X38 Chipset

[Gary Key]

Originally posted by: dinos22
G'Day Gary

How's things

good to see some early numbers. Can you put me out of my misery and put up some SuperPi numbers at say 3.6Ghz and put some posts up on memory settings with maxmem if you can

waiting another few weeks is waaay to long to find out for myself hahaha

what's the go with X48 chipset?????

dinos22

With the engineering board - it is a over 2 seconds faster than the P35 at 8M with a Q6600 at 9x400, DDR2-1200, 5-5-5-18 and around 1 second faster at 8M with the same components and BIOS settings at stock with with memory at 1066, 4-4-4-12. I will have full results including 32M next week with a retail board and BIOS.

 

[Gary Key]

Originally posted by: helix2

Originally posted by: Gary Key

Add to all this mess, the new G92 graphics cards from Nvidia with DirectX10.1 support. While 8800GTX's now still cost near $600.

Do not count on the G92 until sometime next year (probably early spring) if that helps the budget. You can expect the price slides on the GTX cards to occur over the holiday season based on current chip/memory prices to the manufacturers for Q4 production rates. There is no need for NVIDIA to release the follow-up until AMD is ready with the R700 series.

Any word on when their next sli chipset will be out? hoping that the next one is less buggy than 680i/650i

Looking at late November right now, could get delayed a little depending on the AM2+/Phenom launch and QC cycle.

 

[SerpentRoyal]

-2.0 seconds @ 8M SP would probably translate to about -0.3 second @ 1M SP. That's a boost of 30 to 40MHz core speed...probably not that big of a deal for me.

 

[Sunbeam65]

I'm building a $4K-$6K rig and I think I'm going to go with the ASUS Blitz Formula (Special Edition), unless the ABIT IP35 reviews are super good.

It makes me sad not to go with an NVIDIA 680i SLI dual video solution such as EVGA A1 but there's too many rumors of NVIDIA chipset instability.

IS THIS STILL TRUE???, or has NVIDIA 680i chipset quality stabalized?

Am I making a good decision here?

 

[Sunbeam65]

Oops, Please disregard, I'm posting same question on more recent thread.

 

[asdftt123]

I've been contempating whether or not to purchase this board or not...I was pretty sure I wanted to in the beginning, but now I'm really not. I wanted X38 mainly for its PCI-e 2.0 capabilities so I could upgrade to a new GPU in the future and not worry about bandwidth. Because Nehalem is actually closer than we think and this chipset will be obsolete since it won't support Nehalem, I'm beginning to think it's not worth it for me. I'm not a heavy overclocker and I don't plan to push my rig to its limits, but I would like to do some light OCing with the Q6600 (maybe 3.2ghz), and I don't want DDR3. Should I even consider shelling out the extra $100 over a P35 board or should I get a Gigabyte P35 board? Thanks.

 

[Nickel020]

Get a P35 and spend the 100$ on a better CPU or video card. But wait a few days for the first X38 reviews - maybe it's the *miracle* chipset we've all been waiting for that will allow everyone to overclock to 5GHz at 1,0V - who knows? (Point is, we don't know how much those boards will be and if an by how much they will be better exactly)

 

[SerpentRoyal]

Originally posted by: asdftt123
I've been contempating whether or not to purchase this board or not...I was pretty sure I wanted to in the beginning, but now I'm really not. I wanted X38 mainly for its PCI-e 2.0 capabilities so I could upgrade to a new GPU in the future and not worry about bandwidth. Because Nehalem is actually closer than we think and this chipset will be obsolete since it won't support Nehalem, I'm beginning to think it's not worth it for me. I'm not a heavy overclocker and I don't plan to push my rig to its limits, but I would like to do some light OCing with the Q6600 (maybe 3.2ghz), and I don't want DDR3. Should I even consider shelling out the extra $100 over a P35 board or should I get a Gigabyte P35 board? Thanks.

Others here didn't have any issue pushing the Q6600 north of 3.8GHz using the $70 Abit IP35-E. The big wildcard is the quality of your CPU. There's no guarantee when you overclock a CPU.

 

[bryanW1995]

Originally posted by: Gary Key

Originally posted by: nrb

Originally posted by: Gary Key
[Everest Results - X38 Stock]

How about posting thats somewhere where I can actually see it? I have to say, this practice of posting images on Anandtech in such a way that they don't make it past firewalls was already getting old about three years ago. Any chance you guys might do something about that? Ever?

See if this works -

X38 - Q6600 OC - Air - Vista - 4GB - DDR2-1200

I have not pushed it yet, but 3.6GHz on air with 1.375V and completely stable under Vista with 4GB of memory at DDR2-1200 is an indicator of a fairly good Q6600 and probably motherboard.

have you tried that q6600 in other boards for comparison? that does look extremely good.

 

[dinos22]

Originally posted by: eva2000

Originally posted by: Gary Key
It is possible for Quad-CrossFire support on the X38 based on the amount of PCI-e lanes available and controller design, whether anyone will do it up front is a best guess at this time.

woah we need psu with 4x 8pin PCI-E hehe

future is looking expensive but fun

hahahha eva you forum slut you

maybe we'll just need to use multiple PSUs i guess

 

[dinos22]

Originally posted by: Gary Key

Originally posted by: dinos22
G'Day Gary

How's things

good to see some early numbers. Can you put me out of my misery and put up some SuperPi numbers at say 3.6Ghz and put some posts up on memory settings with maxmem if you can

waiting another few weeks is waaay to long to find out for myself hahaha

what's the go with X48 chipset?????

dinos22

With the engineering board - it is a over 2 seconds faster than the P35 at 8M with a Q6600 at 9x400, DDR2-1200, 5-5-5-18 and around 1 second faster at 8M with the same components and BIOS settings at stock with with memory at 1066, 4-4-4-12. I will have full results including 32M next week with a retail board and BIOS.

hey mate

that's interesting. It seems it's not that much quicker afterall......man i can't wait to get all this testing under way....my 4GB DDR3 Corsair 1600MHz kit's crying all the time.....begging me to find a half decent board with all the crap i've been sticking it in so far hahah

 

[Nemesis 1]

Originally posted by: dinos22

Originally posted by: Gary Key

Originally posted by: dinos22
G'Day Gary

How's things

good to see some early numbers. Can you put me out of my misery and put up some SuperPi numbers at say 3.6Ghz and put some posts up on memory settings with maxmem if you can

waiting another few weeks is waaay to long to find out for myself hahaha

what's the go with X48 chipset?????

dinos22

With the engineering board - it is a over 2 seconds faster than the P35 at 8M with a Q6600 at 9x400, DDR2-1200, 5-5-5-18 and around 1 second faster at 8M with the same components and BIOS settings at stock with with memory at 1066, 4-4-4-12. I will have full results including 32M next week with a retail board and BIOS.

hey mate

that's interesting. It seems it's not that much quicker afterall......man i can't wait to get all this testing under way....my 4GB DDR3 Corsair 1600MHz kit's crying all the time.....begging me to find a half decent board with all the crap i've been sticking it in so far hahah

Hi! Dinos. Explain something to me please. How is it that people in power( web masters) Keep saying wait for newer steppings of K10 they will be great. Yet none are saying Penryn is using A0 stepping on a totally new process with tweaks. And is doing great. What happens when Penryn B0 steppings come out and than B1 steppings. Amd gains from new steppings on an old process using soi. Yet Intel will see NO improvements with a totally new process tech . Using A0 steppings . Sure are a lot of AMD fanbois using Intel cpu's . Sheep in wolves clothing . Because in the last year the only thing that has been said about AMD Vs. Intel that I find 100% true . Is Intel is the preditor here, That is a fact. But I will go with the wolf on this one as AMD lacks a good sheppard.

X38 isn't for the faint hearted its for the pure enthusiast . Because its the only way well see A 15-20% performance increase over c2d. Using a penryn cpu.
X38 is for Penryn. X38 is were penryn will shine . X38 with penryn along with multipol drives in raid 0 is were the real performance will come from. X38 = $$$$
No need to say the benefits aren't worth the extra price as thats for people to make choice their $$$ their choice. Just because MANY won't spend for ultra performance no need to cut that extra cost because your unwilling to spend the cash for 20% more performance.

Now this only applies to Intel . If AMD K10 is 20% faster than C2D than and only than will more people be interested in gaining that insignifecant 20% performance gain . LOL!

 

[bryanW1995]

you won't be praising intel quite so much when amd goes bankrupt and you have to pay $1500 for a 2.0 ghz nehalem in 2011.

 

[Nemesis 1]

Originally posted by: bryanW1995
you won't be praising intel quite so much when amd goes bankrupt and you have to pay $1500 for a 2.0 ghz nehalem in 2011.

Were is proof Intel raises prices if AMD gets b-it-ch slapped. I have seen AMD raise prices on cpu whewn they lead in performance.

Show me were in the last 30 years intel has raised prices on new process of Cpu . NOT counting the extreme versions.

 

[KHarvey16]

People just assume Intel will raise prices if AMD goes under. I'm not really sure why, given Intel is already making plenty of money at current pricing and would only make more if AMD weren't around. And if Intel goes crazy with their pricing, they'll leave a hole in the market which someone will gladly fill. They aren't that foolish.

 

[Heidfirst]

Originally posted by: KHarvey16
And if Intel goes crazy with their pricing, they'll leave a hole in the market which someone will gladly fill.

like whom?
You don't come from nowhere to compete against someone like Intel overnight & without oodles of $$$.

Intel had to drop prices to be competitive with AMD when AMD was ascendant & you can bet that if they didn't have a viable competitor they would creep up again.

 

[KHarvey16]

Originally posted by: Heidfirst

Originally posted by: KHarvey16
And if Intel goes crazy with their pricing, they'll leave a hole in the market which someone will gladly fill.

like whom?
You don't come from nowhere to compete against someone like Intel overnight & without oodles of $$$.

Intel had to drop prices to be competitive with AMD when AMD was ascendant & you can bet that if they didn't have a viable competitor they would creep up again.

Of course they will creep up, but things won't be anywhere near as bad as some suggest. Trust me, if Intel was pricing themselves out of the mainstream someone will swoop in and take advantage.

 

[dinos22]

Originally posted by: Nemesis 1

Originally posted by: dinos22

Originally posted by: Gary Key

Originally posted by: dinos22

Hi! Dinos. Explain something to me please. How is it that people in power( web masters) Keep saying wait for newer steppings of K10 they will be great. Yet none are saying Penryn is using A0 stepping on a totally new process with tweaks. And is doing great. What happens when Penryn B0 steppings come out and than B1 steppings. Amd gains from new steppings on an old process using soi. Yet Intel will see NO improvements with a totally new process tech . Using A0 steppings . Sure are a lot of AMD fanbois using Intel cpu's . Sheep in wolves clothing . Because in the last year the only thing that has been said about AMD Vs. Intel that I find 100% true . Is Intel is the preditor here, That is a fact. But I will go with the wolf on this one as AMD lacks a good sheppard.

X38 isn't for the faint hearted its for the pure enthusiast . Because its the only way well see A 15-20% performance increase over c2d. Using a penryn cpu.
X38 is for Penryn. X38 is were penryn will shine . X38 with penryn along with multipol drives in raid 0 is were the real performance will come from. X38 = $$$$
No need to say the benefits aren't worth the extra price as thats for people to make choice their $$$ their choice. Just because MANY won't spend for ultra performance no need to cut that extra cost because your unwilling to spend the cash for 20% more performance.

Now this only applies to Intel . If AMD K10 is 20% faster than C2D than and only than will more people be interested in gaining that insignifecant 20% performance gain . LOL!

the best advice i could give you is wait and see what the actual results are once we all get going with the new CPUs at XS and other forums

i am keeping my fingers crossed that AMD is as good as some people are claiming but i'm not really going to start believing in heresay regardless if who the source is

and THE MOST important question for me needs to be answered soon and that is whether they will have a Coldbug and how severe

 

[Nickel020]

http://www.fudzilla.com/index....=view&id=2811&Itemid=1

I think they're right, Intel doesn't want AMD to perish completely, they just want to keep them small.

 

[Nemesis 1]

This could be true.But maybe the 60 billion dollars AMD wants from intel for bully marketing has intel thinking . We have to beat these clowns down bad now. They started a lawsuite against us when they had the performance edge. Intel probabably has done nothing wrong for quite a while. So maybe intel wants to show the courts AMD just plain cann't compete' Even after they reversed engineered intels processor . If any one is done criminal actions its been AMD and IBM . IBM always lerking watching . If ever their has been a monoply its IBM . Their right there with ATT and Standard oil. They even played footsie with hitler during the war after we entered it. IBM even used its monoply powers to make Intel let AMD produce x86 processors. Intel was FORCED to do this by IBM. THan AMD reversed enginneered intels cpu and started marketing as their own .

Is anyone in CHINA reading this . Its alright you can reverse engineer Intels cores . It legal US courts have already set presedence. On top of that the full details is hidden from the public of the settlement another illeagal act( public traded company Stock holders have right to full disclosure as they own the company). Full disclosure is law. Unless you are the law.

 

[Gary Key]

Originally posted by: Nemesis 1

Originally posted by: dinos22

Originally posted by: Gary Key

Originally posted by: dinos22


Hi! Dinos. Explain something to me please. How is it that people in power( web masters) Keep saying wait for newer steppings of K10 they will be great. Yet none are saying Penryn is using A0 stepping on a totally new process with tweaks. And is doing great. What happens when Penryn B0 steppings come out and than B1 steppings. Amd gains from new steppings on an old process using soi. Yet Intel will see NO improvements with a totally new process tech . Using A0 steppings . Sure are a lot of AMD fanbois using Intel cpu's . Sheep in wolves clothing . Because in the last year the only thing that has been said about AMD Vs. Intel that I find 100% true . Is Intel is the preditor here, That is a fact. But I will go with the wolf on this one as AMD lacks a good sheppard.

X38 isn't for the faint hearted its for the pure enthusiast . Because its the only way well see A 15-20% performance increase over c2d. Using a penryn cpu.
X38 is for Penryn. X38 is were penryn will shine . X38 with penryn along with multipol drives in raid 0 is were the real performance will come from. X38 = $$$$
No need to say the benefits aren't worth the extra price as thats for people to make choice their $$$ their choice. Just because MANY won't spend for ultra performance no need to cut that extra cost because your unwilling to spend the cash for 20% more performance.

Now this only applies to Intel . If AMD K10 is 20% faster than C2D than and only than will more people be interested in gaining that insignifecant 20% performance gain . LOL!

Actually for Penryn, the A1 stepping is great. I have already seen a 3% improvement with a retail X38 over the engineering board and overclocking is a completely different, but we are waiting for the retail BIOS release later this week before presenting final X38 numbers.

 

[Nemesis 1]

To late Gary I have already seen the results . 3% on the x38 is a bit low isn't it.

I am assuming you ran DDR3/1600 with no O/C . Why not say more about overclocking and how its differant. I would have never used the words differant to discribe how A0 stepping O/C . I would have said it O/C better myself. As of right now I have seen a 17% improvement pretty much across the board over Conroe at same clocks. Thats as of yesterday. I am driving back down there next week . He is going to print everthing out for me. I will get to watch . I am excited.

 

[Midnight Rambler]

This could be true.But maybe the 60 billion dollars AMD wants from intel for bully marketing has intel thinking . We have to beat these clowns down bad now. They started a lawsuite against us when they had the performance edge. Intel probabably has done nothing wrong for quite a while. So maybe intel wants to show the courts AMD just plain cann't compete' Even after they reversed engineered intels processor . If any one is done criminal actions its been AMD and IBM . IBM always lerking watching . If ever their has been a monoply its IBM . Their right there with ATT and Standard oil. They even played footsie with hitler during the war after we entered it. IBM even used its monoply powers to make Intel let AMD produce x86 processors. Intel was FORCED to do this by IBM. THan AMD reversed enginneered intels cpu and started marketing as their own .

Regrettably, there are few of us in AT forums old enough to be able to relate to any of the truths in the above.

 

[T2k]

Originally posted by: Nemesis 1
This could be true.But maybe the 60 billion dollars AMD wants from intel for bully marketing has intel thinking . We have to beat these clowns down bad now. They started a lawsuite against us when they had the performance edge. Intel probabably has done nothing wrong for quite a while. So maybe intel wants to show the courts AMD just plain cann't compete' Even after they reversed engineered intels processor . If any one is done criminal actions its been AMD and IBM . IBM always lerking watching . If ever their has been a monoply its IBM . Their right there with ATT and Standard oil. They even played footsie with hitler during the war after we entered it. IBM even used its monoply powers to make Intel let AMD produce x86 processors. Intel was FORCED to do this by IBM. THan AMD reversed enginneered intels cpu and started marketing as their own .

Is anyone in CHINA reading this . Its alright you can reverse engineer Intels cores . It legal US courts have already set presedence. On top of that the full details is hidden from the public of the settlement another illeagal act( public traded company Stock holders have right to full disclosure as they own the company). Full disclosure is law. Unless you are the law.

I'm trying hard, VERY HARD to read some sort of meaning into this utterly incoherent, clueless ramblings of yours... what is the POINT you are trying to make here (apart from being apparently totally clueless on historical facts like how Intel took off back then etc)?

 

[Idontcare]

Originally posted by: Nemesis 1
This could be true.But maybe the 60 billion dollars AMD wants from intel for bully marketing has intel thinking . We have to beat these clowns down bad now. They started a lawsuite against us when they had the performance edge. Intel probabably has done nothing wrong for quite a while. So maybe intel wants to show the courts AMD just plain cann't compete' Even after they reversed engineered intels processor . If any one is done criminal actions its been AMD and IBM . IBM always lerking watching . If ever their has been a monoply its IBM . Their right there with ATT and Standard oil. They even played footsie with hitler during the war after we entered it. IBM even used its monoply powers to make Intel let AMD produce x86 processors. Intel was FORCED to do this by IBM. THan AMD reversed enginneered intels cpu and started marketing as their own .

Is anyone in CHINA reading this . Its alright you can reverse engineer Intels cores . It legal US courts have already set presedence. On top of that the full details is hidden from the public of the settlement another illeagal act( public traded company Stock holders have right to full disclosure as they own the company). Full disclosure is law. Unless you are the law.

Dude, where do you get your history from? I was here for much of the time periods you are writing about and you have some grave misconceptions.

Ever hear of how North Korean children view the outside world thanks to their constant immersion in North Korean government contrived propoganda?

If there was ever a parallel to be drawn to young impressionable children of Intel employees then you are presenting yourself on this forum as the poster-child.