• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Broadwell-E 10 cores?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
My thread lives!

March 2016 release, aka Q1 2016.


I've been waiting forever to make my new build and I've pushed it from Ivy-E, to HW-E and I was set to go with BD-E and now am rewarded with a pleasant surprise. Previous documentation had the same old 6/6/8 setup. I'm confident the delay allowed this.

Now it's making me wonder if I should wait till sky-e, lol..

Yeah having four SKUs with the top being deca core is pretty exciting.

In fact, I really hope this opens the door for the ~$300 price bracket to return (with a 6C/12T, 28 PCIe chip filling that.)
 
Speaking of which, what are the odds that Intel is genuinely concerned about Zen being an 8-core inexpensive monster, and that they might actually lower their prices on BDW-E to compete? (In an attempt to drive AMD out of business.)

Then again, Intel isn't generally the sort of company to engage in price wars. Maybe core wars? Intel 10-core for same price as AMD 8-core Zen? That would indeed be interesting. I'm not sure who would buy Zen at that point, except for AMD die-hards.
 
They might in fact be concerned about Zen, but if intel really wanted to drive AMD out of business they would have done it already the last few years. They could easily have made a mainstream hex core, made hyperthreading more widespread, made iris pro much more available, and cut prices if they wanted to drive AMD out of business. I dont buy all the Zen hype, but at least AMD should be more competitive if they can hold out for Zen and HBM APUs.
 
Let's see, desktop broadwell reduced power consumption considerably, while being slightly faster and having edram on top vs haswell. Can we expect broadwell-e to be as power-efficient, a sub 100W TDP for the junior SKU perhaps?
 
Speaking of which, what are the odds that Intel is genuinely concerned about Zen being an 8-core inexpensive monster, and that they might actually lower their prices on BDW-E to compete? (In an attempt to drive AMD out of business.)

None. Note that I'm sure that Intel would love to get Purley out ASAP so it's possible that Skylake-E could be released before Zen is. I really think Intel is doing the 10 core to find out how much $$$ they can wring out of people.
 
Intel 10-core for same price as AMD 8-core Zen? That would indeed be interesting. I'm not sure who would buy Zen at that point, except for AMD die-hards.

Probably the eight core Zen will be much cheaper than the ten core Intel.

With that mentioned, what interests me most is 6C/12T Intel vs. 6C/12T Zen.

P.S. Will be interesting to see how AMD sized the Zen core. If Zen is big (high IPC/high frequency) that will help AMD for enthusiast desktop, but then I would think performance per watt on the 32C Zen Servers (and possibly mobile) would be less than ideal. If AMD sizes Zen smaller then peformance on the 32C Server should be better, but it will have a harder time competing on enthusiast desktop vs. New Intel processors and maybe even old ones (eg, E5 2670 octocore is now $179.00 shipped on the surplus market).
 
Probably the eight core Zen will be much cheaper than the ten core Intel.

With that mentioned, what interests me most is 6C/12T Intel vs. 6C/12T Zen.

P.S. Will be interesting to see how AMD sized the Zen core. If Zen is big (high IPC/high frequency) that will help AMD for enthusiast desktop, but then I would think performance per watt on the 32C Zen Servers (and possibly mobile) would be less than ideal. If AMD sizes Zen smaller then peformance on the 32C Server should be better, but it will have a harder time competing on enthusiast desktop vs. New Intel processors and maybe even old ones (eg, E5 2670 octocore is now $179.00 shipped on the surplus market).
why not have 2 families? why use a 1 size fits all solution? intel got their xeons, why can't amd have different cpus for different markets?
 
why not have 2 families? why use a 1 size fits all solution? intel got their xeons, why can't amd have different cpus for different markets?

AMD will have two families using the Zen core.

FX (or whatever they end up calling it) for desktop (using Zen 8C/16T).

Opteron for Server (using Zen 32C/64T as the rumored top die).

P.S. Are you thinking there may be a beefed up cat core eventually coming out that would be strong enough in single thread for future high core count Opteron Servers? If so, that would still be independent of the design decisions needed for Zen in order to fit both 8C/16T Enthusiast desktop as well as 32C/64T Opteron Server.
 
Last edited:
Strategy wise do you think Intel might "like" AMD not to be obliterated? If AMD were totally knocked out of the desktop processor business might the monopolies regulators come in if the only show it town is Intel? Maybe Intel want AMD to still be in the game, just way behind rather than being totally out.
 
Nah, Micro Center usually has then for sale on the date of release at good prices.
True. What's nice is the ability to upgrade the BIOS and still use the X99 mb. My 5960x at 4.4Ghz is rock solid. Same mb as you. GREAT product.
 
600$ for a cpu that lasts 5+ years isn't a bad deal at all.

Add to that you can probably sell it for $150-200 in 5 years. TCO = ($600 - $150) / 5 = $90 per year. Alternatively it can be used in a secondary rig for another 5 years. I am being very conservative about the resale value because from my experience the resale value for modern Intel chips is often higher than that. Looking at 2600K/2700K/3770K, they aren't that cheap. Once we passed Lynnfield i7 860/870 chips the more modern Intel architectures hold their resale values like crazy.
 
Last edited:
yeah with as slow as cpus advance now its idiotic to skimp. right now getting a 6600k is a very stupid move compared to a 6700k or 5820k or better if you are going to be running a top end gpu now and upgrading to even faster gpus down the road.
 
when are these guys exactly coming out?

im itching for a upgrade, and a 6950X is calling for me.

Q1 technically means up til march.... does that mean march?

Why couldn't it be xmas.... why is intel not getting with the season of buying and releasing these guys during shopping boom.
 
Last edited:
when are these guys exactly coming out?

im itching for a upgrade, and a 6950X is calling for me.

Q1 technically means up til march.... does that mean march?

Why couldn't it be xmas.... why is intel not getting with the season of buying and releasing these guys during shopping boom.

I'm hearing March and I REALLY hope they make it. If the pricing rumors are true, I'll probably go with the 8-core model or maybe even spring for the 6950X if I think I can hit 4 Ghz with it.
 
when are these guys exactly coming out?

im itching for a upgrade, and a 6950X is calling for me.

Q1 technically means up til march.... does that mean march?

Why couldn't it be xmas.... why is intel not getting with the season of buying and releasing these guys during shopping boom.

because the 0.01% who buys a 6950X will buy it whenever it is released 😛
 
when are these guys exactly coming out?
Late Q2 of 2016, at least according to this leak:

38462


About the same time for Pascal and Arctic Islands, I think, which is good for whole new system builders like myself.
 
Last edited:
Someone's gonna have to lock up my credit cards because I might not be able to make it to Skylake-E even though I've had no problems with an OC'd 4670K...
 
ugh...late q2? That chart makes me sad 🙁

Be thankful we are getting 6/8/10 core split now. Let's face it, Intel didn't have to do it this generation. They could have easily bumped clocks 100-200mhz and sold us 6/6/8 for the exact same prices and waited until 2017 to add more cores. AMD will have nothing to compete with X99 products in almost all of 2016. Props to Intel for doing this. :thumbsup:

Still sucks though that the road-map basically confirms no SKL-E until 2017, which means there will be about a 1.5 year gap between the mainstream platform's new architecture and the subsequent HDET platform on the same architecture. If Kaby Lake is Q4 2016, chances are Cannon Lake is Q4 2017, Ice Lake is Q4 2018. If we assume similar spread of 1.5 years between mainstream and HDET, that would imply we might not even see Ice Lake-E is 1H of 2020. Damn. D:
 
Last edited:
It's 6/6/8/10 cores for Broadwell-E, which leaves open the possibility that the 6/6/8 will retain similar configurations and pricing to Haswell-E, and the 10 core is $2,000.

Some people will still buy the 6950X just because it raised the bar for performance, but it's not exactly the offering that even HEDT buyers are looking for.
 
Be thankful we are getting 6/8/10 core split now. Let's face it, Intel didn't have to do it this generation. They could have easily bumped clocks 100-200mhz and sold us 6/6/8 for the exact same prices and waited until 2017 to add more cores. AMD will have nothing to compete with X99 products in almost all of 2016. Props to Intel for doing this. :thumbsup:

Still sucks though that the road-map basically confirms no SKL-E until 2017, which means there will be about a 1.5 year gap between the mainstream platform's new architecture and the subsequent HDET platform on the same architecture. If Kaby Lake is Q4 2016, chances are Cannon Lake is Q4 2017, Ice Lake is Q4 2018. If we assume similar spread of 1.5 years between mainstream and HDET, that would imply we might not even see Ice Lake-E is 1H of 2020. Damn. D:
This sucks. I would feel bad buying a new "enthusiast" CPU with a two-year old architecture no matter how many cores it has, but everyone will make you feel bad for buying the mainstream i7 actually of the latest architecture because it's a quad-core. Can't win no matter what.

Mainstream hexa-core; why can't Intel give us that...
 
Back
Top