Enthusiasts who count themselves among the AMD faithful continue to look for even the slightest performance advancement as the company’s “Spider" platform” reaches its first birthday. Several factors have held the otherwise strong value-oriented processor/chipset/graphics combination back from also achieving the highest marks in performance as well: bugs in the new processor core, severe limitations in CPU speed and overclocking capability, and the delayed release of AMD’s SB750 high-performance southbridge.
By the time the SB750 was released, most motherboard manufacturers considered the enthusiast-grade 790FX, with its 32 PCI Express 2.0 pathways divisible to support four graphics cards, old news—even though it’s still perhaps the most advanced AM2+ chipset available. But while many insisted that we try to wedge AMD’s newer mainstream-targeted 790GX integrated chipset into the enthusiast space, three brands stuck to AMD’s initial concept by combining the now mature 790FX with the delayed SB750.
The SB750 carries with it the promise of better overclocking through Advanced Clock Control (ACC), a setting many enthusiasts have guessed involves loosening certain thresholds on some of the processor’s internal "tumblers," as AMD puts it. Previous tests have shown that this feature can boost CPU overclocking capability significantly, though its effectiveness varies with individual processors.
Superior CrossFireX support remains the main reason to choose the 790FX over competing solutions, but will the SB750’s ACC feature set these apart from previous products for overclocking? Before we find out, let’s take a closer look at the new motherboards.
- Introduction
- Asus M3A79-T Deluxe
- M3A79-T Deluxe Software, BIOS, and Accessories
- Foxconn A79A-S
- A79A-S Software, BIOS, and Accessories
- Jetway HA04-Ultra
- Test Settings
- Overclocking Results
- Benchmark Results: 3D Games
- Benchmark Results: Encoding
- Benchmark Results: Productivity
- Benchmark Results: Synthetic
- Efficiency
- Too Little, Too Late

and the 9950 is cheaper then the q6600 its cheaper then about every intel processor out there (even the dual cores such as e8400 etc).
Im not saying it is a better processor then what intel has lined up, what im saying is that it is amazing value for money. and much better at overclocking then tomshardware managed to get out of it thats all
I don't know about you guys, but I haven't had a flash going wrong since a P2B board from the stoneage
Call it paranoia from a guy who changes BIOSs on an almost daily-basis on one board or another (me, not Thomas, though he does his fair share of updating, too).
Sadly, this is going to be tough hill to climb for AMD to become competitive again. I wish them luck though, the consumer is always the winner when corporations compete
Ive got a phenom x4 9950 placed on a ASUS m3a32 mvp deluxe motherboard.
With stock voltage and stock cooling ive cranked it up to 3.2ghz and it runs perfectly stable. (this was done by only adjusting the multiplier)
A friend of mine have the same setup as me but an aftermarket cooler (noctua nh-u12p) and hes overclocked it up to 3.4ghz, again running stable.
On Overclocking forums i read about people cranking this CPU up to 3.6ghz on air (noctua nh-u12p) without any wizardry.
This test is flawed and im very dissapointed about tomshardware and what i feel is an effortless test of this setup.
Im an Intel guy myself but at least im honest about intels oponents, and in this case the review should end up with AMD being amazing value for money and that the future is a bit (not alot) brighter for AMD.
and the 9950 is cheaper then the q6600 its cheaper then about every intel processor out there (even the dual cores such as e8400 etc).
Im not saying it is a better processor then what intel has lined up, what im saying is that it is amazing value for money. and much better at overclocking then tomshardware managed to get out of it thats all
Obviously Tom wasn’t interested in doing this review!……..
But I game at 1080P res. wiht a 4870 and I get over 30FPS in most of my games At max setting.
Good Review.
Here ya go Tom.
Article announced end of September. Submission deadline mid-October. DFI wasn't interested or otherwise didn't respond. DFI's inaction is unfortunate.
If that’s the case one could argue that you guys write for manufacturers and not your readers.
Another point: hasn't AMD quite clearly said it won't be able to compete with Intel high end CPU:s and hence is aiming at giving better value for money in the "mid-priced value market"? Yes I'm even quoting some words of the article, but my point is that because of this it's quite unnecessary to go on about "too little to cure AMD’s ills". Of course the AMD64 era was astonishing since AMD did the unthinkable with resources as big as Intel's pocket-money. To believe though that AMD would be able to over perform time after another is wishful thinking. It's already proven that even with a superior product AMD won't be able to get good enough sales. So if you're really interested in finding the cause of the illness, it's not to be found solely inside AMD's headquarters. Without their inventions I doubt we would have seen the light of anything in the series of Intel Core CPU:s.
There are several scenarios where I at the moment without hesitation would prefer Intel. If I did more rendering the Core i7 is a clear winner. Besides that even your Core i7 test showed that the 700/750 scales well when running more AMD/ATi graphic cards.
To choose AMD over Intel isn't stupid, it's just a question of individual computing routines, in some AMD gives great value even beyond the 4-core scenario you mention in the article. When looking through benchmarks it's necessary to evaluate it's impact on your personal computing. In many cases you get a overall figure like "this platform is XX % faster", but the reality is that might be much less because the benchmarks are irrelevant to my computing.
A long rant - and it's not a rant anyway - to convey a simple message:
- Intel has a great line up of CPU:s at the moment
- The above doesn't mean AMD "is ill" or disappointing
/A formed (with constant relapses!) narrow minded overclocker who found out that there's a lot of more fun computing stuff to do
9950 = $185 http://www.newegg.com/Product/Product.aspx?Item=N82E16819103291
Q6600 = $180 (OEM) http://www.newegg.com/Product/Product.aspx?Item=N82E16819115018
or $190 (Boxed) http://www.newegg.com/Product/Product.aspx?Item=N82E16819115017
Either way, a five dollar difference is insignificant - they're about the same price.
From where I'm standing it's the other way around really. It's the manufacturers not included that don't care about the readers. Those who sent in boards are those who care about us readers - and dfi appearently isn't one of them at this point. You can't blame toms for that really.