• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Quad FX Dead? Do We Care?

FrgMstr

Just Plain Mean
Staff member
2FA
Joined
May 18, 1997
Messages
58,298
I have now had two emails pointing me to rumors in the industry about the cancellation of the "4X4" platform from AMD. My questions are, does anyone really care and was dual core gaming ever alive? I think things like Skulltrail and 4X4 / Quad FX are great technology solutions when it comes to video editing and encoding, but do gamers really need to be building or buying dual processor rigs while we currently have quad core boxes that go largely underutilized. Do you guys care about multiprocessor gaming boxes? I honestly could care less because we still live in the land of largely GPU-limited gaming engines. Maybe you feel differently.
 
I don't really care. My quad core doesn't get much of a workout as it is. I've owned dual processor boxes like that in the past and the expense never was justified. I could have had the same performance in games in a system that cost half as much.
 
I do some video editing, and tbh I still don't care, price /] performance is not worth it, rather just build another quad core system for 1/3rd of the price and make a render farm on a gigabit network out of it.
 
yea i can see why it is a complete waste for most people as this thing is ridiculously expensive, but for me a simple quad core is all i need, I don't think i could go back to single or my dad's dual core as a main rig
 
I don't really care. My quad core doesn't get much of a workout as it is. I've owned dual processor boxes like that in the past and the expense never was justified. I could have had the same performance in games in a system that cost half as much.

I agree. The whole Quad FX and Skulltrail is just e-penis at most for rich folks to brag about. Quad cores is plenty for gaming since like you said, it's mostly GPU limited.

 
I'd much rather be playing with nVidia's yet unreleased 1Tflop beast than a dual proc system. Better yet, how about a new core from AMD worth waiting for?
 
I generally find it wiser to spend more money on a pair of the fastest video cards available and pick up medium end processors.
 
Truly multi-threaded games really weren't even on the shelves till recently. So having multi-processors or multi-cores (really the same thing to a program) really didn't make a difference. In addition, the bottleneck during the initial announcement of these platforms was still the GPU, not the CPU, so that made the platforms even less significant.

All those announcements were just about having the "biggest", not being practical... (but then what about high-end gaming ever was practical.) I'd bet that if Intel/AMD/Nvidia/ATI came out with a prototype quad-processor, quad-core, quad-GPU system, it would splattered all over the enthusiast sites.

Honestly, the multi-processor systems blazed the trail, but it's the advent of relatively inexpensive multi-core processors that will drive the transition to truly multi-threaded games. Game developers are smart enough to realize that writing complex, truly multi-threaded games wasn't worth the investment till enough computers had more than one processor/core... that wasn't the case with multi-GPU systems since to utilize the power of multiple GPU's the game developer's code didn't need to be massively modified like it has to to be multi-threaded on the CPU side.
 
well, i used to have dual xeon for my gaming system, all i can say that my dual xeon (Nocca) perform only a bit better than p4...(my xeon is P4 age) and in some games like C&C generals it performs worst, beside that i can do so much stuff @ same time, to me my next system will again be dual or quad socket system, since more and mroe games utalize multi-core/treads,

alot of time i doing Divx encoding + Wc3 gaming, my xeon does very well, so i see there is a birght future for multi-croe/thread/socketed system, maybe not for average joe, but it definatly good for me.
 
I haven't ever worked with a 4x4 or Skulltrail motherboard but in my experience with workstation and server class boards tells me that they are certainly a worse buy than their desktop counterparts. For one thing, they cost more, they require ECC/Registered/FB-DIMMs. Add to that the lack of overclocking options and you can end up with something that costs twice as much as a Core 2 Quad compatible motherboard that gives you LESS performance.

In my mind, the only reason for a gamer to ever purchase a Skulltrail motherboard will be so they can have SLI support and still avoid NVIDIA motherboard chipsets. They'll pay out the ass to do it though.
 
There have been rumors recently that FASN8 was dead, and that Spider replaced it. Just a couple weeks ago, someone told me that FASN8 is still being worked on and there should be a dual socket platform for AMD this round - sometime early next year.. Time will tell.. :cool:
 
There are three interesting graphs in the recent QX9650 review that show good game engine scaling with quad core....at 640x480.

http://www.hardocp.com/article.html?art=MTQxMSw2LCxoZW50aHVzaWFzdA==

The technology is already there (except for Crytek's BS about "multi core support"). The fact is though, the games are so GPU limited we just never see the advantages of it...yet hopefully.

We did this story back in March on Quad Core and the SupCom gaming experience, and there is something to be said for this surely, but even then loading does not scale past the 2nd core well at all. But yes, there is still real world value to it. Just not many games utilizing enough AI to do that IMO.
 
this looks like another tech that will just go in the closet, another niche failure. although not buggy as sin, i'm reminded of the old ati rage fury maxx. what a piece of garbage that was!
 
In my mind, the only reason for a gamer to ever purchase a Skulltrail motherboard will be so they can have SLI support and still avoid NVIDIA motherboard chipsets. They'll pay out the ass to do it though.


Even after 3 dead 680i boards you don't even hate NVIDIA that much do you? ;) No answer needed.
 
I've sat down and with the exception of 2 games, the difference of a well OC Dual core and Quad core is very little in benchs... And in REAL gaming it's not noticeable.

Yes perhaps in 2 or 3 years we'll start seeing the Quad cores used.


4X4 wow so just encase you have 3k sitting around go for it. Other wise I think the money would be best spent on a set of Ultras
 
Dual Core and Quad Core is enough i think for gaming and everyday usage unless your a video encoder. They should be trying to increase processor clock per clock and speed/bandwidth with cpus more than adding more cores. Nothing is even made to use more than 2 cores yet efficiently. I think the big bottleneck these days is also the videocard. Videocards haven't seen major progress in years. Now in days we are only see small leaps and gains. Its like they hit a wall already and slowed down. Dx10 is a waste right now so far. Hopefully something better comes out next year.
 
Dual and quad core single socket systems are plenty. I would never go back to a single core system though because dual core at least allows me to do more things while gaming, smoothly.

Yea, I have a quad core and it may not be getting fully utilized but, more games are making use of it and more online games coming out are making use multi core.

4x4 though for gaming. It's just not worth the cost. It's ok for people doing a lot of video and graphics work but that's it. I would rather spend the extra money on video cards and faster drives, water cooling, etc.

I would rather see work on Dual core CPUs with built in physx processors or stream processors taking up the space of the other 2 cores.

Of course, just teaching programmers how to make use of multi cores would be nice. A lot of the graphics processing is still done in the CPU so learn how to use the other cores.
 
I would use a 4X4 and anything heavier I could get for rendering. Calculating just the lighting solution typically takes more than 24 hours using six cores but network latency is a problem. It wouldn't be with 8 on one mobo.
 
I would rather see work on Dual core CPUs with built in physx processors or GPU type architecture that could be used for that.

How would you manage that, Ageia makes physx processors, which is a separate company, and AMD didn't want to buy them, and intel already bought havoc.
 
I'd be more impressed with multicore video card than CPU for gaming purposes. Right now I don't see a big leap forward in performance with multiple cpu cores given I still run an ancient processor but a decent video card and do just fine with most games still.

Now if they put out a 8 core video card and don't make me use some stupid SLI crap driver, I'd be more impressed. SLI enabled and tuned via drivers seems to be a disaster. Give me multicore goodness with the synced cores processing video ram at near X times the speed (given the number of cores), that's where money is.
 
No one really cares about x64 so how are they to care about dual core, much less quad core.

Tho really it all boils down to software. If the developers don't make the software for the platform why would the consumers care? Problem is it's also vice versa, if the user base doesn't care the developers are not gonna make the push to modify their code.

Funny thing is, like every computer you get now a days (even going back a year or so) has a 64bit/dual core CPU and there is like VERRY LITTLE software out there that makes use of it. Hay developers! The hardware is out there! Bring your code up to snuf!

/me goes back to gaming on his 2x2 3.2GHz computer running 64 bit OS.
 
As a developer I always wanted one, but more for the fun of 'make -j8' than anything else...

Oh yeah, and to compensate for my tiny... um, desk?
 
To be quite honest I care a bit. Right now does 4x4 really do us any good? Nope. But what about 2 years from now? Especially with the advent of things like bulldozer and nethalem. When, and hopefully sooner rather than 10 years later, when programs become more fully multi-threaded those next generation of cpu's will really get a chance to shine on a more matured platform.

Can you really deny the convenience of a cheap desktop solution for multi-cpu gaming and media encoding? As it is quad cores are running for dirt cheap, there would have been nothing nicer than a convenient way to get a dual socket platform for desktop use.
 
I care too, I've been waiting for skulltrail to pop in some 5355s and go. I game on my workstation now and it's also a folding beast. For some, it's just easier to have a single box with two quads then build a folding farm, this was one way to have a box that does it all for you. Probably just end up with a 5100 or 5400 intel chipset board and go from their. Just use ATI cards instead of nVidia.
 
I care too, I've been waiting for skulltrail to pop in some 5355s and go. I game on my workstation now and it's also a folding beast. For some, it's just easier to have a single box with two quads then build a folding farm, this was one way to have a box that does it all for you. Probably just end up with a 5100 or 5400 intel chipset board and go from their. Just use ATI cards instead of nVidia.

The only problem with that is the fact that ATI doesn't make a card as fast as the 8800GTX or Ultra.
 
A lot of people keep on bringing up multi-gpu/multi-gpu socket solutions as being a valid retort against multi-core cpu/socket solutions. But dont such solutions ultimately scale about as well as mutli cpu solutions? I mean grant it they do a bit better but ultimately its not like you're making a better argument there either.

Two things I'd like to see. QuadFire (or whatever its called) reviews to show multi-socket gpu scaling and better cpu coding from programmers. Better gpu coding as well for that matter or just better multi whatever coding overall.
 
Well, if the 'native' quads scale well, then they should do it for the halo effect. There will be someone that might wanna do that. Phenoms are not pathetic, they are a good jump from K8, it's just that c2d is that good and a bit better.

If Phenom can be scaled up fast with frequency, then it will become a BIT more enticing.
 
The enthusiast market drove the Pentium M to the desktop, and that was good. Otherwise, we already know quad cores are overkill. And QuadFX was especially a bad idea from the start.

Enthusiasts anticipated the worst from AMD with a blind fervor, and now they reap what was sown. Doesn't bother me a bit. AMD can't afford to squander their meager resources on such side-projects. I want AMD to focus on sensible projects for consumers like me, who won't give up on them without a fight.
 
The only problem with that is the fact that ATI doesn't make a card as fast as the 8800GTX or Ultra.

Yeah, but I can run my 2 2900XTs on 5000, 5100 or 5400 series chipsets, so it'll make up for it. Plus, we'll hopefully have a folding client out for the 2900/3800s soon.
 
The only reason Quad FX even existed is because Intel had a quad core CPU and AMD did not, this was there temporary answer. Since the introduction of Phenom X4, it's a dead platform and serves no purpose any longer.
 
Currently my Core 2 Duo @ 3Ghz handles everything I throw at it fine.

In Febuary, I'll probably upgrade to a 45nm quad and hopefully hit 4Ghz, then I'm sure that will last me for a long time. (P.S. 2 years is considered a long time for me)

By the time I would need more than 4Ghz x 4 processors, there will probably be native octocore chips with 24Mb of L2 cache and an integrated tri-channel memory controller! (on one socket)

So I don't ever see a need for a dual SOCKET system.
 
The only reason Quad FX even existed is because Intel had a quad core CPU and AMD did not, this was there temporary answer. Since the introduction of Phenom X4, it's a dead platform and serves no purpose any longer.

And what happens when Intel gets to say they have the first octo-core and once again have that luxury for over a year? The way AMD currently does things will not be able to keep up with the way Intel has been doing things for the past 2 generations of cpu's and what I believe they have already alluded to continuing doing with the next generation chip.

AMD's only way to match up with Intel's quick and easy way of doubling the number of cores per chip for new generation chips, which is a process they have refined quite well now and should only get better, was Quad FX. With the advent of the loss of Quad FX Intel's marketing team will be yet again for the third time around be able to say we can do something that AMD cant. And we can deliver to you a market segment that our competition cant.

Will octo-core really matter at that point in time? Probably not. But when has p.r. and marketing ever been about whats actually at the time practical? At the end of the day its ironic how everyone can have the foresight to see why AMD bought ATI for their next generation cpu, among other things, which was 2 years out, but dont have the foresight to see the practical use in marketing and just overall practical use of Quad FX after its been matured and refined.

The current future of cpu processing is the number of cores you can put into a cpu. We've been through the ghz generation and then the IPC generation. Now its the time of the multi-core.

I'm going to stop here for now even though theres so much more to say about the possibilities of Quad FX. All I'm going to say is you guys are cutting it off short of its true full future potential. Was it a shitty idea a year ago? Yup, and if you look through my post history you'll see I was leading the mob to its crucification, but things change and the age of the multi-core is just beginning.
 
I fully agree with DD. By the time it matters on the desktop, octo-core processors will be for sale...
 
I don't care about multi cpu setup. In fact, I don't quite like the idea of running 4 graphic cards either. It's not efficient.

As mentioned, currently its hard to even utilize quad cores. getting another quad core would be a stupid way of wasting money, as application are simply not ready for such level of parallelism. By the time applications could take advantage of 8 cores, we'll already have a CPU with all 8 cores packed into it, with better efficiency, less power consumption and less heat dissipation.
 
And what happens when Intel gets to say they have the first octo-core and once again have that luxury for over a year? The way AMD currently does things will not be able to keep up with the way Intel has been doing things for the past 2 generations of cpu's and what I believe they have already alluded to continuing doing with the next generation chip.

AMD's only way to match up with Intel's quick and easy way of doubling the number of cores per chip for new generation chips, which is a process they have refined quite well now and should only get better, was Quad FX. With the advent of the loss of Quad FX Intel's marketing team will be yet again for the third time around be able to say we can do something that AMD cant. And we can deliver to you a market segment that our competition cant.

Will octo-core really matter at that point in time? Probably not. But when has p.r. and marketing ever been about whats actually at the time practical? At the end of the day its ironic how everyone can have the foresight to see why AMD bought ATI for their next generation cpu, among other things, which was 2 years out, but dont have the foresight to see the practical use in marketing and just overall practical use of Quad FX after its been matured and refined.

The current future of cpu processing is the number of cores you can put into a cpu. We've been through the ghz generation and then the IPC generation. Now its the time of the multi-core.

I'm going to stop here for now even though theres so much more to say about the possibilities of Quad FX. All I'm going to say is you guys are cutting it off short of its true full future potential. Was it a shitty idea a year ago? Yup, and if you look through my post history you'll see I was leading the mob to its crucification, but things change and the age of the multi-core is just beginning.

Unless AMD was able to get Quad FX down in price and up in performance, there is no practical use for it and there is no reason to keep it around. It isn't about haveing foresight, it's about what works and what doesnt. Quad FX might have potential, but potential alone is not good enough to merit it's existance.
 
I generally find it wiser to spend more money on a pair of the fastest video cards available and pick up medium end processors.

in gaming rigs your exactly right, unfortunatly the major gaming PC oems don't share the same feelings. you've gotta have a core 2 Extreme before they even consider the 8 series.
 
Unless AMD was able to get Quad FX down in price and up in performance, there is no practical use for it and there is no reason to keep it around. It isn't about haveing foresight, it's about what works and what doesnt. Quad FX might have potential, but potential alone is not good enough to merit it's existance.

I have to agree.
 
Back
Top