P55 On Boost: Five LGA 1156 Boards Between $200 And $250

[apache_lives]

The 1156 platform the Intels MIDRANGE platform, the 1366 HIGH END - stop expecting high end platform features (16 16) on a midrange platform, its stupid.

Its normal marketing procedure - premium product, premium price, if you cant afford a 1366 rig get a 1156 rig or whatever, stop complaining about a lack of features on a budget platform.

 

[snakyjake]

From the article and comments, I'm reading some big obstacles for me:

"You can't enable USB 3.0 and SATA 6.0 Gb/s at the same time."

So if I have a SATA6 SSD drive, then connect a USB3 thumbdrive or video camera....the system will operate as a USB2 and SATA3 system?

Or if I have a SATA6 SSD drive and a 16x PCIe graphics card, the system will act like a slower SATA3 SSD, 16 GPU system?

Is my understanding correct?

 

[snakyjake]

I look forward to a comparison of AMD AM3 boards.

 

[snakyjake]

I'd like to thank Tom's Hardware for providing a good review for some of us who are mainstream users, not gamers or overclockers.

 

[Crashman]

[citation][nom]redgarl[/nom]Isn't core i5 and LGA1156 socket supposed to be budget platform? Why in hell does core i5 fall into highend market?More I think about it, and more I think Intel screwed up in their goals of achieving budget range LGA1156. Yeah, it's a great CPU, but if you don't plan to SLI or crossfire two strong cards, you aren't going to get much more results than going PII and 790GX AM3... but the cost will be way more noticeable.200$ motherboards... who would want to pay that? Not me anyway... it's not the Intel cpu price that bother me, it's their motherboards![/citation]

The 1156 platform isn't high-end from the connectivity standpoint yet Intel still produces high-end i7 processors for it. If you want to learn how it might be considered for a high-end system, read the first page of the article!

[citation][nom]jcwbnimble[/nom]Vulmer, what "high end" P55 i5 setup cost you $200 less than any x58 setup? A high end P55 MB is $200-$250 and the i5-750 is $200 for a total of $400-$450. You can get a x58 MB for $250 and a i7-920 for $250 for a total of $500.For the extra $100 would gladly go with the x58 which has tri-channel DDR3 and 16x-16x PCIe native for both graphics cards.I think p55 is great for lower end systems with a $100 MB and the i5 chip, but when you start spending ~$250 for a p55 MB, you might as well go the x58 route.[/citation]

You're completely off-base with your motherboard prices. X58 motherboards at around $250 have low-end feature sets. If you want something to compare to a $200 P55 features wise, you end up paying around $300 in X58 and the $100 difference gets you 20 more PCIe 2.0 lanes, pretty much the only difference that matters when it comes to building gaming machines.

 

[NucDsgr]

Good Article on P55 motherboards for performance/upper-mainstream computing. Unfortunately, most upper-mainstream users do not need these enthusiast overclocking features. We want moderate overclocks and IDE legacy support.

Performance figures for all five motherboards essentially the same for synthetic benchmarks that matter to me for normal and overclock operation. There is little to differentiate.

What then matters is features and efficiency. I feel USB 3 or SATA 6G today is premature since implementation of these features on these motherboards is kludged and immature. This capability costs PCIe 2.0 lanes I want for graphics. Devices that efficiently (that actually) exploit these faster links are about 6 months to a year off. Thus SATA 6G or USB 3 is actually liability on until chipsets and motherboards are available that provide SATA 6G amd USB 3 with their own dedicated PCIe 2.0 lanes.

Legacy devices are important to me. I have a couple a PLEXTOR PX-716A and PREMIUM drives that work just fine. So I would choose the GA-P55A-UD3P($160) which supports legacy IDE, has moderately good overclocking, and saves me $90 over the GA-P55A-UD6 ($250) I can invest that $90 in a i860 processor.

The GA-P55A-UD3P also has a LOTES instead of a FOXCONN LGA 1156 socket for those worried about socket failures (observed in FOXCONN sockets during extremely high overclocks at 5GHz with L2 cooling)occuring during normal or moderate overclocks.

 

[Crashman]

[citation][nom]NucDsgr[/nom]Good Article on P55 motherboards for performance/upper-mainstream computing. Unfortunately, most upper-mainstream users do not need these enthusiast overclocking features. We want moderate overclocks and IDE legacy support.[/citation]

Speak for yourself. My upper-mainstream system is VERY OLD and still has absolutely zero Ultra ATA drives. If you have an Ultra ATA drive and it's not worthless by now, it's probably because you bought the wrong drive long after SATA went mainstream. Don't try to assign any of your unique problems to the broader market.

And, BTW, the overclocking comparison isn't extreme since it's done on air cooling.

[citation][nom]NucDsgr[/nom]The GA-P55A-UD3P also has a LOTES instead of a FOXCONN LGA 1156 socket for those worried about socket failures (observed in FOXCONN sockets during extremely high overclocks at 5GHz with L2 cooling)occuring during normal or moderate overclocks.[/citation]

That's probably because LOTES sockets are cheaper. LOTES was just lucky that Foxconn, which normally makes the highest-quailty sockets, had an issue. But I've spent most of the last two months telling people that this is a nonissue anyway.

They don't overclock at those extreme voltage levels. They don't even have the extreme cooling needed to reach those extreme voltage levels.

Dead voltage regulators are a far more common problem, and one that can affect moderate overclocks on less-expensive motherboards.

 

[Stardude82]

Sort of a pointless segment in my opinion. I don't see any improvement over the Sub-$200 boards in terms of overclocking and the extras to me are uninspiring.

 

[vulmer]

[citation][nom]jcwbnimble[/nom]Vulmer, what "high end" P55 i5 setup cost you $200 less than any x58 setup? A high end P55 MB is $200-$250 and the i5-750 is $200 for a total of $400-$450. You can get a x58 MB for $250 and a i7-920 for $250 for a total of $500.For the extra $100 would gladly go with the x58 which has tri-channel DDR3 and 16x-16x PCIe native for both graphics cards.I think p55 is great for lower end systems with a $100 MB and the i5 chip, but when you start spending ~$250 for a p55 MB, you might as well go the x58 route.[/citation]

I picked up an Asus Maximus III Formula for $200
An i5 for $200
And 4gb(2x2gb) 1600 gskill memory for ~$90

My Ideal x58 would have been i7 920, which is priced at ~$290
Mobo would be an Asus P6T which is ~$250
6gb (3x2gb) 1600 gskill memory for ~$150

So i5 = 490
i7 = 690

$200 difference...

I think the maximus is ~$250 now, so only $150 difference if you bought today. Thats still quite a big chunk towards a gpu...

 

[avatar_raq]

Corrections:
1. In the feature comparison table: sub-$200 motherboards--->$200-250.
2. Near Apple iTunes benchmark 7P55D Evo---> Deluxe.

 

[NucDsgr]

[citation][nom]Crashman[/nom]Speak for yourself. My upper-mainstream system is VERY OLD and still has absolutely zero Ultra ATA drives. If you have an Ultra ATA drive and it's not worthless by now, it's probably because you bought the wrong drive long after SATA went mainstream. Don't try to assign any of your unique problems to the broader market..[/citation]

Same fashionably obnoxious troll I corresponded with 6 months ago over an overheated microatx build.

I think this article is a decent review of $200 to $250 P55 motherboards for performance/upper mainstream computing. I found it useful though you obviously did not. This is most unfortunate. This article convinced me I could do better with a lower cost motherboard in the $130 to $160 price range (GA-P55A-UD3P) than the boards reviewed in this article.

And to Hell with SATA 6G and USB 3 until chipsets and motherboards are released which provide dedicated PCIe 2.0 lanes.

If you want to save money with Lynnfield/P55 over that of the Bloomfield/X58, one must consider motherboards in the $130-$160 range with an i750 or i860. This is plenty powerful (and versatile for my needs)

I am speaking for myself (and for others like myself). There are plenty of people who bought so-called legacy optical drives in 2004/2005 that work just fine. I own Plextor PX-716A and PREMIUM drives that continue to offer excellent performance. When I bought these drives SATA optical drives were just being introduced to the market. The performance of modern SATA optical drives, though exhibiting better performance, is not so compelling to toss a functioning $100 IDE optical drive and shell out an extra $50 or so for a SATA optical drive. That's $150 wasted in the opinion of many people. I went over to SATA hard-drives in 2005 when these became available. These are definitely better than IDE harddrives replaced.

BTW check out the article in Anandtech "P55 Extreme Overclockers: Check your sockets!" about FOXCONN Socket 1156 failures during really extreme overclocking while LOTES sockets did not fail. I agree this is an non-issue for stock speeds and moderate overclocks. Amazingly there were people who turned in brand new FOXCONN socketed Mobos for EVGA boards which had LOTTES sockets (and spending $100-$150 more!) because of fears these failures might happen at stock voltages and frequencies. This unreasonable response is foolish and mindless.

 

[Crashman]

[citation][nom]NucDsgr[/nom]Amazingly there were people who turned in brand new FOXCONN socketed Mobos for EVGA boards which had LOTTES sockets (and spending $100-$150 more!) because of fears these failures might happen at stock voltages and frequencies. This unreasonable response is foolish and mindless.[/citation]

It's also amazing that at the lower-end, some people will chose a motherboard with bad VRM and LOTES socket over one with a good VRM and Foxconn socket, for fear that the socket will fail when its obvious the VRM will fail first.

 

[akula2]

I don't quite agree with this review on Gigabyte P55A-UD6. Most of the other top review sites clearly placed UD6 ahead of the Asus, how come it's different in this review? Well, it's nothing wrong with the Gigabyte's 333 implementation; perhaps Gigabyte managed the trade off in somewhat better way even if we compare with the updated Asus MOBO line.

Graphics card runs at PCIe 8x if you use USB3 (OR) SATA3 - Should I really care if am a using a good single card?

 

[Crashman]

[citation][nom]akula2[/nom]I don't quite agree with this review on Gigabyte P55A-UD6. Most of the other top review sites clearly placed UD6 ahead of the Asus, how come it's different in this review?[/citation]

You mean ahead of Asus in Performance? Tom's sets the BCLK to 133 MHz, some other sites use "Auto": If Gigabyte overclocks to 136.3 MHz at "auto" and Asus doesn't, that makes it appear like Gigabyte performs better even if it doesn't. Tom's doesn't play those games, taking great effort to eliminate "Cheating".

Look at the performance comments in the conclusion. Tom's specifically cancels-out Asus' 0.28% overclock at the 133 MHz setting in those comments. Gigabyte's 2.7% "default overclock", if allowed, would have been nearly TEN TIMES as bad. Rather than heap on ten times the criticism for cheating, Tom's takes the "high road" by setting 133 MHz manually.

So to get straight to your question, the reason Tom's numbers are different is because, by doing what it can to prevent cheating, Tom's produces a higher degree of accuracy. This is reflected in the article quote "Since our performance charts reflect base clock massaging more than real-performance differences, we can’t quite base our final analysis on these."

If you don't like accuracy, feel free to disaggree!

 

[Hap]

what happen to reviewing Supermicro line up as well with these

 

[Hap]

what happen to reviewing Supermicro line up as well with these

 

[Crashman]

[citation][nom]Hap[/nom]what happen to reviewing Supermicro line up as well with these[/citation]

Super Micro makes workstation boards in this class, which would have to be compared to workstation boards of other companies.

 

[akula2]

[citation][nom]Crashman[/nom]Tom's doesn't play those games, taking great effort to eliminate "Cheating".Look at the performance comments in the conclusion. Tom's specifically cancels-out Asus' 0.28% overclock at the 133 MHz setting in those comments. [/citation]

I didn't say anything like Cheating. Point is about fairness.

#1 Tom should have added P55-UD6 but NOT P55A-UD6 in this review. Or, Tom should have added the update Asus product line like P7P55D and P7P55D-E.

#2 Earlier I narrowed on Asus and Gigabyte P55-UD4P but I dropped all these boards due to mysterious Foxconn issues and chosen P55A-UD4P due to solid reviews on this MOBO. Now, it's gonna be a big dilemma on deciding 21 workstations I've planned based on these MOBOs!!!

@Gigabyte's 2.7% "default overclock", if allowed, would have been nearly TEN TIMES as bad.

Can you prove x10 claim?

 

[Crashman]

[citation][nom]akula2[/nom]I didn't say anything like Cheating. Point is about fairness. #1 Tom should have added P55-UD6 but NOT P55A-UD6 in this review. Or, Tom should have added the update Asus product line like P7P55D and P7P55D-E. #2 Earlier I narrowed on Asus and Gigabyte P55-UD4P but I dropped all these boards due to mysterious Foxconn issues and chosen P55A-UD4P due to solid reviews on this MOBO. Now, it's gonna be a big dilemma on deciding 21 workstations I've planned based on these MOBOs!!! @Gigabyte's 2.7% "default overclock", if allowed, would have been nearly TEN TIMES as bad.Can you prove x10 claim?[/citation]

Fairness? Gigabyte persistently asked Tom's NOT to use the P55-UD6 on-hand and instead wait for the P55A-UD6 to arrive, which wasn't a problem since the arrival was a couple weeks before the article was published. So it can't be unfair to Gigabyte. Are you saying it's unfair to Asus?

Asus picked the boards it sent, Tom's got the P7P55D Deluxe and the P7P55D-E Premium. The Premium hasn't been reviewed yet because it's in the next price class higher than this one. So I think your concepts of fairness are what need to be made...more fair.

As for the 10x statement, it's math. I can prove that 2.7% is almost 10x as much as 0.28%, That's because 10x 0.28 is 2.8, and 2.7 is almost as much as 2.8.

But I haven't done any mathematical proofs since I graduated. The statement about 2.7 being almost 2.8 is widely accepted as true, so instead I ask, can you prove 2.7 is NOT equal to 0.27 time 10?

 

[akula2]

[citation][nom]Crashman[/nom]So it can't be unfair to Gigabyte. Are you saying it's unfair to Asus?

@can you prove 2.7 is NOT equal to 0.27 time 10?[/citation]

Gross unfair to the overseas customers like me who read confusing reviews, especially when we pumping budget $40,000 plus on building 21 workstations, includes 31 LCD Monitors (Foxconn socket, RAM types, PCIe lanes and now this review!!)

You should be knowing this world is filled with full of such dumb marketing numbers (hate to say it being an employer), real performance is what matters and am no fan of Gigabyte or Asus but I need to trust in my product life and performance. That's all it matters, so I've chose P55A-UD4P. In fact, am more surprised adding an Intel MOBO in this review (most of the performance world out there doesn't give a dime to the Intel boards!).

 

[Crashman]

[citation][nom]akula2[/nom]Gross unfair to the overseas customers like me who read confusing reviews, especially when we pumping budget $40,000 plus on building 21 workstations, includes 31 LCD Monitors (Foxconn socket, RAM types, PCIe lanes and now this review!!)You should be knowing this world is filled with full of such dumb marketing numbers (hate to say it being an employer), real performance is what matters and am no fan of Gigabyte or Asus but I need to trust in my product life and performance. That's all it matters, so I've chose P55A-UD4P. In fact, am more surprised adding an Intel MOBO in this review (most of the performance world out there doesn't give a dime to the Intel boards!).[/citation]

These reviews are written for people who overclock, and proves that Gigabyte and Asus boards have the same performance when set to the same clock speeds. That's the best reason why Tom's does as much as possible to make sure all boards are running at the SAME CPU speed when it compares performance.

As for your statement concerning Intel motherboards, do you consider that fair? Intel had one of the best boards in the bunch.

I think your priorities are totally backwards if you're disqualifying motherboards for having Foxconn sockets because they can't stand up to a 400W CPU load, when you'll probably never push your own past 200W. But like you said, it's probably information overload affecting that type of decision.

 

[akula2]

[citation][nom]Crashman[/nom]As for your statement concerning Intel motherboards, do you consider that fair? Intel had one of the best boards in the bunch.[/citation]
I started learning with my own IBM 80286 computer from 1984, sixth grade! Most of Intel MOBOs market/price themselves between Novice to Performance territory whereas in high performance Intel lacks compared to Gigabyte and Asus. In my experience, observed both brands dominated from right from those Pentium III Tualatin processor days (from 2001 onwards)! Again, all 'extreme' things Intel does are untouchable to 99% folk out there...prefer to invest in a few Apple Macs for my Pharma business requirements.

[citation][nom]Crashman[/nom]I think your priorities are totally backwards if you're disqualifying motherboards for having Foxconn sockets because they can't stand up to a 400W CPU load, when you'll probably never push your own past 200W. But like you said, it's probably information overload affecting that type of decision.[/citation]
http://www.xtremesystems.org/forums/showthread.php?s=902d6987bc21f54cb1bfc00676c3c8a2&t=234723

Such a thing could cost me in hundreds of thousands dollars (halt in computational processing operations etc). No room for any risk in my business!

 

[akula2]

Forgot to add, am actually phasing out 5+ year old 20 Intel 875P workstations. I going through a solid plan for this specific upgrade (branded companies have dropped a high price bomb on me for this project lol), must be completed by the end of February 2010. First, I need to build one workstation and test it to the fullest for 1-2 weeks (driving the Graphics, simulation experiments with 3 LCD monitors and many more) and then GO green for the remaining 20 workstations. You can see clearly, my time is somewhat tight and can't afford any lapses.

 

[Crashman]

That's the thing, if you're looking to build a performance Windows system with similar features to Apple, you're looking at Intel and putting aside its overclocking capabilities.

[citation][nom]akula2[/nom]http://www.xtremesystems.org/forum [...] 2&t=234723Such a thing could cost me in hundreds of thousands dollars (halt in computational processing operations etc). No room for any risk in my business![/citation]

At 1.65V core? Tom's Hardware has VOLTAGE REGULATORS dying at 1.40V core, and you're still worried about sockets? Like I said, completely backwards priorities.

 

[pengivy]

I want to know what version of Prime95 you test?
Is Prime95 v25.9.4 or Prime95 v25.11.2?
I also want to test my system, thanks a lot~~