Report: Intel's Ivy Bridge-E CPUs Will Launch Sept. 2013
Intel's Ivy Bridge-E CPUs are likely to launch in September 2013, according to a slide from VR-Zone.
Previously, we were already aware that Intel's Ivy Bridge-E CPUs would be launching somewhere during Q3-2013, but now we may have a more specific date. VR-Zone has revealed a slide that shows that the CPUs are supposedly coming out in September 2013, among which are the i7-4820K, i7-4930K, and i7-4960X.
The Ivy Bridge-E CPUs would be based on a 22 nm lithography, and like any new generation, they feature a small performance improvement, as we showed you earlier.
The flagship i7-4960X would pack six processing cores and with HyperThreading feature 12 threads. It would have a base clock of 3.6 GHz, which will boost up to the 4.0 GHz mark. It has 15 MB of L3 cache and support for DDR3-1866 memory. All this processing power is made possible through a TDP of a massive 130 Watts.
| CPU Model | Cores / Threads | Base / Turbo | L3 Cache | Memory | TDP |
|---|---|---|---|---|---|
| Core i7-4960X | 6 / 12 | 3.6 / 4.0 GHz | 15 MB | DDR3-1866 | 130 W |
| Core i7-4930K | 6 / 12 | 3.4 / 3.9 GHz | 12 MB | DDR3-1866 | 130 W |
| Core i7-4820K | 4 / 8 | 3.7 / 3.9 GHz | 10 MB | DDR3-1866 | 130 W |
Based on the numbers given, we see the successor to the i7-3820K as the i7-4820K, and the i7-3930K is succeeded by the i7-4930K. Similarly, the i7-3960K is improved upon by the i7-4960K. What we don't see is a successor to the i7-3970X; however, we could assume that down the line, Intel might introduce an i7-4970X to top everything off. This is not shown on Intel's roadmap below, though.


My 3930K already runs at 5ghz on water, I would love a IB-E with 8c/16t though for my photo work
Seriously the X version should have been 8 core @ 150w & the 4820K should have been 6 core without HT. (*save the trouble for manufacturing another diff chip)
Two years ago we had Gulftown, so this is will be a 20-30% improvement at the least.
And it's not like you've seen anything comparable from AMD (in terms of year-over-year improvements, so i think you're beating a dead horse.
The real reason they're not exceeding 6 cores is probably because of power issues and the fact that an 8-core extreme part at $1k would kill AMD's high end completely, as the fastest 6C/12T SKUs would have to be sold for $800 or less.
EDIT: Got confused with the 3970X, SB-E launched Q4, 2011. Sorry!
The real reason they're not exceeding 6 cores is probably because of power issues and the fact that an 8-core extreme part at $1k would kill AMD's high end completely, as the fastest 6C/12T SKUs would have to be sold for $800 or less.
3970K is 3.5GHz, Ivy-bridge is on average ~10% faster than SB-E, so the 4930K should have no problem replacing both 3970X/3930K @ 3.2GHz 6C/12T + 15MB cache. There is already a SB-E Xeon 8C/16T@ 150w @ 3.1GHz. Intel should not have any problems with the smaller 22nm Ivy-E @ 8C/16T @ 150w = 3.3GHz which still faster than 3970X even @ 6 core environment. I generally feel the X version should have at least 2 extra cores to justify the huge price mark up from the 6 core K CPU.
The real reason they're not exceeding 6 cores is probably because of power issues and the fact that an 8-core extreme part at $1k would kill AMD's high end completely, as the fastest 6C/12T SKUs would have to be sold for $800 or less.
I disagree. First - at the high-end Intel is competing with Intel. Even a high end Opteron is preforming worse then Intel's enthusiast offerings when it comes to the most used software. While Opterons are great for bigger number of light threads (specially independent ones),Even though most Adobe and Autodesk software's computational parallelism will span across all cores, it benefits more from IPC ( and GHz ) increases because of the highly hierarchy dependent structure of computations. AMD has no laid foot here in the past years. What I am trying to say is that Intel $1k parts can't kill AMD because AMD does not have a remote, whats left of a direct competitor.
Second - we already have 8C/16T Xeons in the 150W envelope. Drop off all the transistors related to CPU to CPU communication and other specifically server based tasks and instructions, 2MB of L3 and it can maybe arrive in a 130W package.
Just my 50 cents, please feel free to correct me if my arguments were off.
Personally, I would think that it was more a matter of not having anything from AMD to compete with.
On the contrary - over saturating the market with almost the same products for years is a bad business plan as well. Intel's only saving grace is lack of competition. If i was to purchase when IV-E arrives, would definitely grab an older SB-E chip because of the reduced prices. SB-E prices will drop by more than 10%, giving it an edge in price/performance compared to IV-E. Specially when you put the same power envelope into the equation. Not to mention the risk of IV-E using thermal paste between the cores and the heat spreader is a very possible reality. If that is the case then I want to point out the OC SB vs OC IB benchmarks in which SB wins, because it needs only 200/300 MHz to match the IPC increase while the better termals allowed for 400-600 MHz bigger OC. Possibly Haswell-E ( or a suprise comeback from AMD ) will make a change.