Slashdot Mirror
Core 2-Compatible Chipsets Compared
theraindog writes "Intel's Core 2 Duo is clearly the most attractive processor on the market, but which motherboard chipset is it best paired with? The Tech Report has rounded up four of the most common Core 2-compatible chipsets on the market to find out. The chipsets' features are compared and their performance is tested in a wide range of application, peripheral, and even power consumption tests. One emerges as a lemon, two as solid options, and the other as a clear winner." From the article: "Some fanboys still stubbornly cling to their favorite underdog, but most enthusiasts have seen the light and are looking at Core 2 for their next upgrade or system build. The prospect of a Core 2 system build can seem a bit daunting for enthusiasts who have spent years focused solely on the Athlon 64. Core 2 processors need new motherboards for those switching from the Athlon 64, and that requires navigating a whole new world of core logic chipsets. Since the Core 2 processor relies on the chipset for its memory controller, one's chipset choice can also have a much more profound impact on performance. "
"Some fanboys still stubbornly cling to their favorite underdog..."
Gee, that doesn't sound like the author had an opinion up front. No sir.
Fill in your four or five-letter word of wisdom here _ _ _ _ _.
From the article:I can't understand why Nvidia would drop a unique feature like hardware TCP/IP acceleration.
Might I suggest it's because this acceleration has been found to cause serious data corruption on some NFORCE4-based AMD motherboards? I'm surprised the authors weren't aware of this problem.
Myself, I'm wary of NFORCE4 for other reasons. I've recently found that my quad-core Opteron box, built on a Tyan S2895 (K8WE) mobo with an NFORCE4 chipset, will undergo spontaneous data corruption under Linux when I transfer large (>GB) files. I'm not the only one who's had this problem; looks like there's a bug in the NFORCE4 SATA controller. Caveat Emptor.
Tubal-Cain smokes the white owl.
Thanks for the 21 pages...Concusion:
Conclusions
The landscape of Core 2-compatible chipsets is an interesting one. On one hand, you have tried and true chipsets like the 975X Express and nForce4 SLI X16 appearing on boards that have been updated to support Core 2 processors. These chipsets aren't new or particularly flashy, but they're proven designs that don't sacrifice performance when compared with the latest and greatest core logic.
Of course, the Core 2 chipset market isn't all retro refits. Nvidia's nForce 570 SLI is new, at least in name. Unfortunately, the chip's features make it look more like a minor update to the aging nForce4 SLI than a legitimate member of the nForce 500 series. The discrepancies between the chipset's features and those offered by the nForce 570 SLI for AMD processors are striking, and for the life of me, I can't understand why Nvidia would drop a unique feature like hardware TCP/IP acceleration. Extra features have long been a tenet of nForce chipsets, and on that front, the 570 SLI comes up well short.
Compounding the 570 SLI's disappointing feature set is comparatively high memory latency with two-DIMM configurations. This particular problem might be a quirk of the Asus P5NSLI motherboard we used for testing, but it's a retail board with a production BIOS--a board Nvidia itself provided for review.
Intel P965 Express
September, 2006
Unlike the nForce 570 SLI, which feels like little more than an uninspired retread, Intel's P965 Express still has that new chipset smell. It also has a number of new features, including Fast Memory Access optimizations and additional south bridge Serial ATA RAID ports. The P965 also has relatively low power consumption, competitive performance, more robust Matrix RAID options, and evolving CrossFire support that should be complete by the end of the year. No wonder this chipset has proven so popular with motherboard manufacturers, who are building everything from budget $100 wonders to high-end enthusiast boards based on it.
The Intel P965 Express chipset's mix of features, performance, and power consumption make it the perfect partner for Intel's new Core 2 processor and our Editor's Choice. Really, it's an easy call to make. The P965's upcoming CrossFire support kills the one reason we might have recommended the 975X. If SLI's your bag, the nForce4 SLI X16 is a solid--albeit power-hungry--option, but the more attractive nForce 590 SLI should be just around the corner. The nForce 570 SLI, meanwhile, doesn't even measure up to its own name, let alone to the P965
I'll save you the time of scrolling thru all 21 pages by skipping right to the conclusion...
The Intel P965 Express chipset's mix of features, performance, and power consumption make it the perfect partner for Intel's new Core 2 processor and our Editor's Choice. Really, it's an easy call to make.
The core2 is fast in 32bit mode and certainly a step in the right direction for intel. Kudos to them, but fie on their 64 bit support. In deference to the Itanium, they think of it as existing to provide extended memory support, and only because AMD was flanking them. Two important factors drag down core2 64 bit performance: micro-ops fusion, which welds multiple trivial ops into a single macro op, works on 32 bit instructions only, and they support DMA on only memory addressable with a 32 bit pointer. The message from intel is clear: for 64 bit performance, buy an Itanium or an opteron.
Yes, 'fanboys'
Just because I'm not rushing out to buy a Core 2 Duo system to replace or upgrade, and instead 'cling' to my AMD64 system, it doesn't make me an AMD fanboy. Core 2 Duo is looking like a fucking superb processor family, but I think I'lll get a bit more life out of my current system before diving into a complete upgrade / replacement of my current hardware thank you.
As a side note; does anyone have any info on what AMD are planning, if anything, to compete against the Core 2 Duo in the near future? I read something a while back about them switching to 65nm at a new fab, but I don't follow the processor market that closely anymore.
She's built like a steak house, but she handles like a bistro....
In virtually all MB reviews with the same chipset, they are usually all within 5-7% of each other in the benchmarks. This usually doesn't translate into meaningful "sitting in front of the keyboard" performance differences. It's like horsepower in cars - it's hard to tell 195hp from 205hp when you're behind the wheel. (and yes, I like to compare apples to oranges...)
and they support DMA on only memory addressable with a 32 bit pointer.
Indeed. Hence the software bounce buffer in Linux, to make up for the lack of an IOMMU. And while I'm extremely pleased with my recently-purchased E2700 Core 2 Duo box, I wonder how well the Core architecture will fare in quad/oct-core land, with it's memory access issues. The lack of both IOMMU and (more fundamentally) a per-core or per-die memory controller seriously hampers Core's ability to scale.
Tubal-Cain smokes the white owl.
"Some fanboys still stubbornly cling to their favorite underdog, but most enthusiasts have seen the light and are looking at Core 2 for their next upgrade or system build." Nowhere does it say when that next upgrade or system build has to come, or that anyone needs to upgrade from AMD64.
For gaming, which requires the most performance, AMD is soundly beaten. Period. That doesn't mean they won't come out with something to one-up Intel in the future, but right now only a fanboy would recommend AMD if performance is the number one criteria.
If you move away from performance as the only criteria, I think AMD competes well with the cost-to-dollar performance. HOWEVER, if you are willing to overclock, even the sub-$200 Core 2 Duo will outperform any desktop AMD chip you can buy-- at any price.
It's not a matter of fanboy-ism. It's a matter of numbers. Right now Intel wins on performance. That's just a fact. However, AMD could very well get back into the mix and make this argument mute in the near future.
"Politicians find new names for institutions which under old names have become odious to the people."
As a side note; does anyone have any info on what AMD are planning, if anything, to compete against the Core 2 Duo in the near future? I read something a while back about them switching to 65nm at a new fab, but I don't follow the processor market that closely anymore.
AMD has a couple of interesting products comming out:
- First is 4x4, comming out in 4Q2006. This is essentially a dual-socket platform designed for the high-end desktop and low-end workstation/server market. This isn't a product for everyone, but it will make for a very price attractive dual-socket workstation. To start with, it will support two dual-core chips for a total of four cores. AMD has stated that later, when they release quad-core chips, 4x4 will support two of them for a total of eight cores. It's a niche market, but a neat idea.
- Second, AMD is releasing a new core in 2Q2007. This core has double the number of FP (floating point) pipelines, double the L1 cache bandwidth, larger reorder buffers, a L3 cache, and will come in dual and quad-core versions. This chip is going to be a beast, and will be supported in any current socket AM2 mother board. For more, read this: HardOCP, HardOCP. This new core is the direct answer to Core2/conroe, and I expect it to be a good one. It looks really good on paper, and after seeing AMD's delivery of K8, I expect the new core to live up to the hype.
I am a viral sig. Please help me spread.
Two of the chipsets tested support SLI, the nForce 570 SLI and the nForce4 SLI X16. http://techreport.com/reviews/2006q3/core2-chipset s/index.x?pg=10
Your precious P5N32-SLI SE is even tested in the article! So what was your problem again?
Criteria's plural. You want criterion.
Why do the new Intel chipsets have just one P-ATA channel, if any at all? It is ridiculous. 95% of all optical drives are P-ATA, and P-ATA hard drives as of yet are just as fast, if not faster thanks to more mature drivers and technology, than their SATA counterparts. What do they expect people with 2+ perfectly fine last-generation PATA hard drives to do when upgrading to Core 2 Duo? Getting a separate PCI controller (as PCI-E x1 ones are still rare) to already expensive C2D motherboards with just 2 PCI slots, both of which a person could have a much better use for?
Intel had what, 5 years to top K8 and 5%-10% better speeds was the best they could manage?
Well, it depends. Core 2 is really good at 32-bit single threaded tasks that fit inside a 4MB cache or have sequential memory access patters. On these workloads Core 2 will wipe the floor with AMD's K8, hands down. Just look at SuperPI scores for an example.
Here's the question though: What happens with heavily threaded 64-bit tasks that use much more than 4MB of memory in a random access pattern? AMD's K8 wipes the floor with Core 2, hands down.
There is a reason most gamers/encoders buy Core 2 for their desktops and many businesses and universities buy K8 Opterons for their servers and HPC centers.
I am a viral sig. Please help me spread.
I was dissapointed that this review didn't include any G965 reviews.
If you want the new Linux XGL/AIGLX/Compiz functionality to work out of the box with no binary drivers, a G965 board may very well be what you are looking for.
The G965 chipset includes the new GMA X3000 graphics core, which is the only DX10 feature level graphics solution with FLOSS Linux drivers: http://intellinuxgraphics.org/
The new C2 stepping is rumoured to solve many of the performance problems of the first stepping, and although probably still not suitable for FPS gaming under Linux, other less demanding games and desktop users needs could be well met.
but I'm still sticking with my Athlon 64, thanks. Besides, I can't even tell which damn Intel processors are even 64-bit capable – at least AMD makes a clear naming distinction. (I may be wrong here, of course... disclaimer, I used to be an Intel guy myself, but I've since changed.)
Creative misinterpretation is your friend.
I keep seeing this in relation to Intel processors "They aren't really 64-bit! Their 64-bit is t3h slow!" Ok fine but let me ask you two things then:
1) What's your source? Do you have some Intel technical docs that talk about how their chips are gimped in 64-bit? Remember: A post from some guy on some web forum isn't a legit source.
2) If that's the case, why do they perform well in 64-bit mode, in relation to 32-bit mode? You take things like, say, Prime95. Install XP 32-bit and XP-64-bit on the same system and do some benchmarks. You find that Prime95 gains a little bit of speed when both it and the OS are 64-bit compared to when both are 32-bit. So how's that happen? Isn't that counter-evidence to a poor 64-bit design?
I'm not claiming to have the answer here, just saying I see this "Pentium (and now Core2) sucks at t3h 64-bit, buy AMD!" trotted around a lot, but I see very little backing for it. So please, provide some relevant docs. What I'd like to see:
1) Intel tech docs describing the limitations of 64-bit support on the Core 2.
2) Benchmarks showing a non-trivial loss of speed in 64-bit. This means you take a 32-bit OS, 32-bit app, bench it, then a 64-bit OS and 64-bit app on the same system and rebench it.
3) Benchmarks showing the Core 2 vs Athlon on a 64-bit app, 64-bit OS.
If you haven't looked at things like that, you might want to reconsider the confidence with which you state your point of view. I'm not saying it's wrong, I'm just saying that there appears to be counter evidence so you need to get some backing to try and straighten things out.
The worst parts about these fights is that it's millions of words in thousands of forms and blogs over what amounts to a 2% performance difference. BFD.
I read the internet for the articles.
4x4 is AMD's marketing platform for dual-processor, dual-core (4 CPU) systems with dual-board, dual-processor graphics (4 GPU).
Hence 4x4. But it's 8 processor cores, not 16. But heck, when you buy a 4x4 truck do you expect 16 wheels?
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
Intel is likely to make some inroads in the server market except... well, except unfortunately Core Duo doesn't scale as well once you go to quad cpu setups due to the memory bottleneck. So their only real claim to fame is power use. Power is extremely important in the long term, but I don't see anyone rushing to replace all their AMD boxes with core duo just for that when they know AMD will come up with a power-competitive design in fairly short order.
The real problem Intel has is their inability to compete with Hypertransport. AMD is already pushing hard to make it a defacto standard for chip interconnect. Intel is working on their own solutions to the problem, but they are not hitting on all cylinders yet.
If anything is going to drive machine replacement in today's market, it is going to be the new ultra-fast PCI bus technologies. PCI has needed an upgrade for a long, long time. Nothing else will have much of an impact. GiGE is already faster then most hard drives so there isn't going to be much of a consumer push for 10GiGE. Cpu's are already fast enough and machines are already quiet enough. We are a far cry from the old days where every new advance doubled the performance of the previous year's boxes. In today's world magazines proclaim victory and tell people to trash their old machines for barely a 10% improvement, but unless there is a huge improvement in video technology even game players have no real reason to do so any more. The connection to the video card is the only thing left for which significant improvements can drive machine replacement.
-Matt
Systems Involved
Dual P4 Xeon 3.4GHz (will refer to this as "P4")
Dual Core 2 Duo Xeon 5160 3.0GHz (will refer to this as "Core2")
Test suite for our software in Studio .NET 2003
P4: 12 hours
Core2: 5 hours
Compiling our software
P4: 9 hours
Core2: 3 hours
Source code check of our software (using Lint)
P4: 6 hours
Core2: 1hr 20min
I'm not an Intel fanboy but I am understandably smitten with Core 2.
Here in the Netherlands, if you want to buy a processor for a DOY project, AMD is always cheaper, even if Intel says it's got AMD beaten. I can buy an X2 starting at 150,-, while the Intel already is at 220 minimum. That's 70 euro's that I'd rather spend in the bar or at a restaurant or three. Ok, it's much faster then the 3800 I got now, but I've still got a plenty fast dual core. It will probably beat any non-Intel Core 2 Duo out there. And it's a bit like David against Goliath as well. Intel isn't going away, so to have a healthy marketplace, we need AMD.