Sandy Bridge-E benchmarks on Techpowerup.com

[Edrick]

I've said it before, the whole idea of releasing mainstream and then much, much later releasing the high end, is an awful decision.

I agree 100% here. The Nehalam release was how I would expect it to be.

 

[Edrick]

Weird logic here, blame the victim? I blame regulators who let Intel crush the x86 market and got a slap on their wrist with a 1G$ or so settlement. But hey, let's blame the company who desperately tries to compete with the behemoth that is Intel and having only one tenth of Intel's resources (probably less).

Spoken like a true socialist. Lets punish and fine a company for putting out a good product that is better than the competition. Intel was not always huge you realize. And they had a lot more competition in the 80s and 90s. They just happen to win (for various reasons). So the government should step in? I never understood this way of thinking. Punish the people who succeed to help out the people who don't.

AMD was on top for a few years during the P4 days, but they didn't improve upon that. They sat back and waited for Intel to counter. Goes to show even these "behemoth" companies also make mistakes. It is up to the smaller companies to capitalize on that.

 

[ed29a]

Spoken like a true socialist. Lets punish and fine a company for putting out a good product that is better than the competition. Intel was not always huge you realize. And they had a lot more competition in the 80s and 90s. They just happen to win (for various reasons). So the government should step in? I never understood this way of thinking. Punish the people who succeed to help out the people who don't.

AMD was on top for a few years during the P4 days, but they didn't improve upon that. They sat back and waited for Intel to counter. Goes to show even these "behemoth" companies also make mistakes. It is up to the smaller companies to capitalize on that.

So if people point out that a company has violated the law, they are now socialists? Oh boy ... good logic there! Why stop at calling them socialists, I would have gone for terrorists, has a cooler anti patriotic sound!

What does socialism has to do with the fact that a company has violated the law? This law, that is written by constituents elected by the people of a non socialist country. Monopolies are not illegal (see Google), abusing a monopoly is illegal (see Microsoft and Intel). It has nothing to do with socialism, it's the law, and in this case governments should step in and punish the companies not because of socialism or to be mean to the successful companies, but because:

They.
Broke.
The.
Law.

Now, if you don't like the law, get elected and change it. Or lobby for it to be changed.

 

[Ferzerp]

This really isn't the place for your crusade. Please take it to PMs or something.

 

[bryanW1995]

Given how nonchalantly Intel seems in dealing with LGA2011 I doubt they even care about this platform.

They don't, it's just a throw-in made up of chips that couldn't cut it as Xeons. If they'd known how poor BD was going to be they probably would have introduced these skus as lower-end Xeons instead of higher end desktops.

 

[bryanW1995]

yes any 2011 mobo should be able to take Ivy Bridge.


I think thats just the 1155 Ivy Bridge cpus coming out in 4 or 5 months though and 2011 versions will not be out until well after that. honestly I would just wait until Ivy Bridge comes out instead of going Sandy Bridge E.


nearly 4 years later and overclocked i7 920 users really do not have anything worthwhile to upgrade to.

Yeah...thank god my ip 35 pro died or I'd be sitting out another year (or 2)...

This makes it harder to resist upgrading to a 2500K, for a gaming rig, ARRRGH

Why? Your i7 @ 4.21 should be fast enough for just about anything, and could even in some highly threaded instances be as fast as a 2500k @ 4.8 or so.

The Intel Core i7-3930K is actually not bad. For $555 you get better performance than a $1000 990X. So technically on the high-end, if you don't care for 'bragging rights', then an overclocked 3930K will still be very fast for less $. For gamers, definitely not worth it. But this is primarily a powerhouse/workstation platform for someone who actually needs a fast 6-core CPU. For that type of user, the $555 asking price isn't that bad. This power user might be running raided PCIe SSDs, and 2-3 high-end GPUs, in which case the PCIe lanes on the 1155 may be insufficient.

I think the most disappointment part here is that Intel should have released LGA2011 in January of 2011 along-side with the 1155 socket. Being almost 1 year late is disappointing. 3820 delayed until February? Well, there is literally no point to that CPU given that IVB will be out very soon from that point.

They probably just don't have enough chips that are so bad that 1/2 the cores must be disabled, yet so good on the other 4 cores that they're still "good". I have a feeling the 3820 will be the skt 2011's gtx 465 or 5830.

Sigh, when will people learn to stop bitching about power consumption..ON A PERFORMANCE RIG.

Do you drive a Porsche and complain that the engine is to big?

You should have used a viper instead of a porsche as your example. A porsche engine is more like a 3820, it's weak and underpowered for the awesome platform that comes with it.

Well this is a let down..

I've been wanting to upgrade my i7-930. It actually bottlenecks my 3x580s in BF3 and I can't get it stable over 4.2GHz.

Tempted to go for a 2700K now...

sb-e is right up your alley, especially for only ~ $180 more for the 3930k. tri-sli, and you'll probably upgrade those gpus in the future, you could actually see a significant difference with the extra pci-e lanes in x79. You might even go backwards a bit in many games by dropping down to an 1155 rig.

Well after reading "3-way SLI is up to 29% faster", I don't think I can pass...

I mean, how can I spend $1700 on video cards and not use them at their maximum potential.

And you know they're being honest about this b/c NV has zero love for intel. Heck, they'd probably like to open up a can of whoop-ass (tm) on them if they could...

I see what you did there ...

Why stop at video cards? Why not take a 2700K and overclock it to 10ghz too to 'simulate' next generation CPUs too! And then, take some duct tape and tape a few more PCIx slots to the motherboard too to simulate next generation motherboards!

Please don't give jhh any ideas...

 

[bryanW1995]

Yes, because what Intel is doing with the "-E" line is stupid. You can't release the high end so late in the product cycle. The high end previously came first, and if these had come out with the normal SB, the price premium would not make so many people balk. We've always paid a premium for the top end, but when the top end will likely be matched or bettered very soon, for so little in comparison, why buy it?

I've said it before, the whole idea of releasing mainstream and then much, much later releasing the high end, is an awful decision.

SB-e should be more competitive against IB than gulftown was against SB. Even assuming 5.5 ghz top-end OC and 4 cores vs 4.6 ghz SB-e and 6 cores, SB-e should be much closer clock/clock (most recent estimates I've seen are around 7 %). The real issue is that the "high end" is now trailing the mainstream releases by a year instead of preceding them by a year.

 

[Ferzerp]

Competitive in performance, but if we can wait 6 months and get the same performance for 1/3 of the price, buying SB-E is silly.

 

[Edrick]

Competitive in performance, but if we can wait 6 months and get the same performance for 1/3 of the price, buying SB-E is silly.

I doubt 4 IB cores will match 6 SB-E cores in performance.

 

[Ferzerp]

It all goes back to depending on the workload. There comes a point that unless the computing paradigm changes, buying more cores is kind of silly except in specific apps. I'd much rather scale clock speed than cores count at the 4 core + HT level. You *should* be able to scale clock speed on the next gen appreciably higher than you can on the SB-E. It's obviously not out, so we are obviously just guessing, but theoretically that's how it should work.

 

[Edrick]

It all goes back to depending on the workload. There comes a point that unless the computing paradigm changes, buying more cores is kind of silly except in specific apps. I'd much rather scale clock speed than cores count at the 4 core + HT level. You *should* be able to scale clock speed on the next gen appreciably higher than you can on the SB-E. It's obviously not out, so we are obviously just guessing, but theoretically that's how it should work.

You are right, for most people having 4 faster cores is better than 6 slightly slower ones. And in that case IB makes sense. But those are the same people who would never bother with SB-E in the first place. The people who need SB-E, will buy SB-E regardless of when the mainstream IB comes out. Also, IB-E should not be delayed as long as SB-E was, so the X79 platform will have upgrade options in 2012 as well.

 

[Ferzerp]

The focus on throwing cores at the problem of raising raw processor performance is what has us in the situation we're in now where we can't use what we have. It shifts the problem from a few players (in x86's case Intel and AMD) over to every single application developer (who may or may not have the skillset to utilize it properly). It's kind of a cop out to keep scaling cores like that when it outpaces our ability to utilize them.

Of course there is always "fast enough" factor.

Who knows. The future of hardware may be too many cores, but that doesn't mean that it's the choice for a better experience. That's just the direction at the moment because it's easier.

 

[exar333]

I doubt 4 IB cores will match 6 SB-E cores in performance.

Agreed. If IB allows much higher OC ceiling, and if you use mostly IPC and low-threaded workloads, it could be better. Across the board, probably not.

With the X79 platform existing, it will be interesting to see when the first IB-E chip arrives. Maybe an early extreme edition only, and then more SKUs later? Who knows.

 

[bryanW1995]

You are right, for most people having 4 faster cores is better than 6 slightly slower ones. And in that case IB makes sense. But those are the same people who would never bother with SB-E in the first place. The people who need SB-E, will buy SB-E regardless of when the mainstream IB comes out. Also, IB-E should not be delayed as long as SB-E was, so the X79 platform will have upgrade options in 2012 as well.

What makes you say that? It will probably depend more upon yields of IB based Xeons than anything else. 2nd biggest influence will be competition (or lack thereof) from AMD, which seems to dictate a later, rather than sooner, IB-e launch.