>The big surprise here is that none of the current chipsets
>for Athlon 64 have a working PCI lock, whether they claim
>one or not.
Great, hadn't read that. Okay so scratch PCI lock on the K8T800, even though I'm not sure this is a chipset issue, or a problem with that particular board. Anyway, like I said, for people who don't overclock, or only overclock a bit for fun, this is not a huge issue.
>Really now. HMMM guess I have been extremely lucky on every
>single NF2 and i865pe build I have worked on
Then I'm more unlucky than you. I recently built a nForce2 based system for my brother, using a uATX board from Asus (don't recall exact type, one with IGP) and a Barton 2500+. I added two 256 Mb DDR400 modules from Apacer, since they where the same price as DDR333. Guess what ? Windows wouldnt install. BSOD and reboots all the time. Took me a while to figure it out, since the BIOS of that damn board didnt even allow CAS latency tweaking, it ran the DIMMs at SPD speeds, and they wouldnt work at 400 MHz I ended up solving it by swapping the memory modules with my old 333 Mhz one's, and using the apacer modules in another nForce2 system (that does run them @400 MHz, also Asus board BTW, A7N8X2.0).
>And Corsair not working on i865pe, is not at all a worry,
>almost every single review I have seen for the IS7 uses
>corsair XMS for their test setup. Works for them, works for
>me, I don't see the problem.
I've not seen many reviews of A64 boards having any trouble running quality DDR400 RAM either. Really, the THG article you linked to was interesting, but without further investigation, I wouldnt draw definite conclusions. Did those modules work 100% stable at SPD speeds on other boards with nForce2 or i875 ? We don't know. I linked an Anand article showing even several 865 boards had trouble running branded DDR400 memory modules without BIOS tweaking, so what makes you think this really is a A64 chipset issue ?
Where those DIMMS listed on as being recommended or qualified by the MB manufacturer ? Not likely. Did they work on the A64 after altering CAS /precharge etc setting manually ? I bet most would.
Also, did they even bother to install the latest BIOS ? It looks like not; I quickly checked for the Epox board, which was tested with a bios dated 09/29/2003. A quick check on Expox' website reveals there was a new BIOS released on 12/02/03, more than a month before the test.
The Gigabyte board was tested with a BIOS dated 10/17/2003, two new revisions have been released since, one well before publication date of the article
The MSI K8T Neo was tested with a BIOS dated 10/24/2003, ever since, two new bioses have come out, one specifically states 'Fixed Corsair stability issue.'
Anyway, you get my point. Those boards did not have recent BIOS's, even when in many (maybe even ALL ? I didnt bother to check more than 3) cases newer versions where available, and if not then, at least today each and every board tested has a couple of fresh bios releases available which I am fairly certain will address most if not all problems witnessed. I really have to wonder why THG didnt try the latest BIOS for these kind of testings...
So what is the problem; you either have to select qualified DIMMs, or if you're unlucky, maybe manualy tweak BIOS settings, or perhaps flash your BIOS to the latest rev. If that is the price to pay not to have to wait a few months for "better" boards, i'll gladly pay it. I'd flash the BIOS anyway, and tweak & test BIOS settings either way as well. If you're buying new RAM, picking up qualified modules sounds like a good idea as well.
>No not really. P4"C"'s on i86pe or i875p blew people away
>both for the motherboards performance/features, the memory
>bandwidth, and the 800 bus P4's performance. The A64 boards
>have not had that kind of impact.
Hu ? You found the increase in performance of A64/FX not stunning ? The extra features like 64 bit support, Cool&Quiet not all that impressive ? By comparison the upcoming chipset revisions with (maybe ?) PCI/AGP lock and integrated SATA instead of off chip SATA seems like nothing to hold your breath for.
>P4 and A64 are arguably equal performers, both good in all
>areas, but excelling in different areas. Not at all the
>same
First, we are discussing in a thread about someone wanting to build a fast gaming rig. MP3 encoding performance of the P4 seems pretty irrelevant. Secondly, its true that P4 was further ahead last year than A64 is today. But who cares ??
>IMO, the reasons are there, but you ignore what you see and
>still push A64 today as the only choice
Its the obvious choice for a gamer now.
> and jump on anyone who recommends someone who asks for the
>BEST motherboard waits a couple weeks for socket 754
>NF3-250 reviews and boards to come out. It's close, stop
>assuming the wait is months away
Just putting it in perspective; people always recommend to wait, and wait and wait forever. IMHO, there is nothing earthshatteringly better on the horizon, not more than on any other moment, except maybe 2 weeks before the A64 launch. People where waiting for A64 (which was reasonable), then for nForce3 which turned out less than overly impressive, then for faster and or cheaper A64's, now for nForce 250, but by the time its here, people will say wait for S939 and then for PCI-E, then for PCI-E cards to come down in price, then for DDR-II, then no doubt for Tejas and in the meanwhile for 90nm hammers and upcoming pricecuts. Frankly, pretty much all those reasons are likely better than waiting for nForce 250 as well. But today really *is* a great time to buy, like I said, tomorrow will be better, as will next month.. but your best bet is of course waiting a few more years.
> But if his current rig is bareable, waiting is also a wise
>decision.
It always is, as it saves you money. But today really isnt a worse day to buy than any other. Looking back with 20/20 hindsight, what do you think would have been a terrible moment to buy a new rig ? IMHO, at almost any given time, if you made a good purchase decission (like avoiding KT133 boards or SDRAM willamettes), little or nothing substantial would have been gained by waiting another month or so. The "worst" timing I could think off was getting a 533 fsb P4 board just before the 865/875 became available, and at a time AXP was not a really competitive alternative. And even that wasnt all that bad, as even today a P4 3.06 isnt really a slouch, and the upgrade path for early 865 boards is fairly limited as well, not likely exceeding 3.4 GHz.
>Are we really so far apart on this?
I guess not, but you be the judge
= The views stated herein are my personal views, and not necessarily the views of my wife. =
Facebook
Twitter
Instagram