Nehalem and X58 show up in Taipei

Quote:

---

Originally Posted by Anandtech: Gary Key

The board manufacturers informed us they are working on ways to increase the CPU multiplier past the single Turbo speed step and that QPI bus speeds should improve greatly once BIOS tuning is completed. While we wanted to provide clarification to the overclocking question, it is just too early to tell what the enthusiast can expect in the way of overclocking at this time. We will follow up later this week with some additional information after we have more time with the platform.

---

good news as well:up:

http://vr-zone.com/articles/Intel_X5...ured/5827.html
http://en.expreview.com/2008/06/03/c...bos-following/

abit IX58-MAX looks nice!

Quote:

---

Originally Posted by bowman

http://vr-zone.com/articles/Intel_X5...ured/5827.html
http://en.expreview.com/2008/06/03/c...bos-following/

abit IX58-MAX looks nice!

---

yeah, that does have to be one of the best looking mobo's there, i guess that kinda kills the idea that abit was going to drop out after the P45.

I still do kinda like the intel idea of having the ram at the top, although it does kinda put a stall on natural convention helping with cooling the memory.

Quote:

---

Originally Posted by Blauhung

yeah, that does have to be one of the best looking mobo's there, i guess that kinda kills the idea that abit was going to drop out after the P45.

I still do kinda like the intel idea of having the ram at the top, although it does kinda put a stall on natural convention helping with cooling the memory.

---

It reminds me of my NF4 DFI LanParty, but then again, it for some reason has three slots and one redundant slot for the triple channels.. All of these look like they have six, at least the abit does.

Quote:

---

Originally Posted by virtualrain

I'm not sure exactly how CPU-Z works, but is it possible that it could report Nehalem at this point?

---

Yes as it gets the String of the CPU and Mobo. It is like an accessor method

example:

Code:

---

protected static String getCpuMoboId(int id)
    {
        String idname =Integer.toString(id);
       
        return idname;
    }

---

Quote:

---

Originally Posted by virtualrain

There is no reference to Nehalem support in the change log for CPU-Z.

---

Not a full support.

Quote:

---

Originally Posted by virtualrain

]If it is working partially there's obviously a few bugs to work out as the L2 is reporting incorrectly for sure and the "rated FSB" is clearly irrelevant. Core voltage has me wondering as well.

---

Sure as the getInformation may still be working as the bios differs here. The FSB may be the info got from the Quick Link.

I do believe this is real. The CPU-Z author will confirm this soon anyway.

Metroid.

nehalem is code name of core ?
I thought its name of family of new line cpus
like PENRYN's (yorkfield,wolfdale + servers cores, dunno what are called)

Quote:

---

Originally Posted by Duniek

nehalem is code name of core ?
I thought its name of family of new line cpus
like PENRYN's (yorkfield,wolfdale + servers cores, dunno what are called)

---

Nehalem is the overall product code.

Lynnsfield, Havendale, Bloomsfield, Auburndale, Clarksfield etc is all Nehalem cores. Highend servers got the Nehalem-EX, Nehalem-EP etc names.

Nehalem might be named Core 3.

Intel have such weird codenames.

Quote:

---

Originally Posted by Richard Dower

Intel have such weird codenames.

---

Everone got wierd codenames :p:

Quote:

---

Originally Posted by Shintai

Nehalem is the overall product code.

Lynnsfield, Havendale, Bloomsfield, Auburndale, Clarksfield etc is all Nehalem cores. Highend servers got the Nehalem-EX, Nehalem-EP etc names.

Nehalem might be named Core 3.

---

yea
so why on screen from cpu-z (vr-zone) we can see, "code name - nehalem"

fake?

The screenshot is most certainly not fake guys. As for the CPU-Z, chances are that it is not a build that's available to the public.

Quote:

---

Originally Posted by Duniek

yea
so why on screen from cpu-z (vr-zone) we can see, "code name - nehalem"

fake?

---

Fake? Why?

They are all Nehalems. You only gonna know more when the CPU-Z people knows all the family IDs for Nehalem CPUs. Until then all are defined in a broad class.

Quote:

---

Originally Posted by Duniek

yea
so why on screen from cpu-z (vr-zone) we can see, "code name - nehalem"

fake?

---

We can not say for sure if it is fake or real till the CPU-Z author says so. We can believe on something.

Allendale --> 4XXX
Conroe --> E6XXX
Kentsfield --> X6XXX
Wolfdale --> E7XXX/E8XXX
Yorkfield --> 9XXX

I am writing a program that will show the whole Architecture history to make it understandable to newcomers. So we can easily print it out.

Metroid.

Quote:

---

Originally Posted by 4X_FTL

(this forum is still called Extreme Systems, is it not? Or did fugger change it to Extremely Affordable Systems?)

---

Might be best not to start off on the wrong foot here, or you can check out HardOCP :p:

Quote:

---

Originally Posted by Kingcarcas

Might be best not to start off on the wrong foot here, or you can check out HardOCP :p:

---

Hehehe!:D

Quote:

---

Originally Posted by Kingcarcas

Might be best not to start off on the wrong foot here, or you can check out HardOCP :p:

---

With so many elites round here its no wonder there's basically no new blood and every thread you read is basically the same five posters pissing on each other round after round.

Shrugs...I bows to all your superior xtreme'ness. Clearly my comments intended to have but one interpretation and yours was spot-on.

is this correct:

1. nehalem = 1.45 to 1.5 times the performance of penryn architecture clock for clock [single core]

2. nehalem has two threads per core for a maximum total of 8 cores + 16 simultaneous threads.

3. nehalem requires less cpu voltage to maintain similar penryn clockspeeds?

4. nehalem wont be overclockable?

1. nehalem = 1.45 to 1.5 times the performance of penryn architecture clock for clock [single core]
- I've only seen the '33% faster clock for clock' claim.

2. nehalem has two threads per core for a maximum total of 8 cores + 16 simultaneous threads.
- Yes, although the 8core is a fair bit off (late next year).

3. nehalem requires less cpu voltage to maintain similar penryn clockspeeds?
- I don't know about 'clock speeds' but their presentation claims less power consumption for the same amount of performance, which probably refers to being faster clock for clock, more threads, and such.

4. nehalem wont be overclockable?
- The rumor mill claims the non-LGA 1366 Nehalems won't be overclockable. Even if they aren't, the LGA 1366 seems like the platform to have in any case. It's not somehow going to be just $1000+ Extreme Editions like certain sites will have you believe.

Quote:

---

Originally Posted by bowman

2. nehalem has two threads per core for a maximum total of 8 cores + 16 simultaneous threads.
- Yes, although the 8core is a fair bit off (late next year).

---

and only for servers 4 sockets and up

Quote:

---

Originally Posted by 4X_FTL

With so many elites round here its no wonder there's basically no new blood and every thread you read is basically the same five posters pissing on each other round after round.

Shrugs...I bows to all your superior xtreme'ness. Clearly my comments intended to have but one interpretation and yours was spot-on.

---

:rofl::ROTF:

Quote:

---

Originally Posted by jaredpace

is this correct:

1. nehalem = 1.45 to 1.5 times the performance of penryn architecture clock for clock [single core]

---

It is not proved that Nehalem will be 45 to 50% faster clock per clock than Penryn. Intel showed charts last year telling Nehalem features as you can see below and how much the IPC would scale, on that time Intel said 33% on parallelism that means Simultaneously Multi-Thread tasks. The last source we have had is VR-Zone that demonstrated a CPU-Z Nehalem picture and a 3DVantage score, VR-Zone source author Visionary claimed that Nehalem is 45% clock per clock faster than Penryn. See below what bowman said regarding 3DMark Vantage.

Quote:

---

Originally Posted by bowman

Then again Vantage's CPU test is highly optimised and written for up to 32 threads.. So here we see every little bolt, nut and bit they've tacked onto the Penryn core, and the SMT in full effect. We won't see the 8 threads getting used too often yet (even though they look pretty in the task manager :P)

I want to see some non-SMT benchies so we can compare clock for clock, too.

---

Source

So based to what bowman said, it clearly tells us that we do not know yet how much faster is Nehalem over Penryn. I stand still and convict that my 23% arbitrary guess will be just fine.

Quote:

---

Originally Posted by Wikipedia

Parallel computing is a form of computation in which many instructions are carried out simultaneously[1], operating on the principle that large problems can often be divided into smaller ones, which are then solved concurrently ("in parallel"). There are several different forms of parallel computing: bit-level parallelism, instruction-level parallelism, data parallelism, and task parallelism.

---

Source

http://quvtqq.blu.livefilestore.com/...NIfqkjXQ/5.jpg

Quote:

---

Originally Posted by jaredpace

2. nehalem has two threads per core for a maximum total of 8 cores + 16 simultaneous threads.

---

Yes 8 cores = 16 threads, by your statement you have to remove the plus(+) operator and place an equal(=) operator. As you can see below Nehalem is all about Simultaneously Multi-Thread by the vast implemented features added to aid parallelism in Nehalen as seen on the slide below, before it used to be called Hyperthreading.

http://quvtqq.blu.livefilestore.com/...dxO8RgHQ/3.jpg

Quote:

---

Originally Posted by jaredpace

3. nehalem requires less cpu voltage to maintain similar penryn clockspeeds?

---

Yes.

Quote:

---

Originally Posted by jaredpace

4. nehalem wont be overclockable?

---

Yes, but not as much as Penryn does. This is a personal guess opinion.

Nehalem refinement known as Westmere will overtake Penryn regarding overclocking.

http://quvtqq.blu.livefilestore.com/...Jnt1PEiQ/9.jpg

Metroid.

Quote:

---

Originally Posted by Metroid

We can not say for sure if it is fake or real till the CPU-Z author says so. We can believe on something.

Allendale --> 4XXX
Conroe --> E6XXX
Kentsfield --> X6XXX
Wolfdale --> E7XXX/E8XXX
Yorkfield --> 9XXX

I am writing a program that will show the whole Architecture history to make it understandable to newcomers. So we can easily print it out.

Metroid.

---

It isn't fake.. I know where and who it came from and his word is beyond reproach.. That I'd stack money on and my name.:up:

QFT! Metroid. I'm in:up: PM's down.

Quote:

---

Originally Posted by 4X_FTL

With so many elites round here its no wonder there's basically no new blood and every thread you read is basically the same five posters pissing on each other round after round.

Shrugs...I bows to all your superior xtreme'ness. Clearly my comments intended to have but one interpretation and yours was spot-on.

---

:rofl::ROTF::gay:

Clearly when one sees others as elite and no such wording has been shown or given it is proof that they are very likely running in the questionable self-esteem que.

Be nice, you are new and might wish to enter with newness and not sarcasm. Then we can upgrade you from no self-esteem to low self-esteem.

Donnie,

The database is still showing errors can not see anything :(

I hope they will fix it soon, our social cybernetic community is down :(.

At least we can do some of the things.

Quote:

---

Originally Posted by Movieman

It isn't fake.. I know where and who it came from and his word is beyond reproach.. That I'd stack money on and my name.:up:

---

Is there a way to make a joint bet :p:

Metroid.