Intel Core i7 "Ivy Bridge-E" HEDT Processors Start Selling

[radrok]

I fail to comprehend why the extreme series CPUs are based on a previous generation architecture? Intel should have skipped Ivy Bridge-E and simply released Haswell-E.

Makes more sense for Intel to try new architectures on smaller dies than going into full size dies.

Also HEDT gets what's available from the Xeon lineup.

Ivy-e is a success in its intended segment, lower consumption is a boon when building stacked 4p rigs.

 

[vega22]

i am glad i went with a 4770k instead of holding on for these.

 

[Aquinus]

Ivy-e is a success in its intended segment, lower consumption is a boon when building stacked 4p rigs.

Smaller PSUs too, so you can cram more of them in smaller chassis so it can give you more space to fit more servers. Lower power consumption also means reduced cooling costs. Considering this is really the server market, as a system admin, I'd have to say that this is a win for Intel.

Everyone seems to forget that 2011 falls under the server market more than anything else and that the i7s on 2011 are just stripped down Xeons.

 

[radrok]

i am glad i went with a 4770k instead of holding on for these.

The 4770k and these target completely different user bases, if you wait for HEDT means you need more cores or more bandwidth.

These hexas are much more better than 1150s quad when you use all of 12 threads.

You being content with the 4770k means you didn't need 6 cores to begin with

 

[Octavean]

The LGA2011 / X79 (derivative Xeon / C606) was released in late 2011. While not perfect it was a great platform then and now with many noteworthy attributes.

There are some benefits of the new Ivy Bridge-E but I think its important to note that there is a hidden benefit (or at least one that many are not acknowledging). That is to say that Intel has provided an upgrade path as an option on a well established platform before moving on to said platforms successor. A nicety that I think many like to have as an option even if they chose not to take advantage of it.

Personally I think the LGA2011 platform has held up extremely well in general as well as for its age. Understanding the platform and its use case is key though and I don't think a lot of people get it.

This is business as usual and the pricing scheme is in line with its predecessor. Haswell-E will likely be no different in its relative trifecta offerings and pricing scheme.

Haswell-E will presumably be on an LGA2011-3 socket with the same pin count as the current LGA2011 socket. The same people complaining about Ivy Bridge-E now are likely the same people that will complain about Haswell-E requiring a new socket, chipset and motherboards,.....but still retaining the same pin count.

Why complain about something like a platform or segment of the market (like HEDT) that you're never going to buy into anyway,.....?

 

[Fourstaff]

There are some benefits of the new Ivy Bridge-E but I think its important to note that there is a hidden benefit (or at least one that many are not acknowledging). That is to say that Intel has provided an upgrade path as an option on a well established platform before moving on to said platforms successor. A nicety that I think many like to have as an option even if they chose not to take advantage of it.

I think the gesture is nice but personally I don't think its worth the effort, simply because the advantage of each IVB-E SKU is a negligible upgrade compared to the corresponding SB-E SKU. The only sensible upgrade path I can currently think of is from 4 core SB-E to 6 core IVB-E, and even then choosing a 6core SB-E is not much of a downgrade. People new to the 2011 would appreciate PCIE 3.0 among other things they could have brought in. I haven't done my homework on the 4U constant full load and whether the power benefit is worth it over SB-E though, so my point might not be valid for that.

 

[Octavean]

I think the gesture is nice but personally I don't think its worth the effort, simply because the advantage of each IVB-E SKU is a negligible upgrade compared to the corresponding SB-E SKU. The only sensible upgrade path I can currently think of is from 4 core SB-E to 6 core IVB-E, and even then choosing a 6core SB-E is not much of a downgrade. People new to the 2011 would appreciate PCIE 3.0 among other things they could have brought in. I haven't done my homework on the 4U constant full load and whether the power benefit is worth it over SB-E though, so my point might not be valid for that.

Well the effort in releasing Ivy Bridge-E is really only up to Intel. Also note that after upgrading their FAB it makes sense to make use of it given the cost of the upgrade and a smaller process makes sense from that perspective for any processors they deem worthy.

The extreme series is a low yield high margin part one would think which tapers off with the middle and entry level processors. Intel already likely expect lower sales numbers accordingly just as they did for Sandy Bridge-E.

When you think back to Bloomfield and the LGA1366 platform one can point to Gulftown and say there were significant upgrade options before the platform was superseded. While true most Gulftown processors were well out of the price range of people looking to upgrade at ~$1000 or so. Even today a Core i7 975 or 970 will cost in the ~$700+ range. So I think its noteworthy that Intel is providing high-end options and processor refreshes for LGA2011 in a more palatable price range (~300 to ~1000 rather then ~$700 to ~$1000).

People arguing against such an upgrade option are also coming close to making the point that perhaps processors should be soldered to the motherboard. After all, no one wants an upgrade option anyway,....right!?!

Also as you said not everyone is upgrading so the performance delta between Sandy Bridge-E and Ivy Bridge-E isn't necessarily an issue (for some) anymore then the performance delta between Sandy Bridge and Ivy Bridge or Haswell is an issue for those on an older platform. There are also power savings to consider / calculate.

If a platform isn't for you (your use case) then I see nothing wrong with acknowledging that fact but its quite another thing to suggest it shouldn't even exist simply because its wasn't designed for your specific use case.

 

[Fourstaff]

Well the effort in releasing Ivy Bridge-E is really only up to Intel. Also note that after upgrading their FAB it makes sense to make use of it given the cost of the upgrade and a smaller process makes sense from that perspective for any processors they deem worthy.

The extreme series is a low yield high margin part one would think which tapers off with the middle and entry level processors. Intel already likely expect lower sales numbers accordingly just as they did for Sandy Bridge-E.

When you think back to Bloomfield and the LGA1366 platform one can point to Gulftown and say there were significant upgrade options before the platform was superseded. While true most Gulftown processors were well out of the price range of people looking to upgrade at ~$1000 or so. Even today a Core i7 975 or 970 will cost in the ~$700+ range. So I think its noteworthy that Intel is providing high-end options and processor refreshes for LGA2011 in a more palatable price range (~300 to ~1000 rather then ~$700 to ~$1000).

People arguing against such an upgrade option are also coming close to making the point that perhaps processors should be soldered to the motherboard. After all, no one wants an upgrade option anyway,....right!?!

Also as you said not everyone is upgrading so the performance delta between Sandy Bridge-E and Ivy Bridge-E isn't necessarily an issue (for some) anymore then the performance delta between Sandy Bridge and Ivy Bridge or Haswell is an issue for those on an older platform. There are also power savings to consider / calculate.

If a platform isn't for you (your use case) then I see nothing wrong with acknowledging that fact but its quite another thing to suggest it shouldn't even exist simply because its wasn't designed for your specific use case.

You have to realise that Gulftown brought 6 cores at a time when Bloomfield has only 4, but SB-E comes with 6 already, so I don't see a decent upgrade. I would have considered it an upgrade if the performance delta is closer to 10%, or if they offered a 8c/16t part. You can argue that Intel is moving down a node and therefore it makes sense to manufacture IVB-E, but that offers little to the end user in terms of performance. At the end of the day, to me IVB-E is only attractive to people who are migrating into the 2011 platform (i.e. building from scratch, not upgrading from another 2011 chip), or bulk upgrading from 3820 to 4960. I believe the first group will prefer a new (or at least improved) chipset (alá z77). How large the second group is I don't know, nor have any estimate of.

 

[jameskarl]

can i ask??
who are the foreign literature who studies the i7 processor??
please..??

 

[MxPhenom 216]

can i ask??
who are the foreign literature who studies the i7 processor??
please..??

What?

 

[Aquinus]

What?

Good, it's not just me and I'm not going insane.

can i ask??
who are the foreign literature who studies the i7 processor??
please..??

Could you please try to clarify your English? You're post doesn't make any sense.

 

[Prima.Vera]

can i ask??
who are the foreign literature who studies the i7 processor??
please..??

Saint Nicholas came early this year it seems

Or the celebration party at least...