Your link here
|
|
| | | | | | | |
| | | | | | |
|
|
DDR Memory Analysis. Part 4: Corsair DDR2 Modules
We proceed with a new series of articles devoted to the low level analysis
of the most important characteristics of memory modules using our
RightMark Memory Analyzer
test package. The purpose of this analysis is to provide information
on compatibility of a given memory module from a given manufacturer
with different mainboards based on various chipsets. The object of
our next analysis is a matched pair of DDR2 Corsair modules totaling
1 GB.
Manufacturer Information
Module manufacturer: Corsair Memory
Chip manufacturer: Micron Technology, Inc.
Web site of the module manufacturer: http://www.corsairmemory.com/corsair/xms2.html
Web site of the chip manufacturer: http://www.micron.com/products/dram/ddr2sdram/
Module Appearance
Photo of the memory module
Photo of the memory chip
Part Numbering System of Modules and Chips
Module Part Number Expansion
The brief technical documentation for this type of DDR2 Corsair memory modules (http://www.corsairmemory.com/corsair/products/specs/twin2x1024-5400c4.pdf) does not provide information on the expansion of some part number components. Nevertheless, the part numbering system of Corsair modules is rather simple and is easy to decipher on one's own.
| Field |
Value |
Expansion |
| 0 |
CM |
Manufacturer: CM = Corsair Memory |
| 1 |
2X |
Module type: 2X = DDR2 |
| 2 |
512 |
Capacity of one module: 512 MB |
| 3 |
5400 |
Theoretical throughput of the module:
5400 MB/sec (675 MHz) |
| 4 |
C4 |
CAS# Latency: 4 |
Out of additional technical characteristics of the modules we should note the recommended 1.9 V voltage, which exceeds the voltage noted in the official specification of JEDEC (1.8 V), as well as the recommended 4-4-4-12 timings, which (as the documentation claims) are written in the SPD module chip.
Chip Part Number Expansion
| FBGA code |
Part Number |
| D9BQM |
MT47H32M8BP-37E |
As we have previously noted, due to space limitations FBGA-packaged Micron memory chips used in Kingmax DDR2 modules have an abbreviated part marking that is different from the part number. You can decode an abbreviated part marking of any Micron memory chip into a part number at http://www.micron.com/decoder/
| Field |
Value |
Expansion |
| 0 |
MT |
Manufacturer code:
MT = Micron Technology |
| 1 |
47 |
Product family:
47 = DDR2 |
| 2 |
H |
Process technology:
H = SSTL 1.8V VDD CMOS |
| 3 |
32M8 |
Device number: 256Mbit (32Mbit x8) |
| 4 |
|
Device version (n/a) |
| 5 |
FP |
Package code:
FP = lead plating, FBGA, 60-pin 8x12 |
| 6 |
-37E |
Access/cycle time:
-37E = 3.75 ns (266.7 MHz), 4-4-4 |
| 7 |
|
Special options (n/a) |
| 8 |
|
Operating temperatures:
Blank = 0..85oC |
| 9 |
|
Special conditions (n/a) |
It's interesting to note that the PC2-5300 memory modules under review use quite standard Micron DDR2-533 memory chips, which are designed to operate in this mode with standard 4-4-4 timings. In particular, the same chips, but with lead plating, are used in DDR2 Kingmax modules.
SPD module chip data
Description of the SPD general standard:
JEDEC Standard No. 21-C, 4.1.2 - SERIAL PRESENCE DETECT STANDARD, General Standard
Description of the SPD specific standard for DDR2:
JEDEC Standard No. 21-C, 4.1.2.10 - Appendix X: Specific SPDs for DDR2 SDRAM (Revision 1.0)
JC-45 Appendix X: Specific PD's for DDR2 SDRAM (Revision 1.1)
| Function |
Byte |
Value |
Expansion |
| Fundamental Memory Type |
2 |
08h |
DDR2 SDRAM |
| Number of Row Addresses on this assembly |
3 |
0Dh |
13 (RA0-RA12) |
| Number of Column Addresses on this assembly |
4 |
0Ah |
10 (CA0-CA9) |
| Number of DIMM Banks |
5 |
61h |
2 physical banks |
| Data Width of this assembly |
6 |
40h |
64 bit |
| Voltage Interface Level of this assembly |
8 |
05h |
SSTL 1.8V |
| SDRAM Cycle time (tCK) at maximum supported
CAS# latency (CL X) |
9 |
30h |
3.00 ns (333.3 MHz) |
| DIMM configuration type |
11 |
00h |
Non-ECC |
| Refresh Rate/Type |
12 |
82h |
7.8125 ms – 0.5x reduced self-refresh |
| Primary SDRAM Width (organization type) of the
memory module chips |
13 |
08h |
x8 |
| Error Checking SDRAM Width (organization type)
of the memory chips in the ECC module |
14 |
00h |
Not defined |
| Burst Lengths Supported (BL) |
16 |
0Ch |
BL = 4, 8 |
| Number of Banks on SDRAM Device |
17 |
04h |
4 |
| CAS Latency (CL) |
18 |
10h |
CL = 4 |
| Minimum clock cycle (tCK) at reduced
CAS latency (CL X-1) |
23 |
00h |
Not defined |
| Minimum clock cycle (tCK) at reduced
CAS latency (CL X-2) |
25 |
00h |
Not defined |
| Minimum Row Precharge Time (tRP) |
27 |
3Ch |
15.0 ns
45, CL = 4 |
| Minimum Row Active to Row Active delay (tRRD) |
28 |
1Eh |
7.5 ns
2.5, CL = 4 |
| Minimum RAS to CAS delay (tRCD) |
29 |
3Ch |
15.0 ns
5, CL = 4 |
| Minimum Active to Precharge Time (tRAS) |
30 |
2Dh |
45.0 ns
15, CL = 4 |
| Module Bank Density |
31 |
40h |
256 MB |
| Write recovery time (tWR) |
36 |
3Ch |
15.0 ns
5, CL = 4 |
| Internal write to read command delay (tWTR) |
37 |
1Eh |
7.5 ns
2.5, CL = 4 |
| Internal read to precharge command delay (tRTP) |
38 |
1Eh |
7.5 ns
2.5, CL = 4 |
| Minimum Active to Active/Refresh Time (tRC) |
41, 40 |
3Ch, 00h |
60.0 ns
20, CL = 4 |
| Minimum Refresh to Active/Refresh Command Period
(tRFC) |
42, 40 |
4Bh, 00h |
75.0 ns
25, CL = 4 |
| Maximum device cycle time (tCKmax) |
43 |
80h |
8.0 ns |
| SPD Revision |
62 |
10h |
Revision 1.0 |
| Checksum for Bytes 0-62 |
63 |
E0h |
224 (true) |
| Manufacturer’s JEDEC ID Code (only the first
significant bytes are shown) |
64-71 |
FFh, 7Fh,
9Eh |
Corsair |
| Module Part Number |
73-90 |
- |
CM2X512-5400C4 |
| Module Manufacturing Date |
93-94 |
00h, 00h |
Not defined |
| Module Serial Number |
95-98 |
00h, 00h,
00h, 00h |
Not defined |
According to the SPD chip data, this module can operate with the only possible CAS# latency (CL X) of 4, with the 3.00 ns cycle time, that is at 333.3 MHz (in DDR2-667 mode). Memory timings for this case can be written as 4-5-5-15, which actually conflicts with the data provided in the module specification. Nevertheless, absolute values of tRCD (15.0 ns), tRP (15.0 ns) and tRAS (45.0 ns) can be considered standard for DDR2-533 modules, familiar to us from the previous reviews (which is not at all surprising, considering the use of "standard" DDR2-533 chips). Among the other features of the SPD chip is the lack of data on the manufacturing date and the module serial number.
Testbed Configurations and Software
Mainboards based on the chipsets of Intel 915 series
Testbed #1
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i915G
- Mainboard: Intel D915GUX, BIOS 1028 dated 06/29/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Testbed #2
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i915P
- Mainboard: Foxconn 915A01-P-8EKRS, BIOS 6.00 PG dated 06/10/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Testbed #3
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i915G
- Mainboard: Foxconn 915M03-G-8EKRS2, BIOS 6.00 PG dated 05/29/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Testbed #4
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i915P
- Mainboard: MSI 915P Neo2, BIOS V1.3B0 dated 09/08/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Testbed #5
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i915G
- Mainboard: MSI 915G
Combo, BIOS 080011 dated 07/14/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Testbed #6
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i915G
- Mainboard: ASUS P5GDC-V, BIOS 1003.003 dated 08/18/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Mainboards based on the chipsets of Intel 925 series
Testbed #7
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i925X
- Mainboard: Gigabyte 8ANDXP-D,
BIOS F1 dated 06/07/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Testbed #8
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i925X
- Mainboard: Intel D925XCV, BIOS 1259 dated 08/19/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Testbed #9
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i925X
- Mainboard: ASUS P5AD2, BIOS 1004.007 dated 07/02/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Testbed #10
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i925X
- Mainboard: MSI 925X Neo, BIOS 6.00 PG dated 06/18/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Testbed #11
- CPU: Intel Pentium 4 2.8 GHz (Prescott core, 1 MB L2)
- Chipset: Intel i925X
- Mainboard: Foxconn 925A01-8EKRS, BIOS 6.00 PG dated 08/28/2004
- Memory: 2x512 MB Corsair DDR2-675
- Video: Leadtek PX350 TDH, nVidia PCX5900
- HDD: WD Raptor WD360, SATA, 10000 rpm, 36Gb
- Operating system: Windows XP Professional Service Pack 2
- Drivers: Intel Chipset Utility 6.0.1.1002, nVidia Forceware 61.77
Test Results
As we have already written before, we tested memory modules in two modes. The first series of tests (performance tests) were carried out in normal mode, that is with standard timings written in the SPD chip. The second series (stability tests) – in the "extreme" mode with maximum possible timings for a given module on a given mainboard.
Mainboards based on the chipsets of Intel 915 series
Performance tests
As the current generation of Intel 915/925 chipsets does not allow the memory subsystem to operate at the frequencies over 300 (and in most mainboards 266.7) MHz, the first series of tests – performance tests of DDR2 Corsair modules – was carried out at the 266.7 MHz memory frequency (in DDR2-533 mode) with the nominal (according to the DDR2 standard) 1.8 V voltage and with 4-4-4-12 timings, which were set by default in BIOS in all tested mainboards, although the DDR2-533 mode is not written in the SPD chips of the modules.
|
| Parameter* |
Testbed 1
|
Testbed 2
|
Testbed 3
|
Testbed 4
|
Testbed 5
|
Testbed 6
|
| Timings |
4-4-4-12
|
4-4-4-12
|
4-4-4-12
|
4-4-4-12
|
4-4-4-12
|
4-4-4-12
|
| Average memory read bandwidth, MB/sec |
4505
|
4504
|
4497
|
4567
|
4524
|
4530
|
| Average memory write bandwidth, MB/sec |
2086
|
2080
|
2083
|
2124
|
2071
|
2075
|
| Max. memory read bandwidth, MB/sec |
6390
|
6395
|
6396
|
6491
|
6416
|
6434
|
| Max. memory write bandwidth, MB/sec |
4267
|
4267
|
4267
|
4324
|
4267
|
4282
|
| Minimum Pseudo-Random Access Latency, ns** |
54.4
|
54.6
|
54.6
|
53.8
|
54.4
|
54.2
|
| Maximum Pseudo-Random Access Latency, ns** |
63.3
|
63.6
|
63.7
|
62.7
|
63.4
|
63.2
|
| Minimum Random Access Latency, ns** |
128.6
|
128.5
|
128.4
|
126.9
|
128.8
|
128.0
|
| Maximum Random Access Latency, ns** |
153.7
|
153.5
|
153.5
|
151.6
|
153.8
|
153.0
|
*the best results are in bold
**Block size – 16 MB
|
MSI 915P Neo2 (Testbed #4) is again leading among the mainboards based on the 915 series of Intel chipsets. It is followed by ASUS P5GDC-V (Testbed #6) and MSI 915G Combo (Testbed #5) pushing Foxconn (Testbeds #2, 3) and Intel (Testbed #1) mainboards to the last place. As in previous tests of Kingston and Kingmax modules, no advantage can be seen of the i915P chipset over i915G (or vice versa).
Stability tests
According to our test methods, the second series of tests – DDR2 Corsair stability tests – were carried out with minimum possible timings, which would not lead to glitches. In mainboards allowing memory voltage control from BIOS we set this voltage to 1.9 V recommended by the module manufacturers. Memory voltage in other mainboards was supposed to be nominal 1.8 V.
Note that Corsair modules also revealed the strange behavior, previously demonstrated in our tests by Kingmax modules. Namely, both memory types allow any tRAS values in the configuration registers of the chipset, from 15 to 3, without any reaction. In this connection, we again have to content ourselves with the first three timing parameters only.
|
| Parameter* |
Testbed 1
|
Testbed 2
|
Testbed 3
|
Testbed 4
|
Testbed 5
|
Testbed 6
|
| Timings |
4-4-3
(1.8 V)
|
4-3-3
(1.8 V)
|
4-3-3
(1.8 V)
|
3-3-3
(1.9 V)
|
3-3-3
(1.9 V)
|
3-3-3
(1.9 V)
|
| Average memory read bandwidth, MB/sec |
4526
|
4539
|
4543
|
4749
|
4676
|
4690
|
| Average memory write bandwidth, MB/sec |
2128
|
2174
|
2168
|
2251
|
2336
|
2295
|
| Max. memory read bandwidth, MB/sec |
6430
|
6448
|
6456
|
6552
|
6465
|
6487
|
| Max. memory write bandwidth, MB/sec |
4267
|
4266
|
4267
|
4324
|
4267
|
4282
|
| Minimum Pseudo-Random Access Latency, ns** |
54.3
|
54.3
|
54.3
|
52.0
|
52.8
|
52.6
|
| Maximum Pseudo-Random Access Latency, ns** |
63.3
|
63.2
|
63.2
|
60.3
|
61.4
|
61.0
|
| Minimum Random Access Latency, ns** |
125.8
|
122.3
|
122.4
|
117.5
|
119.3
|
118.9
|
| Maximum Random Access Latency, ns** |
151.4
|
147.9
|
148.1
|
142.5
|
144.4
|
144.0
|
*the best results are in bold
**Block size – 16 MB
|
The data in this table clearly demonstrates the direct relation between the voltage and the minimal CAS# latency. So, DDR2 Corsair modules operate steadily with tCL = 3 and 1.9 V. But when it's reduced to nominal (1.8 V), the minimal stable value of tCL = 4. Thus, maximum operating stability of the memory subsystem and the minimal timings – 3-3-3 – are demonstrated by MSI (Testbeds #4, 5) and ASUS (Testbed #6) mainboards. Since it's impossible to control memory voltage in Foxconn mainboards, they demonstrate worse results – minimum 4-3-3 timings. And at last, the worst results, both by stability (4-4-3 timings) and by performance in normal and extreme modes are demonstrated by Intel D915GUX.
Mainboards based on Intel 925X
Performance tests
Among the mainboards based on Intel 925X, the first series of tests again puts forward the product from MSI – MSI 925X Neo mainboard (Testbed #10). The second place, with minimal differences, is taken by Gigabyte 8ANDXP-D (Testbed #7) and ASUS P5AD2 (Testbed #9) mainboards. Foxconn 925A01 (Testbed #11) is almost on the par with them and by some parameters (maximum real memory bandwidth) it even outscores them. The worst results (noticeably lesser real memory bandwidth and noticeably larger memory latency) are demonstrated by Intel D925XCV (Testbed #8).
|
| Parameter* |
Testbed 7
|
Testbed 8
|
Testbed 9
|
Testbed 10
|
Testbed 11
|
| Timings |
4-4-4-12
|
4-4-4-12
|
4-4-4-12
|
4-4-4-12
|
4-4-4-12
|
| Average memory read bandwidth, MB/sec |
4682
|
4527
|
4669
|
4717
|
4653
|
| Average memory write bandwidth, MB/sec |
2464
|
2432
|
2456
|
2461
|
2425
|
| Max. memory read bandwidth, MB/sec |
6412
|
6363
|
6400
|
6515
|
6422
|
| Max. memory write bandwidth, MB/sec |
4287
|
4267
|
4282
|
4327
|
4267
|
| Minimum Pseudo-Random Access Latency, ns** |
51.8
|
54.4
|
51.9
|
51.3
|
52.1
|
| Maximum Pseudo-Random Access Latency, ns** |
60.0
|
63.1
|
60.1
|
59.5
|
60.5
|
| Minimum Random Access Latency, ns** |
123.4
|
128.2
|
123.4
|
122.4
|
124.0
|
| Maximum Random Access Latency, ns** |
148.1
|
153.2
|
148.2
|
146.9
|
149.0
|
*the best results are in bold
**Block size – 16 MB
|
In performance tests of DDR2 Corsair modules with mainboards based on Intel 925X the first place is again taken by the MSI product – 925X Neo mainboard (Testbed #10). The second place, almost with the same result, is shared by three mainboards – Gigabyte 8ANDXP-D (Testbed #7), ASUS P5AD2 (Testbed #9) and Foxconn 925A01 (Testbed #11). Corsair modules demonstrate the worst memory bandwidth and latency on the Intel D925XCV mainboard, as was the case with previously tested Kingston and Kingmax modules. As before, the memory performance on mainboards based on i925X is on the whole higher than that on the i915P/G mainboards.
Stability tests
|
| Parameter* |
Testbed 7
|
Testbed 8
|
Testbed 9
|
Testbed 10
|
Testbed 11
|
| Timings |
3-3-3
(1.9 V)
|
3-3-3
(1.95 V)
|
3-3-3
(1.9 V)
|
3-3-3
(1.9 V)
|
4-3-3
(1.9 V)
|
| Average memory read bandwidth, MB/sec |
4886
|
4688
|
4871
|
4912
|
4681
|
| Average memory write bandwidth, MB/sec |
2486
|
2455
|
2470
|
2447
|
2394
|
| Max. memory read bandwidth, MB/sec |
6487
|
6435
|
6469
|
6577
|
6473
|
| Max. memory write bandwidth, MB/sec |
4287
|
4267
|
4282
|
4327
|
4267
|
| Minimum Pseudo-Random Access Latency, ns** |
49.5
|
51.7
|
49.6
|
49.2
|
52.0
|
| Maximum Pseudo-Random Access Latency, ns** |
57.9
|
60.0
|
57.9
|
57.5
|
60.4
|
| Minimum Random Access Latency, ns** |
113.2
|
117.4
|
113.5
|
112.5
|
117.7
|
| Maximum Random Access Latency, ns** |
137.5
|
142.1
|
137.6
|
136.4
|
142.6
|
*the best results are in bold
**Block size – 16 MB
|
As you can see from the table data, all tested mainboards based on i925X allow increasing the memory voltage (to 1.9 or 1.95 V and higher). Thus, the modules under review operate steadily almost on all mainboards, when the CAS latency = 3, except for Foxconn 925A01 (Testbed #11), where Corsair modules operate steadily only when tCL = 4, though the voltage is 1.9 V. Due to the increased CAS latency, this mainboard demonstrates memory performance results comparable to those of Intel D925XCV (Testbed #8).
Results
Our tests demonstrate that DDR2 Corsair modules are stable to operate in DDR2-533 mode with really fantastic 3-3-3 timings on most mainboards, which support the 1.9V memory voltage recommended by the manufacturers. The best results, both in performance and in memory stability, are again demonstrated by MSI (925X Neo and 915P Neo2), ASUS (P5AD2 and P5GDC-V) and Gigabyte 8ANDXP-D mainboards; the worst – by the Intel mainboards (D925XCV and D915GUX).
|
|
|
|
|
| | | | | | | |
| | | | | | |
|
|
|
|
