DFi UT x48-t3rs, X3360
This is a discussion on DFi UT x48-t3rs, X3360 within the DFI Intel Motherboard / CPU forums, part of the Intel motherboards / CPU category; So, I have tested the RAM with very low voltage (1,5v) which is supposed to be JEDEC standard for Elpida ...
![]() |
| | LinkBack | Thread Tools |
| | #129 (permalink) |
| Senior Member Join Date: Nov 18 2009 Location: Bratislava
Posts: 278
|
So, I have tested the RAM with very low voltage (1,5v) which is supposed to be JEDEC standard for Elpida Hypers MNH at 1600CL8. Intel X48 can complain bout it, but it allowed me to test Skew Rates for memory. Dram Modes when system was able to bootup were 1, 6 and 7. The data strengts are just have to be found, but it will be similar to the OCZ memory this time. Currently I had to set quite high voltage for Northbridge. 1,4. pretty bad since intel reccomends 1,375v as maximum. I will read some documentation about 90nm process. I think that the manufacturing limitation (same as for 90nm cpus) could be useful here. At all most of the unstability has been just supressed, but not removed. Usually the system crashes upon turning off. |
| | |
| | #130 (permalink) |
| Senior Member Join Date: Nov 18 2009 Location: Bratislava
Posts: 278
|
There were some Errata update for X48+ICH9 X48 is not as much updated 1. Dont use PCI-E 1.1 cards in 16x slots. might cause trouble 2. Cold boots on all bearlakes - was referred to overclocking, but thats not true. The problem is caused by Voltage Regulator for PCI-E and DMI. On all boards after 2008 should be fixed. On this particular board seems that its fixed ![]() 3. PCI-E or DMI may reset-freeze during some specific "meta-stability condidion". Not very good to said, but it means that graphic card or southbridge will be "disconnected" if some configuration will be applied. The ICH9 Errata contain 16 separate errors. Last update was on march 2011 just in quick - USB migh give more power to USB device that is maximum by design. - dont use SATA6 drives on this board "D - and try to avoid sleep modes. |
| | |
| | #131 (permalink) |
| Senior Member Join Date: Nov 18 2009 Location: Bratislava
Posts: 278
|
Ok. My cross testing of X48 is finished. 1. CPU Using quadcores causes that each of the two physical chips access memory only with half bandwidth as socket allows. CPU as single device can access total bandwidth... seems to be more connected with cache memory as with cores. At all this does not lower the CPU performance. Also each quad core from this generation will wear CPU VRM and FSB as two separate processors which of course lowers the overclockability. 2. Memory Chips able to run 1333 cl 5-5-5 with 4gb configuration will work on 8gb config only as 1333 cl 8-8-8. Higher frequency will require more NB core voltage (not the memory voltage). 3.PCI-E 32 lanes at PCI-E 2.0 standard is guaranteed. Each PCI-E 2.0 device can access memory independently on host (CPU). This is something that event newer chipsets have not reached. Crossfire, or maybe faster SSD drives still have possibilty to use their bandwidth. Any cooperation between these two devices is then "hardware accelerated" without need of cpu or software. Advantages of the X48: - PCI-E design - Possibility to run DDR3 at speeds which can feed up the CPU and PCI-E independently - quad core support - 400Mhz can be reached without any voltage addition to NB for 4gb memory. - quite high utilization of CPU socket bandwidth Disadvantages - only 8gb of ram - Terrible support for all banks populated. - 90nm manufacturing process (same as infamous Pentium 4) |
| | |
| | #132 (permalink) |
| Senior Member Join Date: Nov 18 2009 Location: Bratislava
Posts: 278
|
I did some tests this day. If FSB is set to 333 I can set all voltages to defaults, including NB Core even when all memory banks are full and running on 1333Mhz. Sad to say but X48 with 8gb ram has decreased its basic FSB 400 to 333Mhz. On the other hand its really nice to see that intel had really trouble implementing DDR3. From my point of view X48 is one of the few chipsets of its generation which fully comply to basic specification of its generation with only one exception - base frequency should not be 400 but 333. On the other hand, the board (no matter how) MUST be able to run on 400Mhz while chipset is not pumping more than stock voltage and with memory on 1600Mhz. However my experience shows that its not possible. |
| | |
| | #133 (permalink) |
| Senior Member Join Date: Nov 18 2009 Location: Bratislava
Posts: 278
|
400Mhz FSB with 8gb at stock NB core voltages is not possible even with 1,9v on chips. Those values are "stock" for the chipset, but too much for the memory. Therefore the ram will crash (i believe due overheating). Currently running config: NB 1.37v RAM 1,736v (raised from 1.712v because of dropping below stable voltage) Skew 1-1-6-6 (timings nominal) 1600 CL8 CR1 Measurement with Linx shows that 1600 Cl 8 CR2 is vice-versa 1600CL7 CR2, which is still not a problem on performance side. However currently no benefits of larger ram are visible. Possibly due the lack of software available. will see in next 5 years... |
| | |
| | #134 (permalink) |
| Senior Member Join Date: Nov 18 2009 Location: Bratislava
Posts: 278
|
Yesterday: Random BSODs, disk drive not found. Today: Copper heatsink slipped off the Jmicron ATA controller. The plastic adhesive tape has melted due high temperature (around 50 degrees). How in the hell can disk controller generate such heat... Ok, the ambient temperatures in the computer are quite high. Harddrive is dead... |
| | |
| | #136 (permalink) |
| Senior Member Join Date: Nov 18 2009 Location: Bratislava
Posts: 278
|
No. I put copper heatsink on chip directly over sign "Intel ICH9R". Thats the JMicron chip. Current updates on system: Bought OCZ Revodrive X2 100GB After one day of total desperation I found correct Bios setting and placed the card to the PCI-E slot B at 4x/NC config. PCI-E slots A and C have issues with Revo on this board. Memory, chipset and CPU has been restored to default values, since the card made system more vulnerable to bad overclocking. New windows 7 installation placed on the Revo. Onboard SATA switched to AHCI mode, Velociraptor still holds "backup" copy of system and its primary used for overclocking. Trying older Arctic cooling 550 PSU which had better results with the system despite lower voltages delivered. Goal: FSB400 CPU 3400 RAM 1600 CL 8-8-8-x 8gigs Stock settings are also pretty impressive to me. Difference between goal settings and stock is 20%. Measured by LinX at 771Mb problem size its 38 VS 48 mflops. However no real impact on any application has yet been observed. At all stock settings give really nice temperatures - 60 degrees on CPU at full load. Other new hardware: Avermedia 707 trinity - TV tuner, with DVB-T and DVB-S including decoding module. I want to watch HBO HD Sits fine in PCI-E Slot APlanned hardware: Single slot graphic card - something like this" PowerColor HD6850 1GB GDDR5 Single Slot AX6850 1GBD5-I2DH. Last PCI slot is currently covered by graphic cooler, but I might require it for my SCSI controller since I own older SCSI Canon Scanner. |
| | |
![]() |
| Bookmarks |
| Tags |
| dfi, x3360, x48t3rs |
« Previous Thread
|
Next Thread »
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | |
Similar Threads for: DFi UT x48-t3rs, X3360
| ||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Initial notes, findings and tips DFI UT X58-T3EH8 | eva2000 | DFI Intel Motherboard / CPU | 0 | 04-08-2009 03:17 PM |
| [Bjorn3d.com] DFI Lanparty LT X48 T3RS | News | Reviews & News Online | 0 | 09-12-2008 07:41 AM |
| DFI UT X48 T2R - C1 Error | sunjovasie | DFI Intel Motherboard / CPU | 3 | 19-09-2008 12:17 AM |
| [Anandtech.com] DFI X48 LT T2R: Floats like a Butterfly… | eva2000 | Reviews & News Online | 0 | 11-05-2008 01:14 AM |
| DFI STREET IS NO MORE ++++ NEW DFI CLUB FORUM IS UP!!!!!!!! | dinos22 | General Hardware | 52 | 26-01-2007 09:17 PM |
All times are GMT +11. The time now is 03:12 PM.

















