Anand doesn't like 4x4

[Eska]

clicky

 

[DrMrLordX]

Shows pretty much the same thing we've seen in other reviws: 4x4 fails to outperform Kentsfield and sucks up twice the juice while losing benchmarks. Ouch.

 

[Eska]

Originally posted by: DrMrLordX
Shows pretty much the same thing we've seen in other reviws: 4x4 fails to outperform Kentsfield and sucks up twice the juice while losing benchmarks. Ouch.

yea but it's official now cause Anand said so.

 

[Shimmishim]

Originally posted by: Eska

Originally posted by: DrMrLordX
Shows pretty much the same thing we've seen in other reviws: 4x4 fails to outperform Kentsfield and sucks up twice the juice while losing benchmarks. Ouch.

yea but it's official now cause Anand said so.

ftw

 

[tylerw13]

Originally posted by: Eska

Originally posted by: DrMrLordX
Shows pretty much the same thing we've seen in other reviws: 4x4 fails to outperform Kentsfield and sucks up twice the juice while losing benchmarks. Ouch.

yea but it's official now cause Anand said so.

ya i hate to say this but amd coming out with this thing now was just a bust....they should have just waited and focused on their up and coming processors

 

[Furen]

There's like a 60W power draw swing between idle and load on the AMD system. This means that each CPU has a power draw swing of around 30W. 95W at idle? Sounds to me like AMD's got some very leaky processors.

 

[Lonyo]

Originally posted by: Furen
There's like a 60W power draw swing between idle and load on the AMD system. This means that each CPU has a power draw swing of around 30W. 95W at idle? Sounds to me like AMD's got some very leaky processors.

I doubt they use 125W under load, that's just the max they can use, nor what they will use.

 

[coldpower27]

Originally posted by: Furen
There's like a 60W power draw swing between idle and load on the AMD system. This means that each CPU has a power draw swing of around 30W. 95W at idle? Sounds to me like AMD's got some very leaky processors.

It seems as the article stated that Cool'N'Quiet was not functioning properly on their test bed so I guess they were idling at 3GHZ a piece.

 

[Furen]

Well, the CPUs BETTER use at least 125W at full load, otherwise the motherboard, harddrive, memory and graphics card (at idle) are drawing 250W+. Here's my theory: I think AMD couldn't get decent yields for 3GHz, and pulled an Intel Pentium D 3.2 EE on us (which actually draw close to 150W at full load) by increasing the voltage massively (hence the 1.475 vcore) in order to achieve the clocks it wanted for its paper launch (availability is scheduled for early 2007). I think it is a horribly stupid thing to do, since all it does is tarnish the Quad FX name before its even out. Maybe AMD didn't expect everybody to measure power draw?

 

[nyker96]

neither do I.

 

[peternelson]

Can anyone point me to a benchmark review of QuadFx 4x4 using anything other than 32 bit XP?

Didn't think so.

These reviewers need to run some XP Pro x64, Server 2003 SP1, Vista, Linux before writing the final words.

Also given that AMD is already testing true quadcore Barcelona (also in today's news), the 8 core potential of this platform comes closer.

 

[Furen]

Posted by: ted
to know about 4x4 performance using windows vista, please follow this link
http://babelfish.altavista.com/babelfis...etta-amd-per-i-quattro-core_index.html

That shows that a NUMA-aware OS does help AMD a whole lot, the same can be said about 64bits. That still doesn't solve the +200W power draw problem, though...

 

[DrMrLordX]

Benching a gamer enthusiast platform under any of the 64-bit Windows variants makes no sense given the terrible driver selection currently available. There's a reason why few people actually use those OS variants on home machines.

Not only do I expect the picture to look about the same under a 64-bit OS, but I also don't really care what the results would actually be.

 

[Rike]

It's funny, to me, AMD processors always seemed to represent simple, elegant, and efficient designs. The overall design of this 4x4 system seems to be the opposite of those virtues and it still doesn't get the job done as well as other solutions. I hope, for AMD's sake, they don't dedicate many resources to promoting this because Quad FX (4x4) looks like one for the trivia books.

 

[Conky]

Originally posted by: DrMrLordX
Benching a gamer enthusiast platform under any of the 64-bit Windows variants makes no sense given the terrible driver selection currently available. There's a reason why few people actually use those OS variants on home machines.

Not only do I expect the picture to look about the same under a 64-bit OS, but I also don't really care what the results would actually be.

Exactly. This is supposed to be an enthusiast's "quad" CPU solution and gaming enthusiasts for the most part don't care about 64-bit apps.

I am embarrassed for AMD that they felt the need to cobble together such a lackluster response to Intels real (or at least on one socket) quad. I'm more embarrassed for anybody trying to spin this trainwreck "quad" into a viable solution. Anand was right... they should've just taken their lumps until next summer and then brought a real product to market.

 

[Regs]

I just hope it didn't actually take 4 years to "innovate" this monstrosity.

 

[cmdrdredd]

Originally posted by: peternelson

Can anyone point me to a benchmark review of QuadFx 4x4 using anything other than 32 bit XP?

Didn't think so.

These reviewers need to run some XP Pro x64, Server 2003 SP1, Vista, Linux before writing the final words.

Also given that AMD is already testing true quadcore Barcelona (also in today's news), the 8 core potential of this platform comes closer.

This doesn't even matter when 99.999% of people will and are running WinXp 32bit anyhow.

 

[dev0lution]

WTH is this supposed to be? A gaming platform? A workstation? Server technology on the desktop? It doesn't seem to do any of the above exceptionally well and until 64bit OS, multi-threaded apps and decent Quad SLI drivers are available, it's all just a big "What if". Who knows where things will go between now and then...

This is AMD's P4 launch. I'm seriously surprised that they even bothered to launch this, instead of scrapping the whole thing and putting those engineers hard at work at getting K8L out a bit earlier. Maybe they've been listening to the ATI guy who launched Crossfire

If you need a specialized system for media apps or server-like performance, then go buy a server or a workstation. It's somewhat insulting to claim "innovation" when you just cobbled together a couple of old dual-core chips onto the same board that gobbles power like mad. It's kind of funny that AMD pokes fun at the QX6700 for being two C2D cores on the same dice, but their solution gets spanked for less wattage and price (once the gouging stops in a month or so). There's some good marketing slogans in there... 2x the Power (consumption) and 2x the sockets for almost the same performance! sweeeeet!

And you have to buy 2 FX's at a time? What happened to the flexible upgrade angle? These things aren't server compatible and aren't desktop compatible with AM2. Don't even think of posting "boohoo...I have no upgrade path" threads in a couple months if AMD kills this thing off.

Seriously, I'm in awe at what a poor showing this is...

 

[Furen]

It looks like TR is the only site that got Cool'n'Quiet working on the Quad-FX platform (and one of the few that tested on a NUMA-aware OS). Power draw at idle was ~210W (much better) but the load power was still sick. Performance was also very good, no doubt due to AMD64 (a cookie for AMD) and NUMA. So I stand corrected, it's not the damn power-hungry chipset that draws all kinds of power at idle, it's the damn 1.475v 3.0GHz Opterons that leak like a sieve even when idle.

 

[Idontcare]

Furen, exactly, I'm with you on this. Sad how folks are focusing on how minimal the waste heat could be if they had bought 4 processors and proceeded to do nothing with them.

Only one number should matter when one is talking about an extreme performance product, and that is its extreme performance.

Reading all this talk over whether these quad-core platforms are the equivalent of 3 100W light bulbs or 6 100W lightbulbs makes me feel silly. What are folks buying these systems for if 6 100W lightbulbs is their metric of success?

 

[harpoon84]

Originally posted by: Idontcare
Furen, exactly, I'm with you on this. Sad how folks are focusing on how minimal the waste heat could be if they had bought 4 processors and proceeded to do nothing with them.

Only one number should matter when one is talking about an extreme performance product, and that is its extreme performance.

Reading all this talk over whether these quad-core platforms are the equivalent of 3 100W light bulbs or 6 100W lightbulbs makes me feel silly. What are folks buying these systems for if 6 100W lightbulbs is their metric of success?

On that note, Prescott was quite a success then?

It's better than A64 at encoding and multimedia, and while A64 was better at gaming, 'games are all GPU limited anyway' (argument used for C2D vs A64).

I think power efficiency is quite relevant, especially since it also determines the PSU requirements and cooling requirements (and subsequently, noise) of the platform.

 

[Idontcare]

Originally posted by: harpoon84

Originally posted by: Idontcare
Furen, exactly, I'm with you on this. Sad how folks are focusing on how minimal the waste heat could be if they had bought 4 processors and proceeded to do nothing with them.

Only one number should matter when one is talking about an extreme performance product, and that is its extreme performance.

Reading all this talk over whether these quad-core platforms are the equivalent of 3 100W light bulbs or 6 100W lightbulbs makes me feel silly. What are folks buying these systems for if 6 100W lightbulbs is their metric of success?

On that note, Prescott was quite a success then?

It's better than A64 at encoding and multimedia, and while A64 was better at gaming, 'games are all GPU limited anyway' (argument used for C2D vs A64).

I think power efficiency is quite relevant, especially since it also determines the PSU requirements and cooling requirements (and subsequently, noise) of the platform.

harpoon84, I have no idea what your metric of success would be to answer your question but certainly from the $Billions worth of Prescotts sold they were apparantely good enough for something. If they were great at encoding then I am guessing that's what folks spent all that money buying them for. Not my situation to judge.

My point was supposed to be simple, really, which was that if you are a consumer who finds themself in a situation that demands they seek out the performance offered by 4x4 and QX6700 systems then I am guessing that 4-sigma of those consumers do not find themselves compelled to care about the waste heat. They are going to purchase a 4x4 or a QX6700 based on some other, presumably entirely performance based, metric of success.

If you want a computer that consumes 2 lightbulbs worth of power then you are likely the kind of consumer who is going to find lower performing computer systems are going to adequately address your performance needs, secondary metric of success, while adhering to your primary metric of success which is power consumption.

Neither consumer is superior to the other, this is not a chest beating event.

 

[DrMrLordX]

Originally posted by: Idontcare
Furen, exactly, I'm with you on this. Sad how folks are focusing on how minimal the waste heat could be if they had bought 4 processors and proceeded to do nothing with them.

Only one number should matter when one is talking about an extreme performance product, and that is its extreme performance.

Reading all this talk over whether these quad-core platforms are the equivalent of 3 100W light bulbs or 6 100W lightbulbs makes me feel silly. What are folks buying these systems for if 6 100W lightbulbs is their metric of success?

uh? The problem isn't the power consumption alone. If the QuadFX platform could actually do something worth mentioning that put it head-and-shoulders above the QX6700, then the extra power would be "worth it" to many people. Fact is that it doesn't. Under a 32-bit OS like XP, an environment in which most QuadFX systems will be run (or 32-bit Vista), QuadFX loses to the QX6700 all the while costing more money and chewing up more power. And let's be clear here that it chews up significantly more power AT LOAD. The idle power figures aren't the issue here, at least not for me.

Furthermore, the high stock vcore on the FX-74 indicates that these chips have little, if any headroom for overclocking. OC a Kentsfield to 3 ghz (which can and has been done repeatedly) and you'll whip the QuadFX system soundly, presumably even in a 64-bit OS (that few people will be using with either the QX6700 or QuadFX). The power consumption figures for the Kentsfield would look less impressive after an overclock, but the performance gain would be considerable.

Bottom line is this: if it's going to suck up more power, it'd better give me something in return. The QuadFX platform currently does not produce performance that can justify its power consumption.

For the record, the Prescott sucked because it chewed up more power/produced more heat than the Northwood and did not yield appreciable gains in performance.

 

[jazzboy]

 

[jazzboy]

it's not the damn power-hungry chipset that draws all kinds of power at idle, it's the damn 1.475v 3.0GHz Opterons that leak like a sieve even when idle.

So is 1.475v the stock VID for the FX-74 then? If so then OUCH! That'll certainly suck up power. And I've never seen a 90nm CPU sold before where the stock VID was above 1.4 so I wonder if there's also an issue with the lifespan of that CPU (esp. if you use a motherboard which overvolts).