• 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.

Discussion Intel current and future Lakes & Rapids thread

Page 9 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
LTC8K6 said:
"A third improved process, "14nm++", is set to begin in late 2017 and will further allow for +23-24% higher drive current for 52% less power vs the original 14nm process."

https://en.wikichip.org/wiki/14_nm_lithography_process

Whatever that means...
Click to expand...

What is drive current? 😀

jpiniero said:
At this point I am assuming you won't see any die on Intel at 10 nm that's of any kind of decent size. Maybe at the very end of the node. For servers I think they are going to go straight to Tigerlake-SP and EMIB.
Click to expand...

Cascade Lake-SP first, maybe Coffee Lake and Ice Lake as well. I think EMIB comes with Ice Lake or after Tiger Lake? I cannot remember.
 
TheF34RChannel said:
I think EMIB comes with Ice Lake or after Tiger Lake? I cannot remember.
Click to expand...

10++, so Tiger. The Lenovo leak mentions "Icelake Refresh" (ie: Tigerlake) in 2019 and 7 nm Sapphire Rapids in 2020. We'll see if Intel can meet those dates. Obviously any SR tiles that are 7 nm will be really tiny.
 
jpiniero said:
10++, so Tiger. The Lenovo leak mentions "Icelake Refresh" (ie: Tigerlake) in 2019 and 7 nm Sapphire Rapids in 2020. We'll see if Intel can meet those dates. Obviously any SR tiles that are 7 nm will be really tiny.
Click to expand...

So that's:

2017 Coffee Lake
2018 Ice Lake
2019 Tiger Lake
2020 Sapphire Rapid (7nm server part?)

My path will then be CFL-S > TGL-S > ??
 
jpiniero said:
At some point, server is going to have to "skip one" to get in front of client.. and Tiger is really just Ice on 10++. So it does make sense.
Click to expand...

Considering that Sklylake-SP differs so much from base Skylake, it won't even matter whether its Icelake or Tigerlake. Icelake-SP, or whatever the server version is called probably has a bigger core change than Tigerlake does.
 
Server first being a Icelake 'refresh' doesn't make sense to me. Seems like Intel would want to define the basic uarch for servers and then have the client team cut features, cache sizes etc.
 
formulav8 said:
When is the next real cpu update supposed to be? It looks like 14nm through 14++ is the same cpu core arch. What's going on? Even 10nm doesn't look to be much when it comes to core arch changes.
Click to expand...

I'm with him. I do not need to upgrade my 3770K, but want some really enticing reasons to do so anyways... Recent years have yielded very marginal performance and feature improvements, or in some cases even regression *shivers*. Furthermore, new process tech does not seem to be translating into the kind of performance I have come to expect as we saw from 22nm > 14nm. Basically, at this point in time, is there any indication that any of their future offerings are worth getting excited about yet?
 
formulav8 said:
When is the next real cpu update supposed to be? It looks like 14nm through 14++ is the same cpu core arch. What's going on? Even 10nm doesn't look to be much when it comes to core arch changes.
Click to expand...

With Ice Lake AFAIK... Or is the complete overhaul and revolutionary new design after TGL?
 
This
I am missing the C2D level of improvement. Does anyone have any info of what is going on inside Intel? They seem from the outside living in their own world....
 
LTC8K6 said:
I think we will be hard pressed to see a difference in gaming between 7700K, 7740X, and the CFL 6C/12T chip.

The 7700K remains the gaming king, with it's only challenger being another 4 core chip, the 7740X.

Unless 14nm++ is a bigger improvement than anyone expects...
Click to expand...

In benches maybe but not real usage. I disagree for real-world scenarios.

1. Multiplayer like BF1 is hardly ever benched and here multi-core wins (Stock broadwell-e beat 7700k easily for example)
2. Streaming
3. General background tasks. A PC for benching usually runs 0 background stuff, probably of the network, no AV and so forth. So not really a realistic scenario. Hence 2 more cores actually help in real-world when 4 are 100% used for your gaming load, and background process will lead to fps drop.
 
beginner99 said:
In benches maybe but not real usage. I disagree for real-world scenarios.

1. Multiplayer like BF1 is hardly ever benched and here multi-core wins (Stock broadwell-e beat 7700k easily for example)
2. Streaming
3. General background tasks. A PC for benching usually runs 0 background stuff, probably of the network, no AV and so forth. So not really a realistic scenario. Hence 2 more cores actually help in real-world when 4 are 100% used for your gaming load, and background process will lead to fps drop.
Click to expand...
I wonder if a 6C BW-E beats a 7700K that easily in BF1? I'm sure it beats it, because BF1 does like more cores.

But the CFL chip won't have the extra cache that a 6C BW-E has, so it may be somewhat slower in games, which seem to like cache.
It won't have the memory bandwidth that a BW-E chip has, either. It will have about half.

BF1 seems to be the game that is always trotted out for the number of cores argument. 🙂
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top