//mAr
Jul 27, 2005, 07:07 PM
[page=Introduction & Specs]
Introduction
First we would like to thank Mips-Computer (http://www.mips-computer.de) for supplying us with this DDR-Ram.
After the success of the old "BH-5" PC3500 LVL2, Mushkin tries to continue the success story with the new Redline XP4000. They select the best Winbond UTT CH-5 and use them on these modules to have a big clearance above the specifications.
Also the specs are very good: 250 MHz with tightest timings of 2-2-2 (cas latency - ras to cas - row precharge) at 3.3-3.5V, but ACTIVE COOLING is required (in our review we use a Noiseblocker SX2 120mm fan @ 5V).
The XP4000 is also covered by Mushkin's Lifetime Warranty. If you haven't bought the RAM through mushkin.com you will have to register the memory's serial number, which is printed on the modules, for warranty. In case you might have any problems, Mushkin has a very good support via telephone or email.
Winbond UTT
Winbond UTT stands for UnTesTed, which means the ICs aren't tested by Winbond for any speed or timings and they are also unlabeled.
http://www.techpowerup.com/memdb/img/chip23.jpg
So the memory module manufacturer has to select the ICs, which is a lot of work, but they can charge a premium for it.
The UTT CH-5 are similiar to the old Winbond CH-5. They can run 2-3-2-X up to 22X MHz with 2.8V - 2.9V. However, the new UTT CH-5 loves voltage even more than the old CH-5. With more than 3.0+V they can get 2-2-2-X. Some old CH-5 did the same, but rarely.
Compared to the UTT BH-5, which can run 2-2-2-X with low voltage, the UTT CH-5 aren't as good as UTT BH-5 at the first sight. But the UTT CH-5 scale better with voltage. In most cases UTT BH-5 have their sweet spot at 3.3V - 3.4V and about 245 - 255 MHz. With 3.4V - 3.6V, UTT CH-5 can get about 255 - 265 MHz, which means higher performance and higher overclock.
Specification
<table border="1" class="resulttable" cellspacing="0" cellpadding="3">
<tr>
<th scope="row">Clock Speeds:</th>
<td scope="row">250 MHz</td>
</tr>
<tr>
<th scope="row">CAS Latency (Tcl)</th>
<td scope="row">2</td>
</tr>
<tr>
<th scope="row">RAS-to-CAS Delay (Trcd)</th>
<td scope="row">2</td>
</tr>
<tr>
<th scope="row">RAS Precharge (Trp)</th>
<td scope="row">2</td>
</tr>
<tr>
<th scope="row">Cycle Time</th>
<td scope="row">-</td>
</tr>
<tr>
<th valign="top" scope="row">Voltage</th>
<td valign="top" scope="row">3.3V-3.5V</td>
</tr>
<tr>
<th valign="top" scope="row">Package</th>
<td valign="top" scope="row">TSOP</td>
</tr>
</table>
From the Mushkin Product Page
The XP4000 REDLINE eliminates the memory as a potential bottleneck when overclocking.
THE REDLINE MEMORY MODULES ARE DESIGNED TO RUN AT HIGHER VOLTAGES UP TO 3.5V AND SUFFICIENT COOLING IS REQUIRED.
Active Cooling Required<br />
Due to the extreme speed of these modules along with the added voltages used, Mushkin requires end users to Actively cool their memory modules. Modules that are not properly cooled will/may not meet rated timings and speeds, and may become susceptible to premature failure. We generally recommend 15-25cfm of direct airflow over the ram modules which can be easily accomplished with one 60mm or 80mm low rpm fan. This will insure increased stability and lifespan of your Redline modules.
Recommended Platforms<br />
Mushkin tests all Redline memory modules on Athlon64 platforms. Furthermore, when running two modules, only Socket 939 platforms are validated to run at rated specifications/timings. We currently recommend and validate the Redline memory series on the DFI LanParty UT NForce4 Ultra-D, DFI LanParty UT NForce4 SLI-DR, and DFI LanParty NForce4 SLI-DR boards as they provide the best combination of voltage options and compatibility for the Redline Series. Other motherboards may work, but may require modification for adequate voltages, and may not produce optimal results. Single Channel Athlon64 systems may have a hard time running 2x512mb dimms at rated speeds due to memory load on the single channel memory controller. Pentium4 and AthlonXP systems generally will not obtain rated specifications secondary to motherboard/memory controller limitations. These modules are compatible with most motherboards for Pentium4/AthlonXP CPUs, however end users are advised that lower frequencies are achieved.
The DFI Lanpary NF4 Series boards are the only validated boards for using this RAM at full speed.
On a small printed note Mushkin gives a small installation guide for the DFI Lanparty NF4 Series.
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/pape_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/pape_big.jpg)
Packaging
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/packfront_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/packfront_big.jpg) http://www.techpowerup.com/reviews/Mushkin/XP4000/images/packback_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/packback_big.jpg) http://www.techpowerup.com/reviews/Mushkin/XP4000/images/ram_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/ram_big.jpg)
[pagE=Test Setup & Testing]
Test Setup
Here you can see the system we tested on.
<table border="1" cellpadding="3" cellspacing="0" class="ramtable">
<tr align="center">
<th colspan="2" scope="row" style="font-size:larger;text-align:center">Test System</th>
</tr>
<tr>
<th width="116" scope="row">CPU:</th>
<td scope="row">AMD Athlon64 4000+ (S939; 1024KB; SanDiego)</td>
</tr>
<tr>
<th scope="row">Motherboard:</th>
<td scope="row">DFI Lanparty nForce4 SLI-D</td>
</tr>
<tr>
<th scope="row">Memory:</th>
<td scope="row">2x 512 MB Mushkin Redline XP4000</td>
</tr>
<tr>
<th valign="top" scope="row">Video Card:</th>
<td valign="top" scope="row">2x Pine XFX 6800 GT SLI @ 450/1200</td>
</tr>
<tr>
<th valign="top" scope="row">Harddisk:</th>
<td valign="top" scope="row">2x 74GB WD Raptor RAID 0</td>
</tr>
<tr>
<th valign="top" scope="row">Power Supply:</th>
<td valign="top" scope="row">OCZ ADJ 520W</td>
</tr>
<tr>
<th valign="top" scope="row">Software:</th>
<td valign="top" scope="row">Windows XP SP2, Forceware 77.60</td>
</tr>
</table>
All parts are selected with high performance and good overclockability in mind to reach high test results. We used BIOS 510-FIX2 (updated version of Mushkin's recommended 510-2) for the DFI LanParty nForce4 SLI-D.
All AlphaTimings in the Bios were set to auto, except for:
CPC - Enabled
DRAM bank Interleaving - Enabled
Idle Cycle Limit - 256
Dynamic Counter - Enabled
R/W Queue Bypass - 16x
Bypass Max - 07x.
TRAS - 5
TRC - 7
TRFC - 14
For JEDEC DDR400A Benchmarks all alpha timings were set to auto.
Testing
<table width="739" border="1" cellpadding="3" cellspacing="0" class="ramtable">
<tr>
<th colspan="9" scope="row" style="font-size:larger">Mushkin Redline XP4000</th>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">CPU Clock & <br />
Memory Ratio</th>
<th scope="row" width="58">Memory<br />
Speed</th>
<th scope="row" width="121">Memory<br />
Timings</th>
<th scope="row" width="70">Everest<br />
Read</th>
<th scope="row" width="76">Everest<br />
Write</th>
<th scope="row" width="54">Everest<br />
Latency</th>
<th nowrap="nowrap" scope="row" width="70">Quake 3<br />
Timedemo</th>
<th scope="row" width="59">3DMark<br />
2001SE</th>
<th scope="row" width="52">SuperPi<br />
Mod 1M</th>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 200 1:1</th>
<td scope="row" width="58">200 MHz</td>
<td scope="row" width="100">2-2-2-5 2.9V</td>
<td scope="row" width="70">5390 MB/s</td>
<td scope="row" width="76">2496 MB/s</td>
<td scope="row" width="54">47.1 ns</td>
<td scope="row" width="70">365.2 FPS</td>
<td scope="row" width="59">28129</td>
<td scope="row" width="52">37.8s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 250 1:1</th>
<td scope="row" width="58">250 MHz</td>
<td scope="row" width="100">2-2-2-5 3.1V</td>
<td scope="row" width="70">6716 MB/s</td>
<td scope="row" width="76">3113 MB/s</td>
<td scope="row" width="54">38.5 ns</td>
<td scope="row" width="70">453.1 FPS</td>
<td scope="row" width="59">34126</td>
<td scope="row" width="52">30.0s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 251 1:1</th>
<td scope="row" width="58">251 MHz</td>
<td scope="row" width="100">2-2-2-5 3.1V</td>
<td scope="row" width="70">6719 MB/s</td>
<td scope="row" width="76">3121 MB/s</td>
<td scope="row" width="54">38.4 ns</td>
<td scope="row" width="70">455.4 FPS</td>
<td scope="row" width="59">34180</td>
<td scope="row" width="52">29.9s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 257 1:1</th>
<td scope="row" width="58">257 MHz</td>
<td scope="row" width="100">2-2-2-5 3.2V</td>
<td scope="row" width="70">6902 MB/s</td>
<td scope="row" width="76">3206 MB/s</td>
<td scope="row" width="54">36.9 ns</td>
<td scope="row" width="70">464.1 FPS</td>
<td scope="row" width="59">35438</td>
<td scope="row" width="52">28.9s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 262 1:1</th>
<td scope="row" width="58">262 MHz</td>
<td scope="row" width="100">2-2-2-5 3.4V</td>
<td scope="row" width="70">7028 MB/s</td>
<td scope="row" width="76">3271 MB/s</td>
<td scope="row" width="54">36.1 ns</td>
<td scope="row" width="70">471.2 FPS</td>
<td scope="row" width="59">35809</td>
<td scope="row" width="52">28.4s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 265 1:1</th>
<td scope="row" width="58">265 MHz</td>
<td scope="row" width="100">2-2-2-5 3.5V</td>
<td scope="row" width="70">7101 MB/s</td>
<td scope="row" width="76">3305 MB/s</td>
<td scope="row" width="54">35.7 ns</td>
<td scope="row" width="70">477.8 FPS</td>
<td scope="row" width="59">36512</td>
<td scope="row" width="52">28.2s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">JEDEC DDR400A</th>
<td scope="row" width="58">200 MHz</td>
<td scope="row" width="100">2,5-3-3-8 2.6V</td>
<td scope="row" width="70">5130 MB/s</td>
<td scope="row" width="76">1710 MB/s</td>
<td scope="row" width="54">54.3 ns</td>
<td scope="row" width="70">354.9 FPS</td>
<td scope="row" width="59">27615</td>
<td scope="row" width="52">39.2s</td>
</tr>
</table>
Very good results, but a motherboard with enough VDIMM is required.
Reduced latencies did not give much overclocking headroom. This shows, that this memory is designed for low latencies. Faster than DDR500 with tight timings is very fast, compared to 2.5-3-3-7.
In the factory, the SPD Timings were not set to the correct values of 2-2-2-5 for DDR500. They were set to 2-3-2-6 for DDR400. Those are the best possible timings for DDR400 using 2.6V. So it's not completely wrong.
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/265_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/265_big.jpg)
265 MHz @ 2-2-2-5 (3.5V)
Benchmarks
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/everestread.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/everestwrite.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/everestlatency.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/quake3.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/3dmark2001.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/superpi.gif
For easier comparison with other modules, we set a maximum voltage of 3.1V und tested until we found the highest clock frequency and fastest timings for this memory. The Benchmarks Everest Read, Everest Write and Quake 3 were run. We then calculated the performance increase in percent compared to some standard DDR400 memory running at JEDEC DDR400A (2.5-3-3-8). The average percentage of the three benchmarks is listed in following table:
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/vsjedec.gif
[Page=Value & Conclusion]
Value & Conclusion
<table width="100%" cellpadding="5" cellspacing="0" id="result">
<tr><th>http://www.techpowerup.com/images/dollar.gif</th>
<td>
The 2x512 MB Dual-Channel Kit is sold for $207 directly through mushkin.com, which is a good price for such an overclocker's dream.</td>
</tr><tr>
<th>http://www.techpowerup.com/images/thumbup.gif</th>
<td>
Tightest latencies
Very good performance
Very nice red heatspreader
Lifetime Warranty (which covers Vdimm up to 3.5V)
</td>
</tr>
<tr>
<th>http://www.techpowerup.com/images/thumbdown.gif</th>
<td>
Lots of Voltage is required
Active Cooling is required
SPD defaults to semi wrong values
</td></tr>
<tr><th>9.2</th>
<td>The Mushkin Redline XP4000 accommodates selected Winbond UTT CH-5 chips. For overclockers it's probably the fastest DDR-RAM in the world with the best performance. This memory can run the tightest timings without any problems. But don't forget to cool it with a fan, which is a requirement for stable operation.<br />
You will also need a motherboard which can give you the high DDR voltages, the memory needs to really shine. If your board supports only up to 3.3V VDDR, or even less you should look elsewhere.</td></tr>
<tr><th></th><td>http://www.techpowerup.com/images/editorschoice.gif</td></tr></table>
Introduction
First we would like to thank Mips-Computer (http://www.mips-computer.de) for supplying us with this DDR-Ram.
After the success of the old "BH-5" PC3500 LVL2, Mushkin tries to continue the success story with the new Redline XP4000. They select the best Winbond UTT CH-5 and use them on these modules to have a big clearance above the specifications.
Also the specs are very good: 250 MHz with tightest timings of 2-2-2 (cas latency - ras to cas - row precharge) at 3.3-3.5V, but ACTIVE COOLING is required (in our review we use a Noiseblocker SX2 120mm fan @ 5V).
The XP4000 is also covered by Mushkin's Lifetime Warranty. If you haven't bought the RAM through mushkin.com you will have to register the memory's serial number, which is printed on the modules, for warranty. In case you might have any problems, Mushkin has a very good support via telephone or email.
Winbond UTT
Winbond UTT stands for UnTesTed, which means the ICs aren't tested by Winbond for any speed or timings and they are also unlabeled.
http://www.techpowerup.com/memdb/img/chip23.jpg
So the memory module manufacturer has to select the ICs, which is a lot of work, but they can charge a premium for it.
The UTT CH-5 are similiar to the old Winbond CH-5. They can run 2-3-2-X up to 22X MHz with 2.8V - 2.9V. However, the new UTT CH-5 loves voltage even more than the old CH-5. With more than 3.0+V they can get 2-2-2-X. Some old CH-5 did the same, but rarely.
Compared to the UTT BH-5, which can run 2-2-2-X with low voltage, the UTT CH-5 aren't as good as UTT BH-5 at the first sight. But the UTT CH-5 scale better with voltage. In most cases UTT BH-5 have their sweet spot at 3.3V - 3.4V and about 245 - 255 MHz. With 3.4V - 3.6V, UTT CH-5 can get about 255 - 265 MHz, which means higher performance and higher overclock.
Specification
<table border="1" class="resulttable" cellspacing="0" cellpadding="3">
<tr>
<th scope="row">Clock Speeds:</th>
<td scope="row">250 MHz</td>
</tr>
<tr>
<th scope="row">CAS Latency (Tcl)</th>
<td scope="row">2</td>
</tr>
<tr>
<th scope="row">RAS-to-CAS Delay (Trcd)</th>
<td scope="row">2</td>
</tr>
<tr>
<th scope="row">RAS Precharge (Trp)</th>
<td scope="row">2</td>
</tr>
<tr>
<th scope="row">Cycle Time</th>
<td scope="row">-</td>
</tr>
<tr>
<th valign="top" scope="row">Voltage</th>
<td valign="top" scope="row">3.3V-3.5V</td>
</tr>
<tr>
<th valign="top" scope="row">Package</th>
<td valign="top" scope="row">TSOP</td>
</tr>
</table>
From the Mushkin Product Page
The XP4000 REDLINE eliminates the memory as a potential bottleneck when overclocking.
THE REDLINE MEMORY MODULES ARE DESIGNED TO RUN AT HIGHER VOLTAGES UP TO 3.5V AND SUFFICIENT COOLING IS REQUIRED.
Active Cooling Required<br />
Due to the extreme speed of these modules along with the added voltages used, Mushkin requires end users to Actively cool their memory modules. Modules that are not properly cooled will/may not meet rated timings and speeds, and may become susceptible to premature failure. We generally recommend 15-25cfm of direct airflow over the ram modules which can be easily accomplished with one 60mm or 80mm low rpm fan. This will insure increased stability and lifespan of your Redline modules.
Recommended Platforms<br />
Mushkin tests all Redline memory modules on Athlon64 platforms. Furthermore, when running two modules, only Socket 939 platforms are validated to run at rated specifications/timings. We currently recommend and validate the Redline memory series on the DFI LanParty UT NForce4 Ultra-D, DFI LanParty UT NForce4 SLI-DR, and DFI LanParty NForce4 SLI-DR boards as they provide the best combination of voltage options and compatibility for the Redline Series. Other motherboards may work, but may require modification for adequate voltages, and may not produce optimal results. Single Channel Athlon64 systems may have a hard time running 2x512mb dimms at rated speeds due to memory load on the single channel memory controller. Pentium4 and AthlonXP systems generally will not obtain rated specifications secondary to motherboard/memory controller limitations. These modules are compatible with most motherboards for Pentium4/AthlonXP CPUs, however end users are advised that lower frequencies are achieved.
The DFI Lanpary NF4 Series boards are the only validated boards for using this RAM at full speed.
On a small printed note Mushkin gives a small installation guide for the DFI Lanparty NF4 Series.
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/pape_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/pape_big.jpg)
Packaging
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/packfront_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/packfront_big.jpg) http://www.techpowerup.com/reviews/Mushkin/XP4000/images/packback_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/packback_big.jpg) http://www.techpowerup.com/reviews/Mushkin/XP4000/images/ram_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/ram_big.jpg)
[pagE=Test Setup & Testing]
Test Setup
Here you can see the system we tested on.
<table border="1" cellpadding="3" cellspacing="0" class="ramtable">
<tr align="center">
<th colspan="2" scope="row" style="font-size:larger;text-align:center">Test System</th>
</tr>
<tr>
<th width="116" scope="row">CPU:</th>
<td scope="row">AMD Athlon64 4000+ (S939; 1024KB; SanDiego)</td>
</tr>
<tr>
<th scope="row">Motherboard:</th>
<td scope="row">DFI Lanparty nForce4 SLI-D</td>
</tr>
<tr>
<th scope="row">Memory:</th>
<td scope="row">2x 512 MB Mushkin Redline XP4000</td>
</tr>
<tr>
<th valign="top" scope="row">Video Card:</th>
<td valign="top" scope="row">2x Pine XFX 6800 GT SLI @ 450/1200</td>
</tr>
<tr>
<th valign="top" scope="row">Harddisk:</th>
<td valign="top" scope="row">2x 74GB WD Raptor RAID 0</td>
</tr>
<tr>
<th valign="top" scope="row">Power Supply:</th>
<td valign="top" scope="row">OCZ ADJ 520W</td>
</tr>
<tr>
<th valign="top" scope="row">Software:</th>
<td valign="top" scope="row">Windows XP SP2, Forceware 77.60</td>
</tr>
</table>
All parts are selected with high performance and good overclockability in mind to reach high test results. We used BIOS 510-FIX2 (updated version of Mushkin's recommended 510-2) for the DFI LanParty nForce4 SLI-D.
All AlphaTimings in the Bios were set to auto, except for:
CPC - Enabled
DRAM bank Interleaving - Enabled
Idle Cycle Limit - 256
Dynamic Counter - Enabled
R/W Queue Bypass - 16x
Bypass Max - 07x.
TRAS - 5
TRC - 7
TRFC - 14
For JEDEC DDR400A Benchmarks all alpha timings were set to auto.
Testing
<table width="739" border="1" cellpadding="3" cellspacing="0" class="ramtable">
<tr>
<th colspan="9" scope="row" style="font-size:larger">Mushkin Redline XP4000</th>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">CPU Clock & <br />
Memory Ratio</th>
<th scope="row" width="58">Memory<br />
Speed</th>
<th scope="row" width="121">Memory<br />
Timings</th>
<th scope="row" width="70">Everest<br />
Read</th>
<th scope="row" width="76">Everest<br />
Write</th>
<th scope="row" width="54">Everest<br />
Latency</th>
<th nowrap="nowrap" scope="row" width="70">Quake 3<br />
Timedemo</th>
<th scope="row" width="59">3DMark<br />
2001SE</th>
<th scope="row" width="52">SuperPi<br />
Mod 1M</th>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 200 1:1</th>
<td scope="row" width="58">200 MHz</td>
<td scope="row" width="100">2-2-2-5 2.9V</td>
<td scope="row" width="70">5390 MB/s</td>
<td scope="row" width="76">2496 MB/s</td>
<td scope="row" width="54">47.1 ns</td>
<td scope="row" width="70">365.2 FPS</td>
<td scope="row" width="59">28129</td>
<td scope="row" width="52">37.8s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 250 1:1</th>
<td scope="row" width="58">250 MHz</td>
<td scope="row" width="100">2-2-2-5 3.1V</td>
<td scope="row" width="70">6716 MB/s</td>
<td scope="row" width="76">3113 MB/s</td>
<td scope="row" width="54">38.5 ns</td>
<td scope="row" width="70">453.1 FPS</td>
<td scope="row" width="59">34126</td>
<td scope="row" width="52">30.0s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 251 1:1</th>
<td scope="row" width="58">251 MHz</td>
<td scope="row" width="100">2-2-2-5 3.1V</td>
<td scope="row" width="70">6719 MB/s</td>
<td scope="row" width="76">3121 MB/s</td>
<td scope="row" width="54">38.4 ns</td>
<td scope="row" width="70">455.4 FPS</td>
<td scope="row" width="59">34180</td>
<td scope="row" width="52">29.9s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 257 1:1</th>
<td scope="row" width="58">257 MHz</td>
<td scope="row" width="100">2-2-2-5 3.2V</td>
<td scope="row" width="70">6902 MB/s</td>
<td scope="row" width="76">3206 MB/s</td>
<td scope="row" width="54">36.9 ns</td>
<td scope="row" width="70">464.1 FPS</td>
<td scope="row" width="59">35438</td>
<td scope="row" width="52">28.9s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 262 1:1</th>
<td scope="row" width="58">262 MHz</td>
<td scope="row" width="100">2-2-2-5 3.4V</td>
<td scope="row" width="70">7028 MB/s</td>
<td scope="row" width="76">3271 MB/s</td>
<td scope="row" width="54">36.1 ns</td>
<td scope="row" width="70">471.2 FPS</td>
<td scope="row" width="59">35809</td>
<td scope="row" width="52">28.4s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">11 x 265 1:1</th>
<td scope="row" width="58">265 MHz</td>
<td scope="row" width="100">2-2-2-5 3.5V</td>
<td scope="row" width="70">7101 MB/s</td>
<td scope="row" width="76">3305 MB/s</td>
<td scope="row" width="54">35.7 ns</td>
<td scope="row" width="70">477.8 FPS</td>
<td scope="row" width="59">36512</td>
<td scope="row" width="52">28.2s</td>
</tr>
<tr>
<th nowrap="nowrap" scope="row" width="126">JEDEC DDR400A</th>
<td scope="row" width="58">200 MHz</td>
<td scope="row" width="100">2,5-3-3-8 2.6V</td>
<td scope="row" width="70">5130 MB/s</td>
<td scope="row" width="76">1710 MB/s</td>
<td scope="row" width="54">54.3 ns</td>
<td scope="row" width="70">354.9 FPS</td>
<td scope="row" width="59">27615</td>
<td scope="row" width="52">39.2s</td>
</tr>
</table>
Very good results, but a motherboard with enough VDIMM is required.
Reduced latencies did not give much overclocking headroom. This shows, that this memory is designed for low latencies. Faster than DDR500 with tight timings is very fast, compared to 2.5-3-3-7.
In the factory, the SPD Timings were not set to the correct values of 2-2-2-5 for DDR500. They were set to 2-3-2-6 for DDR400. Those are the best possible timings for DDR400 using 2.6V. So it's not completely wrong.
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/265_small.jpg (http://www.techpowerup.com/reviews/Mushkin/XP4000/images/265_big.jpg)
265 MHz @ 2-2-2-5 (3.5V)
Benchmarks
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/everestread.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/everestwrite.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/everestlatency.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/quake3.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/3dmark2001.gif http://www.techpowerup.com/reviews/Mushkin/XP4000/images/superpi.gif
For easier comparison with other modules, we set a maximum voltage of 3.1V und tested until we found the highest clock frequency and fastest timings for this memory. The Benchmarks Everest Read, Everest Write and Quake 3 were run. We then calculated the performance increase in percent compared to some standard DDR400 memory running at JEDEC DDR400A (2.5-3-3-8). The average percentage of the three benchmarks is listed in following table:
http://www.techpowerup.com/reviews/Mushkin/XP4000/images/vsjedec.gif
[Page=Value & Conclusion]
Value & Conclusion
<table width="100%" cellpadding="5" cellspacing="0" id="result">
<tr><th>http://www.techpowerup.com/images/dollar.gif</th>
<td>
The 2x512 MB Dual-Channel Kit is sold for $207 directly through mushkin.com, which is a good price for such an overclocker's dream.</td>
</tr><tr>
<th>http://www.techpowerup.com/images/thumbup.gif</th>
<td>
Tightest latencies
Very good performance
Very nice red heatspreader
Lifetime Warranty (which covers Vdimm up to 3.5V)
</td>
</tr>
<tr>
<th>http://www.techpowerup.com/images/thumbdown.gif</th>
<td>
Lots of Voltage is required
Active Cooling is required
SPD defaults to semi wrong values
</td></tr>
<tr><th>9.2</th>
<td>The Mushkin Redline XP4000 accommodates selected Winbond UTT CH-5 chips. For overclockers it's probably the fastest DDR-RAM in the world with the best performance. This memory can run the tightest timings without any problems. But don't forget to cool it with a fan, which is a requirement for stable operation.<br />
You will also need a motherboard which can give you the high DDR voltages, the memory needs to really shine. If your board supports only up to 3.3V VDDR, or even less you should look elsewhere.</td></tr>
<tr><th></th><td>http://www.techpowerup.com/images/editorschoice.gif</td></tr></table>