A64 CPUs can't do 250 on air at stock multipliers (the low end ones can) so they will be running less than optimum ram speed. That's where you could lower CPU multipliers or use a board like the DFI with lots of intermediate RAM ratios. "
I'm not sure I'm following.
With a 9x CPU multiplier, DDR266 memory divider and 312MHz reference clock the memory would be running at DDR400.
With a DDR200 option you could go up to 400MHz on the reference clock. That means that no DDR400 memory will limit the overclock of a CPU with a 9x multiplier.
"1) There is an Asynchronous Latency penalty, which can be tweaked somewhat on boards with better BIOS options like the DFI. It is not, hoever, the kind of asynchronous penalty you see on a FSB board like Intel. "
In my own tests there's no real-world penalty at all. I compared an Athlon 64 running at:
REFCLOCK: 200MHz
Memory Divider: DDR400
CPU multiplier: 9x
LDT: 5x
with:
REFCLOCK: 300MHz
Memory Divider: DDR266 (DDR400 effective)
CPU multiplier: 6x
LDT: 3x
The results where near identical.
"In other words, the easiest way to consistently improve memory performance is 1:1 memory speed."
There's no memory fast enough to run at 1:1 with an Athlon 64. ;)
An Athlon 64 at 9*200 is on a 1800/9 ratio with DDR400 memory.
I'm sorry but I stand by what I said before, there's no reason to invest in memory if you want to overclock your Athlon 64, only if you want to overclock the memory as well.
Kind of on the subject, I hope the round up of AMD PCI-E boards (there is one coming right?) tests the best reference clock the motherboards can achieve without memory as a limiting factor, unlike the reviews before.
PrinceGaz - Tuesday, April 12, 2005 - link
Excellent review, both for the modules it covered and what it didn't.One small point- there is no such thing as 1:1 memory timings with A64 processors. The reduced latency and higher performance that a 1:1 ration gave when the processor to chipset FSB was running synchronously with the chipset memory-controller, is irrelevant with the Athlon 64 as there is no intermediate bus operating at a differnt speed to the memory controller to cause overheads. Selecting a lower memory speed just changes the CPU:Memory ratio in the processor.
The memory on an Athlon 64 system works just as efficiently (though ay a lower bandwidth of course) if set to DDR333 as it does at DDR400, which means there is no real penalty when overclocking in choosing a lower memory speed to compensate for the increasing bus speed.
Olaf van der Spek - Tuesday, April 12, 2005 - link
http://anandtech.com/memory/showdoc.aspx?i=2392&am...> Transcend is another memory that costs just $100 for a Gigabyte and yet manages to nearly reach DDR550 in overclocking.
550?
The table claims 510 (2 x 223), but 2 x 223 = 446.
Olaf van der Spek - Tuesday, April 12, 2005 - link
Is Lavalys sponsoring this article? Why is that paragraph repeated on every page?JustAnAverageGuy - Tuesday, April 12, 2005 - link
I would have liked to see Mushkin Blue($147 per GB)
http://www.newegg.com/app/viewproductdesc.asp?desc...
and Corsair VS 2.5
($174 per gb)
http://www.newegg.com/app/ViewProductDesc.asp?desc...
I won't whine about the voltage, that's been done before :)
Olaf van der Spek - Tuesday, April 12, 2005 - link
-LX - Tuesday, April 12, 2005 - link
Why isn't the OCZ4001024WV3DC-K on the OCZ site???CobraT1 - Tuesday, April 12, 2005 - link
If you are interested in the OCZ Value VX, note the differences in the two part numbers, one with a "W" and one without.Value VX = OCZ4001024WV3DC-K
2.5-3-3-7 (picture) Supports EVP (Extended Voltage Protection)
Value = OCZ4001024V3DC-K
3-4-4-8 Does not Support EVP
See this link for both.
http://www.newegg.com/app/searchProductResult.asp?...
Hope this clears up the confusion.
segagenesis - Tuesday, April 12, 2005 - link
Wesley - Fair enough. But when that ATACOM link posted in #44 shows 3-4-4 even on the label in the picture its hard to tell who to believe (and its hard to read the part # on it). If its all the same chip then fine... but why label it differently then? Buyer beware?Maybe I am off base...
Turin39789 - Tuesday, April 12, 2005 - link
I get real tired pushing ferrari's out of my driveway. There isnt any racing alcohol available to me, sometimes I have my neghbor tow me to work in his chevrolet cobalt