Jump to content

scyphe

Members
  • Content Count

    242
  • Joined

  • Last visited

Everything posted by scyphe

  1. Thanks for the nice story (was it copy/pasted from another post about the RHT?), and my neck definitely creaked since I'm an oldtimer as well and jumped the train when Athlon arrived. Before that I was all Intel, overclocking anything from my first Pentium (P75) to celerons (both P2 and P3-based). Anyway, I admire Intel for getting their thinking cap on after the abysmal P4 (although some of them were nice overclockers) and returning to what worked instead of forcing a heavy and overblown architecture out on the market. The Conroe WILL be a real sweet deal for many many people, both thanks to the actual hardware and also thanks to the reinvented marketing machine @ Intel. I mean, they're out to make money, and they've lost a lot of market to AMD, and now they've charged their capacitors to the max, ready to retake lost territory. However, this has got nothing to do with what I initially posted which, to be honest, is a rumour from theInq. What it says though is that Intel will not support SLI on their own chipsets, they'll just support Crossfire. There will most likely be third party fixes for that, but the fact is, if the rumour is true, that Intel is using it's position to favor ATI, and there can be several reasons behind that, number one being that nvidia have stolen a lot of market from Intel on the chipset front ever since they went Intel. It's the kind of jabs that looks ugly, and I just thought people would like to know about it. In case there will be trouble using SLI on the new Intel chipsets they might be better off running nforce-based boards for Conroe's. I just recieved my premium lapping kit so I'm gonna prepare for a long, arm-killing session of lapping my Freezer 64 Pro (with groves and ridges on the area deep enough to be used as a sand shovel... )
  2. From theinquirer.net: WE GOT some interesting input on Intel not supporting SLI on its Conroe-supporting motherboards. Up until now, we, and our, sources have blamed it all on Nvidia but it turns out we might be wrong about that. There is a big chance that Intel doesn't want to allow Nvidia to run SLI, as it doesn't want to make a big thing about SLI on its own Conroe machines. Intel is in the bed with ATI and signed a cross-licensing agreement quite some time ago, the upshot of which is that it officially supports ATI's Crossfire on Intel 975/965 boards. We hear that Intel doesn’t want to let Nvidia's SLI to run on its boards, at least not officially. Wow.... talk about burning bridges...
  3. I don't know about you guys but if I bought a brand new board and it was broken I'd expect to recieve a brand new one in return, not a second hand refurbished one... Never had a problem with that so far though when it comes to motherboards. The only refurbished things I've gotten back so far was a replacement drive for a dead IBM (yes, the deathstar). However, I've never had to send in a board to the manufacturer, I've just returned it to the retailer and get a new one back. Whatever they do with the faulty equipment I don't know.. They probably just return them to their distributors. I don't care though, as long as I don't have to RMA it myself.
  4. I won't go one way or another until the hardware is out there, not hyped on paper with selected tests and comparisons. When the Conroe has been in the stores for a month or two (mass availability) and reviews have become mature and when AMD can show their RHT actually makes a big difference, and taking into considerations other things (like price to change platform once again) I may feel inclined to go one way or the other. Right now though it's all just interesting reading until the products are available. But it's good for business with all early adopters. When they've learnt about the system and performance, they'll write up on it and then go somewhere else if they're not happy with it.. Great for people with $$$ to spare. I guess what I mean is that I won't go AM2(RHT) or Conroe until both are mature and proven. And at the moment they're not. Besides, I haven't had my Opty that long and I can't afford to switch platform every ~3 months with new CPU, RAM etc...
  5. Almost, I got it off warp2search.net that in turn got it from theinquirer.net. I've actually heard about RHT before but had no idea when it would surface. My guess at that time was K8L, but it will be interesting to see if RHT will perform as we all hope it will.
  6. X-Fi + nForce4 = trouble... and Creative Labs is still working on a complete solution (which includes bios upgrades since it's actually a timing problem). They've released new betadrivers that may solve most issues for users.
  7. Edit by ExRoadie link goodness... http://theinquirer.net/?article=32589 As we all have seen, people all over the net believes the Conroe will wipe the floor with AMD's AM2-CPU's (mainly because the CPU's didn't perform differently compared to their s939-heritage). However, AMD have a neat trick up it's sleeve that it will unleash by the time the Conroe is released worldwide. It seems that all AM2 CPUs were outfitted with a support for Reverse-HyperThreading, an architectural change which enables software to think that it is working on a single-core alone. By combining two cores, the company has been able to produce the six IPC "core" that will go head to head against four IPC "core" from Conroe/Merom/WoodCrest combo. It seems that in certain cases, even an old AMD Athlon 64 3800+ can wipe the floor with Core 2 Duo E6300 CPU. In single-treaded apps, Core 2 Duo is expected to struggle against Reverse HyperThreading CPUs, which work at higher clock frequencies and produce higher instruction per clock ratios (IPC). AMDs Reverse-HT is a dynamic technology, and with Microsoft's Windows update and a new processor driver, the driver will copy the graphics drivers of today's 3D accelerators. The driver will detect the app, see if it is multithreaded or not and turn the ReverseHT on, or leave it off. Talk about an added boost to performance in 99.9% of the games, suddenly gaining almost 50% extra performance (from the second core running in "singecore-mode" with the first core). It will be a nice thing to have even if the feature won't beat the Conroe to a pulp (waiting for the K8L for that). All in all, pure goodness for all of us. Now, think about this for a moment, in the midst of this Conroe-craze where everyone seem to have bought into the massive hype-campaign Intel have managed to create (yes, you want a Conroe, you believe it will be the best even though you haven't even seen one in real life.. power of marketing hype)... With a DualCore AM2 CPU (Opteron 12xx for instance) running in RHT-mode, and it's overclocked to 3Ghz (just like 165's begs to be clocked to). Where does that leave us? It leaves us with a CPU: * 6 IPC's (compared to Conroe's 4) * Running at a higher frequency than the Conroe's (stock) What does that mean? It means we should wait a minute before jumping trains for the next big thing. 99.5% of the apps out there are STILL mainly single threaded (esp. games), and a dualcore AM2 CPU running @ ~3Ghz in RHT-mode will slay... read again: SLAY!!!! SLAYER m/
  8. Holy hell, 1.6+ volts. I won't use voltages over 1.5v on air. And as has been stated, most CCBWE's are statistically duds. CCBBE's are the current sweetheart among Opteron's. CCB1E's comes second. Personally I doubt I'll ever get it past 2.7Ghz with my current system. I have to push the voltages to 1.5+v and dual prime fails on core #0 after a few minutes when the temp reach and surpass 52 degrees. I'd need to improve cooling considerably to get 2.8-2.9 out of my CCB1E. Just a bad core #0 I guess. But I'm quite content with a rock stable primed 2.7Ghz. Oh, in case it has been omitted: do NOT attempt to overclock RAM and CPU when trying to find the max. You'll just end up not knowing which part breaks the stability. Keep RAM underclocked with relaxed timings at all times while trying to find the max stable CPU O/C.
  9. Well, there's no added pricecut apart from what's been announced already. The 1mb L2 dualcore cpu's will just vanish from the channels. Here's the full pricing chart btw.. with the announcement of retirement of the 1mb cache cpu's, the chart makes a lot more sense now with the lack of price reduction on the existing 1mb cache cpu's. AMD's price chart
  10. Do what? He's already removed the IHS and is running watercooling and have managed to drop temps to 42-45c. What else is there to do except going phase/dry ice?
  11. I don't think you could drop it. To be honest, those temps are dandy seeing as 165's are dualcore and produce more heat than their singlecore contemporaries, esp. seeing as you're overvolting and overclocking. I dualprime @ 48c but that's on air. All in all you've got a sweet setup.
  12. It's a golden time for Opty 165+ CPU's seeing as AMD is retiring all their desktop dualcore CPU's with 1mb L2-cache per core, leaving only 512mb-L2-per-core. The Tech Report - AMD Confirms X2 chips with 1mb L2 cache TOAST Of course, the same goes for the current X2's that have 1mb L2 per core. If you want to get a dualcore with 1mb L2 cache per core you'd better be quick about it. The only CPU left is the FX62 and most people would have to sell their houses, cars and souls to be able to buy that. Well, there's always Opterons (for s939) :nod: :nod:
  13. Here's my current setup, dual-primed. When I've gotten rid of my übercrappy freezer pro (the ridges on the copper base are just ridiculous.) and put a BT on it I'll up the voltage. ps. I had some weird RAM divider set in BIOS. Fixed after I noticed, didn't fancy another long prime just for that.. They're not O/C ram anyway so I'm running them @ 200Mhz.
  14. I ran an 11+ hour dual prime the other day, no problems. And RAM is running @ stock speeds. Anyway, it's a non-issue now. Both NIC's disabled in BIOS and Realtek 8139 NIC is running just fine. It's just mysterious that they both started corrupting downloads at a point where I wasn't doing anything in particular.
  15. Update: I just put my realtek NIC in and tried downloading the same file I've failed 8+ times with the marvell & nvidia onboard NIC's. And it worked flawlessly the first time. Conclusion? Something with the board. No idea what may have happened.
  16. I connect straight to the wall (CityLAN 100mbit tp), no routers or modems used I don't use a firewall (disabled) I didn't install nVidias network software (ActiveArmor etc.) No ad/spyware found I've tried using both Firefox and IE, same result It doesn't matter how "far away" the file is. I tried downloading a file from the CityLAN server (2 hops) and it was just as corrupted as the 91.31 drivers @ nvidia.com While downloading the 84.21 drivers from nvidia.com: if it's not the onboard hardware I have a second guess: my cats. they've been known to chew on network cables but haven't seen them doing it for quite a while now. btw, the cable they chewed on isn't used now. if it indeed is the cable, I'm gonna use a brush and soak the whole thing in tabasco.
  17. Wow, this came out of nowhere. Started last night for no apparent reason. I've used the Marvell until today, worked fine all the time until last night when every damn downloaded exe/zip was corrupted. I tried downloading other drivers but they were corrupted (!!). I tried the drivers that came with WinXP, same crap, file corrupted. I had an old driver version that I downloaded previously. Same thing.. Today I enabled the nVidia LAN and damn me if it doesn't do the same damn thing. Every exe/zip I download is corrupted. Meanwhile there's no problem for me browsing and stuff. It's just when I download a file of some sorts that it's corrupted. :confused: :confused: :confused: :confused: :confused: :confused: Some circuit on the board burned or something that killed both LAN-ports? Good thing I have an old Realtek 8139-based NIC lying around.
  18. Yeah, ebay is pretty much the only place you can buy known steppings.
  19. Nice one. I've got the same stepping. My core #0 seems to be touchy about speed over 2.7 though. 2.7 is rock stable but anything above will give errors in prime95 sooner or later. I haven't pushed the voltage past 1.44 though. Seems my core #1 will go anywhere though. Wouldn't surprise me if it passed 3.0Ghz without breaking a sweat. Pity there's no support for asynchronous clocking...
  20. AC Freezer 64 Pro, finish on the base: 0 out of 5.. It's the worst I've ever seen, and I've seen my fair share of HS bases. It's got deep ripples after some circular machining. Seems it's not lapped at all, just cut off and used like that.. And it was a classy lady to mount correctly on my Ultra-D. I had to move the RAM to the yellow slots (thankfully my RAM works in those), take the clipping mechanism apart and insert it in a 90-degree angle, try to get the clip to settle on those two plastic miniature pins (drove me mad) and then try to get the odd end to fasten. Definitely not the easiest kit to use on the Ultra-D. Removed the stock "goo" ("" because it was dry) and replaced with AS5 (which means I won't see the real results until at least 4-5 days of use). Not too happy though, the PWMIC-area gets very little cooling now and sits @ 55C. CPU is averaging 47-48C @ 1.39v. Too high for me. May get a Big Typhoon after all. Pity they're 1+ month away. Seems they go like a warm knife through butter. ps. I first installed it in the only way I could with the clip in default position (fan pointed downwards) with the stock goo. PWMIC reached 61C and CPU reached 52-53C so simply by changing direction where fan was on the right side this time, and using AS5, I dropped PWMIC-temps by 6 degrees and CPU by 4-5C. Still, temps are way too high. Gotta get another case with some clever solution to get a good airflow.
  21. Sweet chip. Seems I can't go past 2.7 on my CCB1E, core #0 drops out in prime95 after ~5-10 minutes no matter what voltage. Core #1 seems to go higher than hell but that's no good if both cores can't go there. :/
  22. This is NOT good.. The PWMIC is hitting 61C... Time to open the can and add another fan.. somehow.. no idea where and how to attach it.. :/ I upped the voltage to 1.440v (+ ~0.05v that CPU-Z usually miscalculates) and temp on CPU & PWMIC increased. CPU is past 50 and PWMIC is past 60. Hmmmm... The temps can be explained by a brand new Freezer 64 Pro with stock thermal goo (crappy) that have barely begun to work yet. The PWMIC though.. 61 degrees.. :eek:
  23. Intersting, I just replaced my old slk-948u with the AC Freezer 64 Pro. Due to the construction of the Ultra-D I had to fit it so the fan points downwards. That means there's no air on the PWMIC-area. The result is that the CPU is cooled and kept at ~46C (put the sink on 15 minutes ago and had to use the stuff already on the copperbase since I was all out of a cleaning agent. I'll replace with AS5 asap, right now I think it'll take a week to get good transfer going with the default grey goo from AC). The PWMIC reaches a staggering 54-55C. When the slk948u was on it topped @ 46C. Guess I have to find a way to fit a fan in that area...
  24. :confused: I'd take the 146 any day over the 3700 if the price was the same. If you want a stable stock chip, go Athlon64, if you want great O/C, go Opteron. Well, that's my experience anyway...
×
×
  • Create New...