If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
Socket 771 Xeon in Socket 775?
I've been trying to do this using a 'modded' Xeon that I bought from
AliExpress (essentially has one of those stickers micro-glued in place and cut down so it's just over the two transposed lands and has a second set of grooves cut in periphery to fit socket). I've tried it in two mobos now, at least one of which is on the whitelist of mobos that will work with this mod and neither even get to BIOS. However when I read up on it I see instructions for unpacking BIOS, inserting CPU microcode, repacking and reflashing the mobo. Is this a "must do" step? If so it wasn't made clear and I'm a bit miffed. That sort of thing isn't my forte and I know how easy it is to brick a mobo by flashing a bad image. Cheers, -- Shaun. "Humans will have advanced a long, long way when religious belief has a cozy little classification in the DSM*." David Melville (in r.a.s.f1) (*Diagnostic and Statistical Manual of Mental Disorders) |
#2
|
|||
|
|||
Socket 771 Xeon in Socket 775?
~misfit~ wrote:
I've been trying to do this using a 'modded' Xeon that I bought from AliExpress (essentially has one of those stickers micro-glued in place and cut down so it's just over the two transposed lands and has a second set of grooves cut in periphery to fit socket). I've tried it in two mobos now, at least one of which is on the whitelist of mobos that will work with this mod and neither even get to BIOS. However when I read up on it I see instructions for unpacking BIOS, inserting CPU microcode, repacking and reflashing the mobo. Is this a "must do" step? If so it wasn't made clear and I'm a bit miffed. That sort of thing isn't my forte and I know how easy it is to brick a mobo by flashing a bad image. Cheers, As far as I know, "microcode" and "recognition" are two separate issues. The CPU support chart for your motherboard, shows the minimum BIOS revision to recognize a CPU. That means the BIOS uses the various special instructions for CPUID, and makes sure it knows what the processor is, what register to read for voltage or multiplier and so on. Microcode on the other hand, is for repairing bugs in the CPU design. It's possible a computer might still boot, if no microcode is loaded. The Intel PIU utility, the "revision" field is in fact the output of the current microcode revision loaded. At T=0, the revision field is 0. The BIOS microcode loaded, loads a 2KB segment into the CPU. The CPU verifies the microcode is for usage with that specific processor, otherwise the patch is rejected. Maybe after the BIOS is finished, the revision field is 0x04. (Intel is constantly releasing new revisions, when more bugs are discovered in testing.) Now, Windows boots. Windows has mcupdate, a microcode loader service. It runs early in boot, and attempts to load microcode. To be accepted, the microcode has to be a greater revision than the BIOS installed one. So now perhaps, Windows installed 0x07 revision microcode. If you use the Intel PIU utility at this point, it would report 0x07 while you're sitting in the Windows desktop. When some Microsoft service pack was failing, one of the symptoms was that the microcode revision was zero. Implying nothing was patching the CPU. Anyway, that's just a bit of history on microcode. It doesn't seem to be the same as the recognition problem. And the CPUSupport chart for the motherboard is where you would look for inspiration there. ******* To give another example of "recngnition" versus "microcode". I attempted to put a Tualatin into my 440BX board. I used a slocket to convert to S370 format. I changed the voltage regulator on the motherboard, to give 1.5V for the CPU (instead of 1.75V). OK, so when the computer starts, the BIOS reports I have just installed a "Pentium II", which is utter ********. It's actually closer to Pentium III era. So the recognition is failing, but nobody got hurt. I don't know if any OSes were relying on that identification for anything important. As far as I know, modern OSes run their own recognition code, duplicating the effort the BIOS put in. I used a separate tool to install Microcode. It turns out one of the BIOS companies, it has a BIOS call for installing microcode without disturbing the main BIOS code block. But both AMI and Award don't have this, only one of them implemented a microcode cache. You could load the cache manually, and the cache would remember the 2KB block, until the day comes that you change out the processor. And then the cache contents get wiped. I would work on the recognition problem first. While in theory, a processor with errata could die at BIOS level, usually they're bug free enough that they work in the BIOS. (Otherwise, the motherboard company would have a hard time releasing the first revision of motherboard.) The other possibility, is the CPU is not contacting the socket properly. I'm not up on this mod, and I don't know what needs to be disconnected or connected to make it work. I did a similar kind of socket mod, to isolate a BSEL signal on my $65 LGA775 board, and used the insulation off a piece of wire to cover one of the LGA775 pins. I did the research for that myself, so know what it's for. I don't know what needs modding to use that particular Xeon. No idea. Paul |
#3
|
|||
|
|||
Socket 771 Xeon in Socket 775?
Once upon a time on usenet Paul wrote:
~misfit~ wrote: I've been trying to do this using a 'modded' Xeon that I bought from AliExpress (essentially has one of those stickers micro-glued in place and cut down so it's just over the two transposed lands and has a second set of grooves cut in periphery to fit socket). I've tried it in two mobos now, at least one of which is on the whitelist of mobos that will work with this mod and neither even get to BIOS. Just to correct this; The machine I tried it in was a Lenovo M58p, the machine that's on the whitelist is an M58e. However when I read up on it I see instructions for unpacking BIOS, inserting CPU microcode, repacking and reflashing the mobo. Is this a "must do" step? If so it wasn't made clear and I'm a bit miffed. That sort of thing isn't my forte and I know how easy it is to brick a mobo by flashing a bad image. Cheers, As far as I know, "microcode" and "recognition" are two separate issues. The CPU support chart for your motherboard, shows the minimum BIOS revision to recognize a CPU. That means the BIOS uses the various special instructions for CPUID, and makes sure it knows what the processor is, what register to read for voltage or multiplier and so on. Microcode on the other hand, is for repairing bugs in the CPU design. It's possible a computer might still boot, if no microcode is loaded. The Intel PIU utility, the "revision" field is in fact the output of the current microcode revision loaded. At T=0, the revision field is 0. The BIOS microcode loaded, loads a 2KB segment into the CPU. The CPU verifies the microcode is for usage with that specific processor, otherwise the patch is rejected. Maybe after the BIOS is finished, the revision field is 0x04. (Intel is constantly releasing new revisions, when more bugs are discovered in testing.) Now, Windows boots. Windows has mcupdate, a microcode loader service. It runs early in boot, and attempts to load microcode. To be accepted, the microcode has to be a greater revision than the BIOS installed one. So now perhaps, Windows installed 0x07 revision microcode. If you use the Intel PIU utility at this point, it would report 0x07 while you're sitting in the Windows desktop. When some Microsoft service pack was failing, one of the symptoms was that the microcode revision was zero. Implying nothing was patching the CPU. Anyway, that's just a bit of history on microcode. It doesn't seem to be the same as the recognition problem. And the CPUSupport chart for the motherboard is where you would look for inspiration there. Interesting, thanks. Th eonly reason I was using the term 'microcode' is that's the term used in the guides I read for unpacking, modifying then repacking BIOS firmware to then flash to the mobo. ******* To give another example of "recngnition" versus "microcode". I attempted to put a Tualatin into my 440BX board. I used a slocket to convert to S370 format. I changed the voltage regulator on the motherboard, to give 1.5V for the CPU (instead of 1.75V). OK, so when the computer starts, the BIOS reports I have just installed a "Pentium II", which is utter ********. It's actually closer to Pentium III era. So the recognition is failing, but nobody got hurt. I don't know if any OSes were relying on that identification for anything important. As far as I know, modern OSes run their own recognition code, duplicating the effort the BIOS put in. Heh! I did a bit of that 'Tualatin in board designed for Coppermine' thing myself. In the end I was using a couple of those "Slot T" adapters for slot CPUs http://duhvoodooman.com/powrleap/lfl/slot-t_1.htm and I even had a 'Socket U' (?) thing made by the same company that enaled use in Socket 370. For ages my best result was a 1.1 GHz Tualatin Celeron running on a 133 MHz FSB for 1.46 GHz. At the time it outperformed the affordable P4s available. The I got a Pentium IIIs and that was great! That's one of the times I've seen a BIOS display "Unknown CPU at xxx MHz" then continue to boot and run flawlessly. I used a separate tool to install Microcode. It turns out one of the BIOS companies, it has a BIOS call for installing microcode without disturbing the main BIOS code block. But both AMI and Award don't have this, only one of them implemented a microcode cache. You could load the cache manually, and the cache would remember the 2KB block, until the day comes that you change out the processor. And then the cache contents get wiped. I would work on the recognition problem first. While in theory, a processor with errata could die at BIOS level, usually they're bug free enough that they work in the BIOS. (Otherwise, the motherboard company would have a hard time releasing the first revision of motherboard.) Thanks for the info. The other possibility, is the CPU is not contacting the socket properly. I'm not up on this mod, and I don't know what needs to be disconnected or connected to make it work. This is the CPU that I bought. http://www.aliexpress.com/item/Intel...377080130.html For reference a second-hand Q9550 is about NZ$100 if you can get one. As they're still a very capable CPU they almost never come on the market. I did a similar kind of socket mod, to isolate a BSEL signal on my $65 LGA775 board, and used the insulation off a piece of wire to cover one of the LGA775 pins. I did the research for that myself, so know what it's for. I don't know what needs modding to use that particular Xeon. No idea. I did that BSEL mod on a 775 CPU also. I used a tiny bit of sticky tape. So after trying the Xeon in a Dell Precision T3400 and the Lenovo M58p I dug out my only other spare Socket 775 motherboard, a Gigabyte G41M - Combo. I checked the CPU suport list http://www.gigabyte.co.nz/support-do....aspx?pid=3505 and it supports similar-specced C2Qs, even the 130w TDP versions (althogh the VRM section doesn't seem *that* beefy). I sat it on the workbench and connected up a PSU and a small panel with power and reset switches, power and HDD LEDs and a piezo beeper that I still have from my benchtesting days. I fitted a stick of RAM, put the Xeon in the socket, some paste and a cooler and fired it up. Imagine my surprise when it not only posted but correctly recognised the CPU as a Xeon E5462! Awesome. Now my only issue is that I don't have a case to put it in. I had a big clear out a while back when there was a chance I might have to move house (I rent). It physically fits the Lenovo case but the connectors for the front switches and LEDs aren't industry standard - as you might imagine. Also there's little in the way of airflow around the single HDD bay (one of the reasons it wasn't in use). Still I can work with it. At least I didn't waste my money on the Xeon and I just have rather more work to do than I'd hoped. Cheers, -- Shaun. "Humans will have advanced a long, long way when religious belief has a cozy little classification in the DSM*." David Melville (in r.a.s.f1) (*Diagnostic and Statistical Manual of Mental Disorders) |
#4
|
|||
|
|||
Socket 771 Xeon in Socket 775?
On 09/04/2015 12:31 AM, ~misfit~ wrote:
So after trying the Xeon in a Dell Precision T3400 and the Lenovo M58p I dug out my only other spare Socket 775 motherboard, a Gigabyte G41M - Combo. I checked the CPU suport list http://www.gigabyte.co.nz/support-do....aspx?pid=3505 and it supports similar-specced C2Qs, even the 130w TDP versions (althogh the VRM section doesn't seem *that* beefy). I sat it on the workbench and connected up a PSU and a small panel with power and reset switches, power and HDD LEDs and a piezo beeper that I still have from my benchtesting days. I fitted a stick of RAM, put the Xeon in the socket, some paste and a cooler and fired it up. Imagine my surprise when it not only posted but correctly recognised the CPU as a Xeon E5462! Awesome. Glad you finally found had some success with yours. I did the mod on my Gigabyte EP-DS3L last year, modifying both the socket and the BIOS, and I was successful in getting it to work with a Xeon E5450. I wasn't able to overclock it without pushing the voltages beyond what I was comfortable with, but I'm still happy with the extra two cores compared to the C2D that had previously. In case you didn't already have it bookmarked, there is a big honking thread about it at: http://www.overclock.net/t/1431723/m...a771-xeon-cpus , which has a lot of different input regarding different boards that people have found to work (or not work). I used the tools available from: http://www.delidded.com/how-to-updat...-phoenix-bios/ to modify my BIOS. I modded the latest BIOS available for my board, and yes, I was nervous as hell during the flashing process. Jon |
#5
|
|||
|
|||
Socket 771 Xeon in Socket 775?
Once upon a time on usenet Jon Danniken wrote:
On 09/04/2015 12:31 AM, ~misfit~ wrote: So after trying the Xeon in a Dell Precision T3400 and the Lenovo M58p I dug out my only other spare Socket 775 motherboard, a Gigabyte G41M - Combo. I checked the CPU suport list http://www.gigabyte.co.nz/support-do....aspx?pid=3505 and it supports similar-specced C2Qs, even the 130w TDP versions (althogh the VRM section doesn't seem *that* beefy). I sat it on the workbench and connected up a PSU and a small panel with power and reset switches, power and HDD LEDs and a piezo beeper that I still have from my benchtesting days. I fitted a stick of RAM, put the Xeon in the socket, some paste and a cooler and fired it up. Imagine my surprise when it not only posted but correctly recognised the CPU as a Xeon E5462! Awesome. Glad you finally found had some success with yours. Thanks Jon. I did the mod on my Gigabyte EP-DS3L last year, modifying both the socket and the BIOS, and I was successful in getting it to work with a Xeon E5450. I wasn't able to overclock it without pushing the voltages beyond what I was comfortable with, but I'm still happy with the extra two cores compared to the C2D that had previously. Yeah, the extra cores and L2 cache was my main goal. Actually the full story is that this second computer of mine is used mainly for a friend gaming. I'm an invalid who is largely unable to do physical stuff and she comes and stays for a couple of days every week or so and does some housework for me. In return I set her up with a mid-level gaming computer for her home and beefed up my 'spare' so she can play when she's here. Both my spare and her main PCs were ex-corporate Dell T3400s that ran E8400s. (My main PC is an ASUS P5Ke-WiFi AP / QX9650 / 8 G / 120 GB SSD machine, my last full build that was started in '07 with a lowly E4500 / 2 G with the intent of being upgradeable and durable - so far so good. Since then with my declining health and financial situation I've found it far better to buy ex corporate machines, there are real bargains to be had. However they're not very upgradeable and I run into issues with proprietary BIOS and components.) A while back I got the chance to buy a PC that was running a Q9550 very cheaply - cheaply enough that I could swap the CPU for the E8400 in my second machine and then on-sell it without loss. Since then the game we both play (Path of Exile), while not being hugely demanding of GPU (it's coded in DX9) has become more CPU-intensive - to the point where she was noticing my spare computer was running better than her computer despite her having an HD 7770 and my spare having a HD 5670. So, apprecative of the help she gives me, I promised her a Q9550. However I've been unable to source one that I can afford so went the modded Xeon E5462 route. Essentially an identical CPU as far as fab size, speed, FSB, multi and L2 cache goes. I was hoping to drop it into my Dell T3400 (a brick outhouse of a computer, extremely well built and reliable) and give her my Q9550 - job done. Now, as the Gigabyte mobo doesn't fit the Dell case it looks like I'm back in the system-building business - at least for my spare machine. In case you didn't already have it bookmarked, there is a big honking thread about it at: http://www.overclock.net/t/1431723/m...a771-xeon-cpus , which has a lot of different input regarding different boards that people have found to work (or not work). Thanks for that, I hadn't found it in my searches. I'll read it properly in a while. I used the tools available from: http://www.delidded.com/how-to-updat...-phoenix-bios/ to modify my BIOS. I modded the latest BIOS available for my board, and yes, I was nervous as hell during the flashing process. Thanks. Yes I found that site and, if I'd had to mod a BIOS I intended to follow instructions there. Thank the gods it doesn't seem like I'll have to now. Instead I just need to modify a case and find another install key for W7 64 bit and 'build' an OS the way I like it for this Gigabyte mobo. I originally put the GA G41M-Combo board aside as I don't like how hot the NB gets (far too hot to touch, even with the X4500 graphics disabled in BIOS) and it's hard (impossible so far) to fan cool it - it's a small board. Even with some CPU fan spill it's still uncomfortably hot. Despite the fact that I have a drawer full of salvaged CPU and NB heatsinks (that I mainy use for experimenting with high power LEDs) there's nothing better there that will fit. I may have to get the hacksaw and drill out and use some of the morphine I've managed to put aside for when the zombie apocolypse comes..... Cheers, -- Shaun. "Humans will have advanced a long, long way when religious belief has a cozy little classification in the DSM*." David Melville (in r.a.s.f1) (*Diagnostic and Statistical Manual of Mental Disorders) |
#6
|
|||
|
|||
Socket 771 Xeon in Socket 775?
Once upon a time on usenet ~misfit~ wrote:
Once upon a time on usenet Jon Danniken wrote: [snip] In case you didn't already have it bookmarked, there is a big honking thread about it at: http://www.overclock.net/t/1431723/m...a771-xeon-cpus , which has a lot of different input regarding different boards that people have found to work (or not work). Thanks for that, I hadn't found it in my searches. I'll read it properly in a while. Wow! "Big honking thread" is right! 7,839 posts! The Lenovo M58 is mentioned on page 1 and discussed for the first few pages - result is it's a no-go. I haven't done anything further to the board / CPU yet, pain is a hard taskmaster. However reading some of that thread got me thinking... I deliberately ordered a CPU with a low-ish TDP as I intended to put it in the Dell which only handles 80 TDP with the stock heatsink and the design of the system means it won't take an aftermarket cooler. (There was a heatpipe cooler option for it so it could handle Q6600s but they're like hen's teeth.) Now that I'm not using the Dell I rather wish I'd bought a Xeon designed for a lower FSB and higher multiplier, the Gigabyte board has overclocking abilities in the BIOS. That said the VRM section of the board isn't very impressive and I'm only running a generic 500w PSU (with two 12v rails, one for the 4-pin CPU power and the other for HDDs, other mobo, GPU etc.). I have a couple of honking great big Skt 775 heatpipe coolers collecting dust that *would* fit the Lenovo case. Oh well, at least the Xeon's working (still only in BIOS so far but it looks good). -- Shaun. "Humans will have advanced a long, long way when religious belief has a cozy little classification in the DSM*." David Melville (in r.a.s.f1) (*Diagnostic and Statistical Manual of Mental Disorders) |
#7
|
|||
|
|||
Socket 771 Xeon in Socket 775?
~misfit~ wrote:
Once upon a time on usenet ~misfit~ wrote: Once upon a time on usenet Jon Danniken wrote: [snip] In case you didn't already have it bookmarked, there is a big honking thread about it at: http://www.overclock.net/t/1431723/m...a771-xeon-cpus , which has a lot of different input regarding different boards that people have found to work (or not work). Thanks for that, I hadn't found it in my searches. I'll read it properly in a while. Wow! "Big honking thread" is right! 7,839 posts! The Lenovo M58 is mentioned on page 1 and discussed for the first few pages - result is it's a no-go. I haven't done anything further to the board / CPU yet, pain is a hard taskmaster. However reading some of that thread got me thinking... I deliberately ordered a CPU with a low-ish TDP as I intended to put it in the Dell which only handles 80 TDP with the stock heatsink and the design of the system means it won't take an aftermarket cooler. (There was a heatpipe cooler option for it so it could handle Q6600s but they're like hen's teeth.) Now that I'm not using the Dell I rather wish I'd bought a Xeon designed for a lower FSB and higher multiplier, the Gigabyte board has overclocking abilities in the BIOS. That said the VRM section of the board isn't very impressive and I'm only running a generic 500w PSU (with two 12v rails, one for the 4-pin CPU power and the other for HDDs, other mobo, GPU etc.). I have a couple of honking great big Skt 775 heatpipe coolers collecting dust that *would* fit the Lenovo case. Oh well, at least the Xeon's working (still only in BIOS so far but it looks good). Seeing a working BIOS screen, is a large step forward. http://ark.intel.com/products/33084/...z-1600-MHz-FSB Processor Number E5462 # of Cores 4 Processor Base Frequency 2.8 GHz TDP 80 W L2 Cache 12 MB FSB Speed 1600 MHz That's a pretty high FSB for your run of the mill LGA775 system. Not a lot of room to overclock it. It would be a treat if it was unlocked, but I doubt they'd do that to the average LGA775 processor. It probably doesn't leave a lot of room for SpeedStep either. When the machine is idle, the multiplier might be 6, and when flat out, the multiplier might be 7. Not a lot of room there. My E8400 for example, goes from 6x to 9x. Still not a lot of room. It would need to be unlocked, to be a "fun" choice. With that kind of FSB. Paul |
#8
|
|||
|
|||
Socket 771 Xeon in Socket 775?
Once upon a time on usenet Paul wrote:
~misfit~ wrote: Once upon a time on usenet ~misfit~ wrote: Once upon a time on usenet Jon Danniken wrote: [snip] In case you didn't already have it bookmarked, there is a big honking thread about it at: http://www.overclock.net/t/1431723/m...a771-xeon-cpus , which has a lot of different input regarding different boards that people have found to work (or not work). Thanks for that, I hadn't found it in my searches. I'll read it properly in a while. Wow! "Big honking thread" is right! 7,839 posts! The Lenovo M58 is mentioned on page 1 and discussed for the first few pages - result is it's a no-go. I haven't done anything further to the board / CPU yet, pain is a hard taskmaster. However reading some of that thread got me thinking... I deliberately ordered a CPU with a low-ish TDP as I intended to put it in the Dell which only handles 80 TDP with the stock heatsink and the design of the system means it won't take an aftermarket cooler. (There was a heatpipe cooler option for it so it could handle Q6600s but they're like hen's teeth.) Now that I'm not using the Dell I rather wish I'd bought a Xeon designed for a lower FSB and higher multiplier, the Gigabyte board has overclocking abilities in the BIOS. That said the VRM section of the board isn't very impressive and I'm only running a generic 500w PSU (with two 12v rails, one for the 4-pin CPU power and the other for HDDs, other mobo, GPU etc.). I have a couple of honking great big Skt 775 heatpipe coolers collecting dust that *would* fit the Lenovo case. Oh well, at least the Xeon's working (still only in BIOS so far but it looks good). Seeing a working BIOS screen, is a large step forward. http://ark.intel.com/products/33084/...z-1600-MHz-FSB Processor Number E5462 # of Cores 4 Processor Base Frequency 2.8 GHz TDP 80 W L2 Cache 12 MB FSB Speed 1600 MHz That's a pretty high FSB for your run of the mill LGA775 system. Not a lot of room to overclock it. It would be a treat if it was unlocked, but I doubt they'd do that to the average LGA775 processor. It probably doesn't leave a lot of room for SpeedStep either. When the machine is idle, the multiplier might be 6, and when flat out, the multiplier might be 7. Not a lot of room there. My E8400 for example, goes from 6x to 9x. Still not a lot of room. It would need to be unlocked, to be a "fun" choice. With that kind of FSB. I didn't want a 'fun choice' for this machine, just a relatively fast quad core. However it turns out this E5462 isn't it - the board won't run stably overclocked (anything over 333 / 1333 is an overclock on this board) so is currently running at 2.33 GHz. As the board seems fine with Xeons (BIOS correctly IDed it, Windows 8.1 installed and running well, Prime 95 stable) I've now ordered an E5440 - the CPU I should have got in the first place. http://www.aliexpress.com/item/INTEL...024572230.html (That's what happens when I spend most of the day thinking about whether to get a modded Xeon, then order one quickly last thing before bed. I didn't look into the details too much, just searched for one that had "Q9550" in the title. I've complained to AliExpress about the E5462 listing title being misleading but it was really my fault for not reading the 'fine print'.) -- Shaun. "Humans will have advanced a long, long way when religious belief has a cozy little classification in the DSM*." David Melville (in r.a.s.f1) (*Diagnostic and Statistical Manual of Mental Disorders) |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Quete CPU Cooler for Opteron Socket F (Socket 1207) w/ Tyan n6650W mobo | Kurt Kufeld | AMD x86-64 Processors | 0 | January 15th 07 03:27 AM |
Quite CPU Cooler for Opteron Socket F (Socket 1207) w/ Tyan n6650W mobo | Kurt Kufeld | Homebuilt PC's | 1 | January 15th 07 12:43 AM |
Pentium3 socket 370 sur Asus P3BF avec adaptateur (slot one-socket 370) | Perriot-Lebarq | Asus Motherboards | 3 | November 13th 05 09:20 PM |
Are Heat Sink fans for socket 754 AMD CPU's compatable with the socket 939 Motherboards? | Keith S. | Overclocking AMD Processors | 0 | July 4th 04 01:35 PM |
Are Heat Sink fans for socket 754 AMD CPU's compatable with the socket 939 Motherboards? | Keith S. | AMD Thunderbird Processors | 0 | July 4th 04 01:35 PM |