First Sandy Bridge Numbers?

[exar333]

Why would a gamer buy sandybridge? That's not what the SKU is designed for.

This is like someone benching Llano or Bobcat but deciding to use SLI-GTX480 instead of benching the Fusion aspect of the CPU because hey there are some gamers out there that might want to see such benches...but that's not an excuse to not bench the part of the chip that defines its very existence.

Even if the performance sucks, from bad driver or bad bios support, its not coolaler job (or is it?) to make sure Intel's prototypes are seen in the best possible light so you'd think a leaker would leak the results as they are. It kinda defeats the purpose of leaking stuff if you only leak the stuff that is pretty much irrelevant to begin with.

As a gamer, I would love Sandy Bridge if it had some sort of hybrid-power scheme for the IGP/discrete GPU. I could save power and use IGP when surfing the web, and then use my GPU when gaming.

I agree though, in general, that part of the performance should be the IGP numbers. It may not be "extreme" but it is an important aspect of the CPU.

 

[Nemesis 1]

:\

You need to spend more time on this forum.
Then you will realize who i exactly am.



Correct.. no one is telling me about 2011.
From what i heard, its a bigger mess over there.

Who do ya think you are ? Someone special? NOT!

 

[taltamir]

Who do ya think you are ? Someone special? NOT!

aigomorla knows what he is talking about

 

[Acanthus]

Until i see application numbers (encoding, games, power at idle, heat specs, overclocking) i don't care about sandy bridge at all.

These benchmarks are pretty worthless indicators of what we can expect.

 

[aigomorla]

aigomorla knows what he is talking about

LoLz.. he means it as a joke.
Well at the most im guessing...
Him and i go way back...

Who do ya think you are ? Someone special? NOT!

()

 

[Ben90]

aigomorla knows what he is talking about

He has the numbers, but I think his perception of those numbers is completely wrong. He is trying to argue that a 700 dollar processor will be a better buy than anything else out currently or SB. He doesn't know how I use my processor and he doesn't know how a i7-970 is actually slower than my 920 due to to the slow cache.

That's right Aigo, for my needs, a 920 RAAPEPPSESES and SLUGHtsers a 980x. Anyone that has a usage pattern or requirements outside of your views is apparently wrong. Just like achieving subambient using a pelt is a bad cooling solution because it doesn't conform to your over engineered WC metric.

 

[aigomorla]

That's right Aigo, for my needs, a 920 RAAPEPPSESES and SLUGHtsers a 980x. Anyone that has a usage pattern or requirements outside of your views is apparently wrong. Just like achieving subambient using a pelt is a bad cooling solution because it doesn't conform to your over engineered WC metric.

Sigh...
Im not even going to argue.
And a 920 is better then SB currently because of the element of Overclocking.

Obviously you havent been reading my comments

excuse me are we looking at overclocking or not?

Because if we ignore overclocking yes, your point is valid.

But i said in regards to overclocking.

Meh whatever.. this is going right back to the whole argument on 32nm.
I'll just sit and wait it out, since my side of the story obviously means nothing to you guys.

Also Ben a westmere/gulftown would rape your 920 in overclocking.
That is a fact.

 

[Ben90]

Oh I'm not doubting westmere/gulftown would clock way faster than my 920, It just would have a problem catching up to my speedy L3$. I would totally wager a pink bet between your processor and mine, but that would require me moving off the stock cooler which is barely keeping a 4.2ghz Bloomfield from catching fire.

 

[Idontcare]

Until i see application numbers (encoding, games, power at idle, heat specs, overclocking) i don't care about sandy bridge at all.

These benchmarks are pretty worthless indicators of what we can expect.

The threshold of relevance is obviously a personal one and I don't begrudge others who have a threshold set such that pre-retail stepping ES results are meritorious.

For me though I truly don't care, and this doesn't happen often, about SB performance until I see retail-stepping bechmarks AND I see it available for ship at Newegg so I can evaluate price/performance.

Gulftown pre-retail meant what? Awesome benches but ridiculous price/performance...so how many people actually care about gultown's performance?

So show me SB prices (including platform...new mobo, etc), show me retail performance, and the hopefully show me Bulldozer results of same nature so I can decide what my 2011 computer is really going to look like.

Untill then I guess we are all pretty much just neffing here in the CPU forumz...

 

[bryanW1995]

First of all WPrime is fully multithreaded, so yes the i7-970 will almost certainly beat a quadcore SB in that benchmark. It is not however an accurate reflection of most applications.

Secondly I have to point out that its a bit premature to talk about how well SB will overclock, as there is no way either of us can accurately predict that. The only thing we can know at this point is that it will be faster than a quadcore nehalem, which is already fairly close to a hexacore nehalem for most purposes (http://www.anandtech.com/bench/Product/142?vs=99)

I'm pretty sure that aigo knows a little bit about SB overclocking, or lack thereof.

 

[Nemesis 1]

aigomorla knows what he is talking about

Yes he does . As do thousands more on the internet. But that wording just stunk.

 

[Nemesis 1]

The threshold of relevance is obviously a personal one and I don't begrudge others who have a threshold set such that pre-retail stepping ES results are meritorious.

For me though I truly don't care, and this doesn't happen often, about SB performance until I see retail-stepping bechmarks AND I see it available for ship at Newegg so I can evaluate price/performance.

Gulftown pre-retail meant what? Awesome benches but ridiculous price/performance...so how many people actually care about gultown's performance?

So show me SB prices (including platform...new mobo, etc), show me retail performance, and the hopefully show me Bulldozer results of same nature so I can decide what my 2011 computer is really going to look like.

Untill then I guess we are all pretty much just neffing here in the CPU forumz...

As usual some sanity you add to thread . Amd reported its earnings today . Lost money without handouts. As for AMDs 32nm and BD not good news. AMD got hammered on servers lost market share. Seems to done well on laptops . But till the market numbers come out for QT . its all talk.

 

[ydnas7]

why do i get the impression that per clock sandybridge's SMT will be materially improved vs nehalem but the single thread IPCC increase will stay marginal. and that this SMT improvement is at the expense of overclockability, but sandybridge is likely to have even better turbo mode.
In short, the 2C4T sandybridge will rock for mass market users who don't overclock, intel i3 sandybridge will be a worthy competitor to not just Athlon x4 but Phenom x4 also.

To achieve better SMT intel might be doublepumping or duplicating some execution units so that both parts of both threads are processed in concurrently. That would reduce turbo ability but the integration of clock-gen will probably more than make up for that.
If its selective doublepumping than sandybridge has a touch of bulldozer without the die size penalty.
both doublepumping and internal clockgen would make 3rd party overclocking less successful. but perhaps a 4C4T sandybridge might be an overclock star.

 

[ydnas7]

perhaps i'm totally wrong about improving IPCC with Sandybridge, maybe its just 5% - 10% improvement, with all the effort on the core spent at reducing energy usage. if so the result is great for mobile users (long battery life) great for servers (all cores run fast, and low energy costs) but not much for PC users. (ie intel keeps the all the profit in the x86 industry, AMD keeps its market-share in PC chips)

Sandybridge 4Core has 1.12 billion transistors (includes 8mb? cache)
Westmere 6 Core has 1.17 billion transistors (includes 12mb? cache)
as cache tends to have high transistor count, Sandybridge 4core might well be a larger chip than Westmere 6 core.

 

[IntelUser2000]

Sandybridge 4Core has 1.12 billion transistors (includes 8mb? cache)Westmere 6 Core has 1.17 billion transistors (includes 12mb? cache)
as cache tends to have high transistor count, Sandybridge 4core might well be a larger chip than Westmere 6 core.

Sandy Bridge 4 core is at ~225mm2 die because the GPU portion is more area efficient than CPU core logic.

 

[Idontcare]

Sandy Bridge 4 core is at ~225mm2 die because the GPU portion is more area efficient than CPU core logic.

xtor density will be high for GPU logic compared to the CPU core logic since the design clockspeeds are much much lower. (don't need the Idrive, don't need the Vcc)

I'd expect the same of AMD's fusion offerings...unless they targeted the GPU clocks to be comparable to the CPU core clocks.

 

[IntelUser2000]

xtor density will be high for GPU logic compared to the CPU core logic since the design clockspeeds are much much lower. (don't need the Idrive, don't need the Vcc)

I'd expect the same of AMD's fusion offerings...unless they targeted the GPU clocks to be comparable to the CPU core clocks.

Right, just saying what's out there though. The die sizes were estimated quite a long time ago.

 

[lopri]

Sorry I started this thread and didn't follow it through so I don't know where the discussion has been going, but I think SB performance will be just fine. (i.e. generational performance leap when all things considered) I remember Intel demo'ing Cinebench run on a SB laptop and it took about 40~45 secs. Considering it's a mobile chip and Cinebench takes all cores, I'm guessing the clock speed was probably below 3.0 GHz. If so that's a pretty good improvement over Nehalem.

 

[Idontcare]

Right, just saying what's out there though. The die sizes were estimated quite a long time ago.

was just adding the reasoning why what you said was true for the folks who might be curious about the why (just explaining the "cause" of your noted "effect"), wasn't meaning to imply you weren't aware of it yourself

 

[Tofurkeymeister]

It seems as if the stock version of sandy bridge will not support overclocking. Be prepared to pay for the unlocked version:

forums.anandtech.com/showthread.php?t=2091809

 

[IntelUser2000]

was just adding the reasoning why what you said was true for the folks who might be curious about the why (just explaining the "cause" of your noted "effect"), wasn't meaning to imply you weren't aware of it yourself

Yea, I'm aware of that.

 

[RussianSensation]

The "high-end" Sandy Bridge doesn't do anything special. Expecting more than 5% increase over the "crappy 115x" Sandy Bridge is ridiculous. If Intel is gonna deliver for Sandy Bridge, the very first versions will tell us how its doing.

Sure it does. LGA2011 will support 6 and 8 core SB processors. You have a dual core processor. This means for you it will be an upgrade to get a 4-core SB. For Core i7 users near 4.0ghz, and with rumors that SB won't overclock at all on LGA1155 unless you get K designated processors with unlocked multipliers, the 1155 platform is shaping up to be a disappointment.

 

[IntelUser2000]

Sure it does. LGA2011 will support 6 and 8 core SB processors. You have a dual core processor. This means for you it will be an upgrade to get a 4-core SB. For Core i7 users near 4.0ghz, and with rumors that SB won't overclock at all on LGA1155 unless you get K designated processors with unlocked multipliers, the 1155 platform is shaping up to be a disappointment.

Go back to the thread, and read the graphs carefully.

The "partially unlocked" version will let you do some bins above the Turbo bin. That is called overclocking, yes?

6 and 8 cores still do nothing for PC segment because extremely small amount of applications take advantage of it. The larger caches, and potentially more memory bandwidth will benefit far more people. And you'll pay through the nose for the 6 and 8 cores.

There was also a rumor that 6 cores will be coming later for the LGA115x platform.

And I still insist that LGA135x with the 3-channel memory will be the high-end, available with same 6 and 8 cores people are expecting as well. MAYBE they'll do socket 2011 for the V8 replacements costing $5k for the bare minimum config. Good luck with that.

 

[aigomorla]

MAYBE they'll do socket 2011 for the V8 replacements costing $5k for the bare minimum config. Good luck with that.

*kicks a can... and then slowly goes off into a corner*

 

[IntelUser2000]

*kicks a can... and then slowly goes off into a corner*

*Intel hands Aigo FREE LGA2011 CPU, he runs away cheering*

Go away.