well to be honest I'm glad AMD arent manufactrinf their own chipsets. This gives 3rd party chipset manufacturers to get creative and really flex their muscle. With Intel u have very little choice and more oten than not their boards ALWAYS carry a compromise.
810E no AGP
825/815 : WTF RD RAM!!!
845: Ok this chipset I actually liked, but it had SDRAM to start with and the DDR's had no AGP!!
865: An overall CRAP board
915:Could have been a great board, BUT... no AGP, only one IDE connector
925: came immidiately after the 915 and offers a couple of nice improvements... those who bought the 915 m YOU SCREWED!
Another characteristic shortcoming of Intel boards is the fact that everyt time they release a board, within a couple of months, some technology (board related) becomes mainstream and ur left in the dark.... ITs as if they always knew the trend but leach the market by saturating with outdated tech then force customers to get another board well within ,what one would consider its useful life-span.
The same applies to private companies that are allowed to use Intel chipsets in their computers, such as say DELL,COMPAQ etc. Intel basically offers these looters so much flexibility that they can pick the board clean, get it manufactured at cut-throat rates in substd manufacturing centres then sell it at the same price as the OEM intel board. Too many holes in the Intel strategy that allow customers to be taken for a royal ride.
The reason why AMD boards (the oler ones atleast) faced so many compatibility issues was cause of teh monopoly Intel had established because of which 90% of compnent manufacturers produced products that were tested with Intel machines. Its not setting a manufacturing standard, its the IT MAFIA.
Apart from the BIG flop SLI boards I thought most modern AMD boards run well without problems.The blame for the SLI failure should be placed squarely on Nvidia who jumped the gun and released chipsets to enable the same.
I've been using AMD chips since the K5-233 was issued. I've had the K6-2, K6-3, Duron, Thunderbird, Athlon, AthlonXP, AthlonXP-M and now Sempron. I've yet to try the AMD64, though I am salivating over one at my handy-dandy computer place. The owner knows I really want that rig and, I swear, she puts in on the counter when she sees my van roll in (LOL).
I have yet to have one problem with any of these chips. I've built several for customers and have several of my own (5 in the house at last count - Lots of kids!). The only criticism that I have had with AMD are the heat issues, but I've managed to get around that by buying quality heat solutions instead of using the boxed fans that come with the CPU's. I've never burned one up, never had one give up the ghost and even had one survive a lightning strike which burned off the leads of the memory and welded a few pieces of the mobo. Just took the chip out, stuck it in another board and went on my merry way. Been running 2 years now with no crashes, no nothing.
I did have an Intell P4 once. It was in a laptop and it actually burned up on me. It was a Compaq laptop and it was hotter than anything I have EVER used. In 2 months, it fried itself and I excercised the store warranty at Circuit City. I switched to an Athlon 1700+ and have had it going on 3 years now. A little warm, but not "egg frying" hot like the P4 was/is.
On the cost front, the price per clock cycle ratio, AMD beats Intel. For innovation, where Intel used to have the lead, AMD has decidely taken over. Even back in the P233 days, the AMD chip, if cooled properly, ran well. In fact, one of my clients still uses her K5-233 to digitize her embroidery and pass the data to her embroidery machines. Again, I had to oversize the heat sink and fan to overcome the heat problem, but the CPU responded well and lasted a long time. I've yet to have an Intel chip last as long as that puppy has.
No thank you...I'll stay with AMD. I'll still put them in my clients machines and recommend them every chance I get. I've not had issues with mobo's. So what if you have to update the bios every so often? I still do that less than I patch Windows.
Bill i find it hard to believe a P4m burned up on you.Im sorry but the laptop doesnt put out enough power to burn up a P4 and the P4 will shut down before it burns up.If it just died ok.But no way it burned out.Also the P4 478( not talking the prescott here) socket runs way cooler than the amd athlon.And start saying ah your just an intel fan boy because im not i acutally like AMD very much but you should give credit were credit is due.
Hey Bill I couldn’t have said it better! I read your post and thought that I had wrote it one late night and forgot! I have 4 AMD machines running all kinds of configurations and my machine has gone through many transformations! I am running a Gigabyte NVIDIA’s Nforce 2 board that is full featured and I am using them all. 2 EIDE drives raid 0 for a dual boot 2000 and XP, 2 SATA Raptors raid 0 with XP for gamming and 1-120 gig HD for storage. I can boot to the EIDE and have a choice between 2000 or XP or select SATA to boot and play all my games, with the swap file of the SATA raid array on EIDE
raid array and visa versa, I get the best performance! I use onboard sound (I had a creative SB Audigy 2 Card but it died and I haven’t replaced it yet) quite a few USB devices, 2 DVD burners! All on an Nforce 2 board that is fast and reliable. I had heard that the original Nforce chip set had problems but not the Nforce 2. I don’t see the basis for Sanders comment “but also particularly bad ones such as the first Nforce2”. Nforce 2 has been the best chipset for AMD since it first came out. That is why Intel is now using NVIDIA chipset for there boards. “NVIDIA Corporation (Nasdaq: NVDA) and Intel Corporation today announced that the companies have signed a broad, multi-year patent cross-license agreement spanning multiple product lines and product generations. Additionally, the companies signed a multi-year chipset agreement for NVIDIA to license Intel’s front-side bus technology. This will enable NVIDIA to deliver the NVIDIA nForce™ platform technology on Intel-based systems.” Nvidia website.
GA-7NNXP 1.5 gigs of PC 3200
2 WD 74 gig Raptors Raid O Win XP Gaming array
2 WD 250 gig Raid O Storage
1 WD 30 gig server 2003
Audigy 2, Creative 6:1
All stuffed into a Cheiftech dx-01wd
I don't want to get people mad around here, but i turns that I think (and that's is just my opinion) that is Intel the one that releases something in the rush and then make turns it always been that way since first pentium with math co processor flaw. And is Intel the one that is making users to change motherboards type. Let me say something, I was an Intel fan or at least I used intel processors up to a great Intel 815 E board. After a year of using such a good invention (derived from another great chipset like the Intel 810) with a celeron 700Mhz then Intel came out wih tualatin.
So I couldn't upgrade my PC to more than 1Ghz without having to change mobo. (Even if there were ways to load a tualatin in a 815 mobo, but that was just to 1.1Ghz... so what? If I want to upgrade, I want at least, to double Mhz). I remember my old celeron 300A, to upgrade that I had to change the board, but I wasn't mad about that since that board came with no AGP, PC66 memory support, old ATA66 controllers, seemed that I needed to change it anyway and the Intel 815 was the best option at the time. But I just think that Intel could take care of their customers at that time by releasing a Penium 3 with more Mhz for their great mobo.
First Athlon was around, and at that time AMD just changed their Slot A Athlons to Socket A thunderbirds, with cache mem on die. So at first I though that I would have be a victim anyway if I was an AMD user. However, AMD released a Slot A version of Thunderbird, with caché on die, so imagine, AMD releasing a full cartridge like processor for what? if the cahe mem wasn't on the PCB! Just for one thing: Make customers happy. And that, that made me mad. I have to change my Intel mobo to upgrade again.
But the story isn't finished here, then came the Willamette story. The First P4, socket 423, do you remember that? I was considering it until performance benchs revealed it was slower than a Pentium 3. It's not the only thing. The memory bridge that Intel released to make the new mobos compatible with SDRAM instead of RAMBUS was buggy/unstable/crap (I think it was Intel 820 or 825, don't remember). AMD with socket A was gaining popularity at the time, and socket A DDR mobos came out. Intel could not release a DDR mobo for a time since it had an agreement with rambus. So the thing turned to be more "AMD-ME please". Intel, suddenly went out with another socket, 478. Rambus agreement just died and Intel was on the road again with DDR. Anyone buying P4 socket 478.
I Could not understand that though.. With the release of Athlon XP I couldn't believe such a performance kicker with so little money. I remember in the year 2001 I did the greatest upgrade I ever did. I've got an Athlon XP 1700(1466Mhz) with a MSI 745 Ultra, the SIS chip was quite popular, there were(are) even fan comunities around the web. It blowed away the celeron 700 on the Intel 815. I was so happy, and for the years that came, I upgraded to an AXP 2400(2Ghz) Then tested a barton 2500 (1.8Ghz) -since the mobo was bus 266 only, I just increased multiplier to reach the 1.8Ghz. Just kicked-.
The story is not ending though. AMD didn't change socket A since it was released til now, and I expect socket A to be alive 1 more year. That was how long? when was the socket A released? in year 2000? Which year are we crossing? 2005? WOW man. Did you know you can load any Athlon XP in the old KT133 chipset by ust raising multiplier? I know that KT133 is slow as hell, but then, you can load an old Duron 850 in (almost) any new socket A mobo. This is flexibility anyone ever wanted.
Now I run an Athlon 64 3000 socket 754, they say socket 754 is going to be replaed by the socket 939, but I know what AMD will do. They will keep customers happy. A64 It's cool and quiet and really powerful and stable on my K8T800. each and every AMD processor runs cooler and cooler. Each and every Intel processor runs hotter and hotter. With prescott heating up the whole planet, DDR2 failing as Rambus did, and Pentium 4 Extremely Expensive Edition being released I think I will never go back to Intel again even if chipsets are great. The first Nforce 2 kicked everything from the beginning to the end. What Sander meant maybe was Nforce 1. Nforce 1 came in 2 flavors, with or without dual DDR. Just for the knowledge of anyone here, the double channel Nforce1 that sported Geforce 2 onboard, was a very powerful choice at the time since the memory bus was great with the slow DDR at the time. One more thing, AMD still invents. The last cool thing released by Intel was Hyper Threading. With considerable performance boost in windows that makes a natural choice for the heavy multitasking guy. To those using Hyper Threading now, I would consider testing an Athlon 64 939 socket 939 with SIS 755FX chipset. Try it out and then tell me.
I have to agree with Sander, being in the computer business I build or supervise the building of a lot of custom computers. The techs me included love to get an order for an Intel sytem especially If its an Intel motherboard. It almost always insures a trouble free build.On the other hand If its a new board designed for an AMD CPU we pick straws to see who the unlucky fellow is to build the system. Case in point we went through 2 Asus A8n-SLI boards before we gave up and bought a Gigabyte SLI. Worst experience I have had lately with a motherboard. And the Gigabyte was no walk in the part either at least it keep working as we struggled with drivers and moveing parts from one slot to another. I have read many forums and it seems the hardcore has come to accept this as normal. I dont agree that hardware should be released to the public to work out the so called bugs. If you pay $200.00 for a motherboard it should work out of the box. Sound cards should not lock up the motherboard nor should 2 thirds of the memory modules designed for the system not work. Also If there are required power supplies it should be clearly noted on the information about the board and on the box. I have watched P/S manufactures ramp up to build compatable power supplies for the new SLI boards. How can you release a product requiring a part not in production yet. Yes there were a few 24 pin power supplies out there but who knew they needed one? First I was told by Asus I could use the 20 pin with no problems as long as they met the requirements, when problems accrued they changed their tune and said no you have to have a 24 pin. Also some say a dual rail wont work, others say its best to use a dual rail. The MSI requires a 24 pin with a -5 volt pin or you can barely hear the onboard sound. The first DFI had a mod to enable SLI but no bridge and then Nvidia changed the chips. All the boards seem to have problems with certain hard drives. Is this expected by the hardcore AMD people? Well I have to say I am by no means an Intel banner waver but I also dont expect these kind of problems either.
I was a Intel fanboy until I built this rig I'm running now.
I have a A64 3200 @ 2.5 gigs on a Epox board which in all ways is better than any Intel I have ever had.
I had a msi neo 2 board that had issues but the epox is spot on!
This is the best performing rig I have built. It is even much faster than my 3.4EE
Prescott @4.2 gigs. And runs much cooler.
Intel was doing it right for several years but now have some real heat issues.
Amd on the other hand has Intel by the balls when it comes to memory bandwidth. The "on board memory controller' seams to make loads of difference.
I did like the Intel 850 chipset and rambus. At the time it was fast and stable(at least for me).
The 865 pe chipset was kick ass also.
And then came that cooker Prescott, Intel has paid the price for that one.
Currently I have an Athlon 2700XP that failed, due to heat, less than a montha after I bought it, AMD replaced it when I mailed it to them. AMD has overheating problems!!! I have an Athlon 900 that is about three or four years old but it failed last week. I have had sveral AMD and pentium computers over the years, the AthlonXP 2700 was the only processor to fail on me over the last ten years. I still have a Pentium II 450mhz up and running with 128MB of ram. Now i need to get a new Athlon 2400+ to replace the 900, this way I can still use the same motherboard. The Pentium II is about six (I think) years old maybe seven. AMD just needs to solve the heatsink problem. They should offer a liquid cooled alternative heatsink as a purchase package so the warranty is not voided.
Now with the advent of ther socket 754 and 939 pin chips along with the hypertransport technology and the OPTERON they will definatly add a serious amount of XEON crushing power to there already strong arsenal. Now if AMD would just sign a contract or partneship agreement with a leading manufacturer of chipsets they could develop and release an AMD designed, tested, and certified chipset that would take care of all the AMD beleived standards and it would leave all the overclocking chipsets up to the third party designers. This would also improve the relabilty of the AMD cpus'.
What do all of you think if AMD did this with the design of a chipset?
actually we arent talking about the CPU's themselves. The issue here is compatibility and reasons why the Intel hs fewer issues in that area when compared to the AMD.
Like I said it takes time for people to get off their path forced via pure inertia and take a good look at the competition and its possiblities. The world was very slow to take notice of the AMD to start with but now the pace has picked up and more manufacturers are coming out with AMD specific hardware. Its improving all the time. So honestly I dont think its a problem that AMD do not make their own boards.
I can understand why someone would think such a thing about AMD chipsets. The poor performing but stable Intel chipset/processor appears to be a better system all around. I think the information presented here is not a true or fair assumption. My reason for saying this is that I repair computers daily. I produce AMD systems exclusively but I work on Intel systems equally. Meaning, I have the same problems with Intel Systems as I do AMD. It's always a driver or software fix. As far as Hardware, I always get the usual headaches. Driver issues, software conflicts. What's worseare the people that were used small hard drives during the Windows Me and Windows 98 era still delete DLL files nd don't know how to manage or how to delete files properly.
I don't see all or that many problems with the chipsets and I don't have to flash the bios on many boards. Every system including Intel that I have built or worked on has required some type of update. Microsoft even scans the system and offers updates for Windows to take care of many of the
issues. The latest technology always has bugs. Intel or AMD. That's why I never purchase the latest. I give the manufacturers a chance and Microsoft a chance to work the bugs out. My system may be outdated technologically, but I can still do video editing, burn DVD's and add any component on the market with my current system. This has enableed me to watch patiently. Once the bugs have worked themselves out I will then upgrade. And it will be an AMD 64 chipset either 754 or 939. The systemt will have to outperform my current system by at least 40%. Which none of them currently Intel or AMD do.. I based this solely on my computer's ability to do work. I compare it to every system that comes to my workbench. My system is modest but performance wise I have had up to Pentium 4 3.2Ghz in my office and it was not faster than my
Barton 2500+. It did not out perform my current system (listed below) by any order of magnitude.
Albatron KX400 8 VX Pro motherboard (And this is just a back up.) I had a couple caps leak just last weak
and I used this to replace my KX 400 8X board still under warranty.
Barton XP 2500+ processor over clocked to 1.97 Mhz. Will go higher but that's fine for now.
80mm fan on the processor
ATI 9600 Pro bios updated to 9700 Pro Still Awesome and excellent in Video Editing and Capturing.
Wireless LAN Card
Soundblaster 5.1 Sound Card
Direct burn to DVD capture card
512 MB OC Systems PC 2700 Memory
Western Digital WD 1600
NEC 3500a 16X Dual Layer +- Burner
Chieftec Dragon Case Sky Blue
As I said a modest system and I haven't seen a single game that would not run on this system or application.
When I build a system even below $500 I make sure it will run those same games. I have not had a single
system that any of my customers could not run the games or their choice. And I do test every component and I don't use components that don't meet the challenge. Maybe more manufactures should do this.
On the subject that another responded to. It does seem that Intel gives us products that have been
tested, yes and oudtated the moment you purchase it. Less than 3 to 6 months after you have purchased an Intel Machine it is obsolete. On the other Hand AMD does seem to present a fresher outlook by allowing 3rd parties to do their thing. I like that and I am forever devoted to AMD. I am an AMD Exclusive builder and I always will be. Intel processors are expensive and much slower in performance compared to AMD. In the Past 5 years I have built only 1 Intel computer. I don't have any reqrets. And my customers are fully satisfied too.
It is interesting to note that no one is mentioning which OS they are using when 'evaluating' a chipset/cpu compatability and functionality.
Windows has been known in the past to have some rather , er, interesting methods of choosing drivers for non-intel equipment . Interesting to the point of stubbornly re-loading 'incorrect' drivers and continuously 'misrepresenting' certain pieces of hardware with the result being lower or less stable than possible configurations .
It could be argued reasonably that the third party orgy nature of the AMD systems leaves loop holes wherein the WinOS ( scatagorically speaking) does not or cannot identify "collage" integrations with enough sophistication to provide the best possible driver implimentation ( IN Win's defense in some cases hardware snafu's may actually make it impossible , but those are actually pretty rare ) . And also that possibly it falls to the cpu and mobo manufacturers to compensate with bios revisions and tweaks . Not to say that ALL integration problems on non intel systems are due to this ( certainly mixing and matching tech-animals has it's caveats ) , but certainly a number of issues can find thier roots with OS management of system resources .
Generally I have found intel / intel systems to set-up and be implimented very smoothly under win os's , and that Amd / whomever sytems often need tweaking . Even so , when given the choice , I will choose AMD systems 32 out of 35 times.
I guess the little bios and drivers tweaks are not as annoying as Intel's draconian "proprietary" marketing strageties , and at the bottom line after all is said and done Amd systems simply offer more bang for your buck .
AMD owners are usually ahead of the flock, and cause a lot of their own problems by pushing their systems to the limits. I can easily get my aging 1800+ rig at home to out-perform the Dell P4s we have here at work.
Intel configurations are stable as long as you're willing to disable virtually everything. Spend a little time on Google looking for applications that don't work when HyperThreading is enabled, for instance. Sure, that may be a problem with 'older' applications, but the entire reason we're still on the x86 platform at all is for backward-compatibility.
My big problem with Intel is that they always try to dictate to their consumers what technology they'll use, i.e. RIMMS and separate 32-bit and 64-bit product lines. AMD is far more likely to go the route of open standards and letting the consumer decide, as in DDR, HyperTransport, 32/64-bit processors, etc.
As for the so-called compatibility issues with AMD rigs, I haven't seen anything that gets more than a 'so what' since VIA fixed their initial Athlon chipset problems. That was several years ago guys, it's time to move on. If AMD can still get grief for those things, then Intel should still be getting grief for the Pentium floating-point flaw that was worse than all AMD compatibility problems throughout their history combined.
I'm also glad that AMD isn't quietly trying to push mainboard manufacturers completely out of the way like Intel appears to be doing. I like having choices when it comes to technology.
Looks like it, most of the people that reply just missed the point I'm making though. I'm not saying all AMD chipsets are quirky, or AMD is, but give a few examples of recent AMD chipsets that are. That I'm making a valid point and not some one off flame is indicated by the various threads in our forums from readers having problems with recent PCIe socket-939 chipset, the Nforce 4 SLI being a prime example.
Eric ur bang on about the utility value of 3rd party chipsets. This has always been a feature of AMD boards. And I have ato say that this is due to the fact that AMD gave people like Nvidia the opportunity to tinker around and produce key-compnent specific hardware.. imagine AGP-CPU specific boards..... that pretty much covers the main use of PC's today.
Despite all their problems, the earlier AMD chipsets offered so much ahead of their time.. that was something to appreciate.
I think “AMD chipsets, quirky by nature?” kind of implies that you think all AMD chipsets have this problem, and “particularly bad ones such as the first Nforce2” seems to imply Nforce 2 were a real problem. I don’t think that is what you meant.
The real problem here is the fact that the chipset manufacturers sell unfinished parts and then we have to find the problems and push for them to update and repair. It seams that SLI and PCIe are the next features that need to have the bugs ironed out.
I've been running an MSI Diamond for the past 2 weeks with an ATI X700 pro and just changed to 2 6600 PCIe SLI setup to test. No problem what so ever doing so I must add. When comparing chipsets, you also have to compare the rest of the system and also the chipset that have comparable features. Doe’s Intel have any SLI chipset out there??? Then now can you compare any Intel stuff to SLI setups. When Intel finally catches up to NVIDIA on this... then you can compare.
Most of the problem that AMD has had in the pas were from Cheap OEM's wanting to sell the cheapest possible systems with a Key few catchy hardware parts in them like ASUS mother board in them and then cut back on the important stuff like a good Power supply. I can't count how many unstable systems I've fixed by just adding a good power supply in it. The 30$ cases and power supply combos just won't do it. This goes for Intel and AMD system now days.
As for heat, The XP did create a lot of heat but nothing like the new 3.6 / 3.8 Intel’s of now days can create. Intel’s fix to this... Let's make the system throttle down when heat goes up. What a good fix... when you need the CPU power lest make it run slower. The user will never know and we can brag of 3.8 GHz in a CPU. ??? Good investment I tell ya. The new AMD 64 run very stable with very little heat, Many neat features and a good choice of stable mother boards. If SLI is too much of a challenge for you stick to an NFORCE 3 250 with AGP, Can’t go wrong with that.