| ||||||
|
Be aware that I will ask all those noob questions that I can't seem to find reliable research on but here goes. I am building a computer and decided to go with G.SKill 8500 CL5D 4GBPI 8500 4GB (2x2) DDR2 - 1066 CLS 5-5-5-15 240pin Dual Channel Memory but they ran out at Ncix.com so I saw something else that looked like the exact same memory except for the other memory had 4GBPK it had a K instead of the I and one was more expensive then the other. I'd tell you which but they've been having troubles with the site and nothing is loading up lol. Otheriwse I know 4GB (2x2) mean two sticks of ram at 2GB a piece, DDR2 I have no idea, 1066 I think is MHz but not sure, CLS 5-5-5-15 is various clock speeds which I'm sure have something to do with overclocking but again not sure, 240pin I imagine is the connecting port. oh and CL5D... no idea what that means ![]() If anyone can answer these innane newb questions then I'm all ears |
| |||||||
|
DDR2 memory is a type of SDRAM. SDRAM stands for Synchronous Dynamic Random Access Memory. The memory is organized like a matrix or chart, with data arranged in rows and columns. The data is stored in blocks whose location are found by the coordinates of the specific rows and columns. Latencies come from the memory looking for the data in these series of rows and columns. The four most common latencies are:
Simply comparing memory latencies with considering the speed at which the memory is running those latencies is silly. This is because the overall latencies in nano-seconds is derived from dividing your total latencies in cycles by how many cycles your RAM can complete in one second. This gives you latencies per operation in seconds. For example: DDR2-800 does 800,000,000 cycles per second. Latencies of 4-4-4-12 add up to 24 cycles per operation of latency. Divide 24 cycles of latencies by 800,000,000 cycles and you get 30 nano-seconds worth of latencies per operation. However, DDR2-1000 with latencies of 5-5-5-15 also net you the same 30 nano-seconds of latencies per operation (30 / 1,000,000,000). However, even though both settings have the same latencies. DDR2-1000 @ 5-5-5-15 is better than DDR2-800 @ 4-4-4-12, this is because DDR2-1000 has more data throughput when compared to DDR2-800. see: Intel Overclocking Redefined: Guide to Successful Overclocking with NB Straps in Mind - Overclock.net - Overclocking.net
__________________ |
| ||||
|
i'll try by breaking down each part for you... i am assuming you are speaking of this kit... G.SKILL F2-8500CL5D-4GBPI PC2-8500 4GB 2X2GB DDR2-1066 CL5-5-5-15 240PIN Dual Channel Memory Kit G.SKILL F2-8500CL5D-4GBPI PC2-8500 4GB 2X2GB DDR2-1066 CL5-5-5-15 240PIN Dual Channel Memory Kit ++++++++++++++++++++++++++++++ G.SKILL = this is the manufacturer name F2-8500CL5D-4GBPI = this is the model number, like Focus is to a Ford Focus, it also designates specs as you will see shortly PC2-8500 = the "PC2-xxxx" is a way manufacturers identify the theoretical bandwidth of memory. PC2-8500 has a theoretical bandwidth of 8500MB/s. i think it is actually 8533MB/s but that doesn't matter. this is simply an identifier to group memory together, it really isn't necessary as it is redundant. 4GB = total amount of memory...4 GigaBytes 2X2GB = distribution of total amount of memory above...2x2GB equates to two 2GB modules DDR2-1066 = this is the important identifier. this is why the PC2-8500 is redundant, they say the same thing. PC2-8500 = DDR2-1066. DDR2-1066 is another made up term to identify memory speed. the 'DDR' portion stands for 'Double Data Rate'. this means that Data is transmitted twice during the clock cycle. so the memory actually runs at 533MHz (which is the memories actual frequency). but because data is transmitted twice with every clock cycle, it is effectively running at DDR2-1066 (533x2 = 1066). by labeling it DDR2-1066, they are identifying that the memory runs at an effective 1066 or DDR2-1066. this is why when memory running at DDR2-1066 shows up in CPU-Z as 533MHz, it is running at full speed. 533MHz = actual frequency, DDR2-1066 = marketing term to make memory look like it is running faster. but 533MHz = DDR2-1066. this is the tough part as there is a bit of back knowledge required to full understand. all you need to really know is DDR2-1066 = PC3-8500. it is the 'speed' of the memory. CL5-5-5-15 = memory timings. these are the main timings that the memory runs at. standard identification of memory is by these main timings. CL5 = Cas Latency 5. the others aren't really important but you will notice most DDR2 is either 5-5-5-15 or 4-4-4-8 or 3-3-3-8. these are common timing groups. tighter is better as it is faster, less latency, but as the memory frequency increases, timings have to loosen. it is a direct relationship. higher frequency (DDR2-1066/533MHz) = looser timings (5-5-5-15 VS 4-4-4-8 for example). often memory running at DDR2-800/400MHz will have 4-4-4-8 timings, most DDR2-1066/533MHz will have 5-5-5-15 timings. 240PIN = simply the amount of pins the DDR2 DIMM slot uses. all DDR2 is 240pin. Dual Channel Memory Kit = identifying that the memory kit is a two stick set of memory designed to run in dual-channel. again, pretty standard at this point in time. ++++++++++++++++++++++++++++++ so too answer your question about the two different kits... G.SKILL F2-8500CL5D-4GBPI PC2-8500 4GB 2X2GB DDR2-1066 CL5-5-5-15 240PIN Dual Channel Memory Kit G.SKILL F2-8500CL5D-4GBPK PC2-8500 4GB 2X2GB DDR2-1066 CL5-5-5-15 240PIN Dual Channel Memory Kit ...they are identical except for the bolded part. this is part of the model number and designates the heat sinks. that is the only difference. the slightly more expensive kit is the PI heatsinks, the PK are the normal baby blue heat sinks. that is all the difference is. i know that was long but you said you wanted to learn ![]() HTH = hope that helps |
| |||||||
| Maybe we're just getting in touch with a side of Jody that we weren't aware of
__________________ Please note: I am not here to provide any kind of official NCIX support on these forums. For faster (and official) service please contact me at Linus@ncix.com, or please contact our customer care team at wvvw.NCIX.com (Canada) or wvvw.NCIXUS.com (America) Heatware: http://heatware.com/eval.php?id=25647 |
| ||||||
|
Thank you to 3oh6 and fefox for a good explanation. I guess to that certain motherboards would utilize ram much better then others. I don't plan to do OCing right away once I build my system (first time build) but it'll be good to start piecing together what my computer may have the capability of doing. The system to be in question is this. Asus P5Q-pro ATX LGA775 P45 DDR2 2PCI-E16 Crossfire 3PCI-E1 2PCI SATA2 sound GBLAN eSATA motherboard Intel Core 2 Quad Q9550 LGA775 2.83GHz Yorkfield 1333FSB 12MB G.Skill 8500 F2-8500CL5D 4GBI PC2-8500 4GB 2x2GB DDR2-1066 CL 5-5-5-15 240pin dual channel (thanks btw for explaining all that meant) Western Digital WD1001FALS Caviar Black 1TB SATA2 7200RPM 4.2MS 32MB 3.5in Dual Processor Hard Drive OEM XFX Radeon HD 4890 1GB 850MHz 1GB 3.9GHz GDDR5 PCI-E 2XDVI HDTV Video Card LG GH22LS30 Black 22X SATA Lightscribe DVDRW OEM all in a nice Coolermaster cm 690 case, powered by a corsair 750W PS with a couple 120mm 3pin case fans to keep it cool, an extra heatsink may be in the works too. I know this all doesn't pertain to RAM but does the choice I made match the system reletively well? |
| |||||||
|
The RAM will do fine for the setup you picked out. Since the RAM you picked is rated up to 1066mhz, you shouldn't worry about the RAM not keeping up while overclocking the Q9550. Just to add onto 3oh6 regarding the timing part. Note very carefully that RAM timings are crucial to a good and stable overclock, while squeezing the most out of the RAM. Common sets are 5-5-5-15, 4-4-4-12 (or 8), and 3-3-3-8. You won't find it lower than that. 3oh6 already mentioned the relationship between timings and frequency, however not ALL RAM will do what you tell them to. It's best still to research on the tweakability of RAM before purchasing. However you made a good purchase with the G.Skills. Make sure you come back later, after you set up your system, and we can work on Subtimings.
__________________ Iapetos Q9450C1 3.2ghz 1.215v, DFI LP UT X48 T2R, Mushkin Redline Ascent 2x2gb 1066mhz cas5, Aeneon Xtune 1066mhz cas5, XFX GTX295, Enermax Revolution 85+ 1050W, Logi Illuminated, SS Ikari Water Loop: 2xMCP655 Parallel, EK Supreme LT, PA120.3 Audio: Shure SRH840, UE Triplefi 10vi Olive mod. UE Superfi3. Pro. http://bfbc2.elxx.net/sig/detail4/pc/Jackquelegs.png |
| |||||||
| The FSB of the CPU linked to the Frequency of the RAM via the MCH Strap (CPU/Mem ratio). As people increase FSB, the memory frequency goes up as well. People really like 1:1, which is FSB = Memory Freq. So for example, 400mhz fsb on a Q9450 would equal to 800mhz effective on the memory using 1:1 ratio. Therefore, purchasing 667mhz RAM would be a bad idea.
__________________ Iapetos Q9450C1 3.2ghz 1.215v, DFI LP UT X48 T2R, Mushkin Redline Ascent 2x2gb 1066mhz cas5, Aeneon Xtune 1066mhz cas5, XFX GTX295, Enermax Revolution 85+ 1050W, Logi Illuminated, SS Ikari Water Loop: 2xMCP655 Parallel, EK Supreme LT, PA120.3 Audio: Shure SRH840, UE Triplefi 10vi Olive mod. UE Superfi3. Pro. http://bfbc2.elxx.net/sig/detail4/pc/Jackquelegs.png |
| |||||||
|
Well i certainly want to thank all you guys !!1 That was really a good explanation of ram to us laymen :) :) , I still don't know how to over clock myt gskill but i learned alot from this write up !! +1 to all you guys:)
__________________ I use paypal only.:) |
| ||||||
|
May I join in and ask a question? I have 3 sticks of OCZ Gold PC12800 DDR1600 2GB. If I run it at 1600MHz, I get a bandwidth of 16889MB/s. If I run it at 1520MHz, I get the same bandwidth, and in both cases, it does not matter if it is CAS8, or CAS9 0r CAS10. If I run it at 1200MHz, I get bandwidth of 15510 MB/s. These bandwidth are much higher than the rated PC12800, and it seems to max out at 1520MHz. But, I have not overclocked the ram to beyond the rated DDR1600, nor below the rated CAS8. So, how is it that I am getting much better bandwidth than the rated value? Does it mean that I can run it at 1520 MHz and it will perform just as well as at 1600MHz, with the benefit of lower ram voltage? The bandwidth is reported from memtest86+ v2.11
__________________ ___________________ Intel i7 720 D0 3.8GHz @1.225v QPI 1.30v OCZ Vendetta 2 Gagabyte EX58-UD3R OCZ pc3-12800 DDR3 gold xtc 6GB @1600 MHz 1.64v CAS 8-8-8-24 1T WD Caviar Black 640GB Corsair TX650W Powercolor HD4850 with fan replaced by 80mm panaflow CM690 case (1 extra case fans) WIN7/Linux Debian dual boot |
| |||||||
|
Check with Everest Memory benchmark.
__________________ Iapetos Q9450C1 3.2ghz 1.215v, DFI LP UT X48 T2R, Mushkin Redline Ascent 2x2gb 1066mhz cas5, Aeneon Xtune 1066mhz cas5, XFX GTX295, Enermax Revolution 85+ 1050W, Logi Illuminated, SS Ikari Water Loop: 2xMCP655 Parallel, EK Supreme LT, PA120.3 Audio: Shure SRH840, UE Triplefi 10vi Olive mod. UE Superfi3. Pro. http://bfbc2.elxx.net/sig/detail4/pc/Jackquelegs.png |
| ||||||
|
Thanks. Everest bench comes much closer, still about 120 MHz higher than rated.
__________________ ___________________ Intel i7 720 D0 3.8GHz @1.225v QPI 1.30v OCZ Vendetta 2 Gagabyte EX58-UD3R OCZ pc3-12800 DDR3 gold xtc 6GB @1600 MHz 1.64v CAS 8-8-8-24 1T WD Caviar Black 640GB Corsair TX650W Powercolor HD4850 with fan replaced by 80mm panaflow CM690 case (1 extra case fans) WIN7/Linux Debian dual boot |
![]() |
|
| Tags |
| 4gbpi , 4gbpk , g.skill , question , ram |
| Thread Tools | |
| Display Modes | |
Similar Threads | ||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| New OCing Numbers. | Chromey | Overclocking, Tweaking and Benchmarking | 8 | June 22, 2009 11:06 PM |
| Anyone Have Numbers on the 4890 Yet? | Soultribunal | HardwareCanucks F@H Team | 8 | April 15, 2009 08:04 PM |
| batch numbers | Rapid-X | Overclocking, Tweaking and Benchmarking | 8 | May 12, 2008 10:29 PM |
| Solid Barcelona numbers and dates?? | Babrbarossa | Rumor Mill | 4 | June 27, 2007 02:40 PM |
| Some interesting numbers in current events | Babrbarossa | Off Topic | 5 | May 23, 2007 05:15 AM |