core i7-2700k is coming

[Idontcare]

Maybe Intel has some insider info on Bulldozer performance and need to nudge performance of 2600K price bracket a bit. But thats only if BD is competitive.

/tinfoilhat

When intel came out with the Q6600 at around $800, I was convinced this was because they had inside info on the performance range of the then not-yet-released Phenom processor.

Then Intel dropped the Q6600 price to $500. Then to $300, all while there was no Phenom to be found.

Then Phenom was launched and it couldn't hold a candle to the Q6600, stock or overclocked.

The Q6600 totally pinned down the upper-end of Phenom's pricing structure.

Of course Intel went ahead and released a Q6700 in the meantime as well, but nobody around here really bough a Q6700 knowing that a Q6600 was going to OC to the same level anyways.

Then Intel did the same thing with the i7-920, ensuring the lower-end product was around $300 but still exceeded the upper-end of performance that a Phenom II was going to deliver.

When I look at Sandy Bridge vs. Bulldozer, I see the mere existence of 2500K for $200 and 2600K at $300 to be an indication of where Intel figured the upper-end of Zambezi would come out.

I don't see the 2700K being any different than the Q6700. Its existence has no bearing on the enthusiasts that buy 2600K's nor the impending zambezi product. But they have to offer their OEM's something new for them to refresh their lines with otherwise the marketing engines go stale. You can't have the same 2600K top-end desktop ads from Jan being used nearly a year later, it just looks bad.

 

[RussianSensation]

I don't see the 2700K being any different than the Q6700. Its existence has no bearing on the enthusiasts that buy 2600K's nor the impending zambezi product. But they have to offer their OEM's something new for them to refresh their lines with otherwise the marketing engines go stale. You can't have the same 2600K top-end desktop ads from Jan being used nearly a year later, it just looks bad.

Exactly. Eventually i7 920 was replaced by 930, i7 860 by 870 and i5 750 by 760 at the same price. I pretty sure 930/870/760 didn't overclock any better than their older cousins. Also, it's not as if AMD stands still. They refreshed 955 to 965 to 980. Intel slowly increasing clock speeds ensures that they can safely retain their pricing strategy by being out of reach in performance vs. AMD parts. Finally, I am not sure if Intel thought BD will be delayed for so many quarters. So the 2700k part was probably planned 6-9 months ago anyway. Once the yields are identical on i7 3.5g vs. an i7 3.4g, it only makes sense that Intel updates its lineup and cements its leadership (and with it 60% gross margins ).

After what happened with A64, X2, Opteron and FX series, Intel is well aware that you can't ever underestimate your competition. The competition is not standing still — you've got to have a plan to continue to grow, anticipate imminent response from your competition and have succession products in the pipeline as businesses and consumers will continue to upgrade to faster technology.

 

[skipsneeky2]

This 2700k is what the i7 930 was to the i7 920.

Expect a i7 2800k soon.

Not worth the purchase unless you got a dual core or building a new system.

Maybe worth it to those like me running 2500k chips and want a fast binned new stepping if i7 2800k is ever to be released...

 

[StrangerGuy]

OMG100MHZ!

Yawn.

 

[BlueBlazer]

OMG100MHZ!

Yawn.

Looks like its more or less confirmed from Bits and pieces: AMD FX-8150 at 8.43GHz, Core i7-2700K frequency confirmed

 

[Riek]

Looks like its more or less confirmed from Bits and pieces: AMD FX-8150 at 8.43GHz, Core i7-2700K frequency confirmed

So the i7 2700 wil be <3% faster than the 2600. Thats basically releasing the same cpu with a different number (since 3% is in the test error margin)

 

[BlueBlazer]

So the i7 2700 wil be <3&#37; faster than the 2600. Thats basically releasing the same cpu with a different number (since 3% is in the test error margin)

It would be more logical to release something like 3.8GHz/4.2GHz turbo SKU, which would be more interesting (even though the numbers are quite conservative IMHO as Sandy Bridge CPUs has so much headroom).

However if with the release of this Core i7 2700K will reduce prices of current Core i7 2600K remains to be seen. Also will there be an even higher ceiling of stable overclock on air cooling (exceed 5GHz?) on these Core i7 2700K? :hmm:

 

[blackened23]

It would be more logical to release something like 3.8GHz/4.2GHz turbo SKU, which would be more interesting (even though the numbers are quite conservative IMHO as Sandy Bridge CPUs has so much headroom).

However if with the release of this Core i7 2700K will reduce prices of current Core i7 2600K remains to be seen. Also will there be an even higher ceiling of stable overclock on air cooling (exceed 5GHz?) on these Core i7 2700K? :hmm:

They could probably do a 3.8ghz part if they really wanted to, but as it would be binned the same as current 2500/2600k parts it would not be profitable - yields would not be very good.

It is extremely doubtful that 2600k will lower in price, although I would be extremely happy if it did. All indications point to 2700k being a silent successor to the 2600k with the price being the same, while the 2600k disappears.

 

[BlueBlazer]

They could probably do a 3.8ghz part if they really wanted to, but as it would be binned the same as current 2500/2600k parts it would not be profitable - yields would not be very good.

I was being conservative because 3.8GHz is base clock (peanuts for Sandy Bridge) and 4.2GHz is turbo on single core (again no biggie for Sandy Bridge), not 4.2GHz on all cores. However that 100MHz increment does look kinda disappointing considering the CPU is capable of much more. :hmm:

It is extremely doubtful that 2600k will lower in price, although I would be extremely happy if it did. All indications point to 2700k being a silent successor to the 2600k with the price being the same, while the 2600k disappears.

If the Core i7 2700K pricing is the same, then Intel needs to sell off current inventories of those Core i7 2600K CPUs somehow. :hmm:

 

[inf64]

If 2700K was a 3.6Ghz base part with 4Ghz Turbo it would beat 990x in ~90&#37; of desktop benchmarks most websites use in their reviews. It would come on the top when it comes to average score and immediately make whole wetsmere and maybe even SB-E lineup - irrelevant. All for a fraction of price. Count in better overclocking and you see intel's problem: chip would be "too good".

 

[blackened23]

If 2700K was a 3.6Ghz base part with 4Ghz Turbo it would beat 990x in ~90% of desktop benchmarks most websites use in their reviews. It would come on the top when it comes to average score and immediately make whole wetsmere and maybe even SB-E lineup - irrelevant. All for a fraction of price. Count in better overclocking and you see intel's problem: chip would be "too good".

This has been discussed before, but simply raising clockspeed isn't a trivial matter. Intel can't simply say "hey guys, lets up the clock!" I hypothesize that maintaining good profitability while having good yields, I suspect, is the driving factor behind this - overclocking headroom has little to do with what kind of yields intel gets. This is the same reason lynnfield wasn't released with faster versions, it wasn't profitable due to yields.

 

[BlueBlazer]

If 2700K was a 3.6Ghz base part with 4Ghz Turbo it would beat 990x in ~90% of desktop benchmarks most websites use in their reviews. It would come on the top when it comes to average score and immediately make whole wetsmere and maybe even SB-E lineup - irrelevant. All for a fraction of price. Count in better overclocking and you see intel's problem: chip would be "too good".

There is however an existing Sandy Bridge based Xeon CPU that has those specifications >> Intel® Xeon® Processor E3-1290 (8M Cache, 3.60 GHz), though the price is kinda ridiculous.