Prices will only go down once Intel is able to compete again.
@Asterox
That's really impressive! I can't wait to see what overclocked 3600s can do.
. We as consumers will profit as prices will innevitably go down so anyone can take a pick and chose the best chip for their needs (be it intel or AMD).
That is awesome. I'm doing a lot of x265 encoding and it's good to see Zen2 outperforming intel. I am so glad I switched to AM4 platform as soon as Zen came out.
Prices will only go down once Intel is able to compete again.
@Asterox
That's really impressive! I can't wait to see what overclocked 3600s can do.
What is the impact of Realtime versus Normal and HPET Active versus HPET Not Active?
So AMD will finally match or exceed contemporary iteration of Core processors in IPC since the Core 2 Duo days? It has been over 12 years. Holy dear god.
I think you are right. When that happens Zen 3 will be out. You guys realize they dropped hyper threading not because intel processors didn't need it. All the security issues went away when hyper threading was removed. Not exactly a good outcome for intel.I’m still thinking intel has a core2duo up its sleeve again.
I’m guessing next year around this time we’ll hear about some little known engineering team intel employs in god knows where that has a “new” processor that performs at twice the speed or power or efficiency and miraculously it will appear on the market shortly after.
No facts just a feeling.
5.275GHz on LN2 and > 1.7V
I think 5GHz on air all cores is probably... optimistic, to say the least.
I think you are right. When that happens Zen 3 will be out. You guys realize they dropped hyper threading not because intel processors didn't need it. All the security issues went away when hyper threading was removed. Not exactly a good outcome for intel.
Conroe was a further developed core of the mobile ones (Banias, Dothan) from the A64 vs P4 era so it could've been predicted to come in a desktop form at some point. At the moment we know about Icelake and this is by no means a second coming of Conroe.I’m still thinking intel has a core2duo up its sleeve again.
I’m guessing next year around this time we’ll hear about some little known engineering team intel employs in god knows where that has a “new” processor that performs at twice the speed or power or efficiency and miraculously it will appear on the market shortly after.
No facts just a feeling.
They'd really need something groundbreaking to pull something like this off again. Guess we'll find out in ~1-2 years.
1.7v? That doesn't seem right somehow. I don't think even 14LPP took that much voltage to hit 5.2 GHz under LN2.
Conroe was a further developed core of the mobile ones (Banias, Dothan) from the A64 vs P4 era so it could've been predicted to come in a desktop form at some point. At the moment we know about Icelake and this is by no means a second coming of Conroe.
The netburst architecture was a terrible one to begin with (mainly regarding IPC) which helped them to get that ~2x IPC increase, I think it was "only" about 20% higher than K8. The same goes for Zen since the architecture it replaced had subpar IPC.
They'd really need something groundbreaking to pull something like this off again. Guess we'll find out in ~1-2 years.
Intel has had their ass ignited many timed in the recent past, or have you forgotten about their mobile endeavor fiasco?My point is intel hasn’t had a fire under their ass to do anything for a long time.
They threw billions of dollars at mobile too, and those little ARM cores chewed them up like candy. Contra revenue was sweet. You can make predictions based on feelings and faith all day long, the harsh reality is this industry works on the foundations of science, engineering and most importantly building and planning years ahead. Money works wonders if you already spent them years ago, not starting today for next year. Apple dumped billions into a Google Maps competitor and it took them 7 years to even get close to their competition, let alone Conroe them.I’m sure when they start throwing billions of dollars at a problem something that was impossible to do last year suddenly becomes possible.
Those electrons just wouldn't move themselves, they had to be pushed while screaming the whole way.
Source:
https://pbs.twimg.com/media/D9-K1ZQXUAApTYn.jpg
My point is intel hasn’t had a fire under their ass to do anything for a long time. Ryzen 2000s were good but not good enough to be a great competitor.
Intel has been doing what all big companies do, pocket the money, I’m sure when they start throwing billions of dollars at a problem something that was impossible to do last year suddenly becomes possible.
Intel's problem is not architectural design. It's funny that you think 10% (is that an average figure?) upon Skylake architecture is "paltry" when AMD is now getting to Skylake ipc level after how many tries? AMD's ONLY advantage at this time is TSMC's 7nm which translates to more performance per watt so they can cram more cores into a chip at the same tdp level.Intel needs to have some of those eureka moments and you can't throw money at that. Sunnycove for example is a mediocre boost (10%) in a core design they have been working on for 12-13 years. It will be a kudos for eaking that out but its another 3-4 years before they have anything different. They aren't just going to chuck that out right away.
Those electrons just wouldn't move themselves, they had to be pushed while screaming the whole way.
Source:
https://pbs.twimg.com/media/D9-K1ZQXUAApTYn.jpg
After 2 triesIntel's problem is not architectural design. It's funny that you think 10% (is that an average figure?) upon Skylake architecture is "paltry" when AMD is now getting to Skylake ipc level after how many tries?
That is all true but don't forget that AMD of today is a moving target. They have Zen 3 ready to intersect Sunny Cove core and Zen 4 one year after that to extend the lead. They are aware of the opportunity that was presented and intel knows that as well. What needs to happen in next two years for intel to regain the competitive edge is :Intel's problem is not architectural design. It's funny that you think 10% (is that an average figure?) upon Skylake architecture is "paltry" when AMD is now getting to Skylake ipc level after how many tries? AMD's ONLY advantage at this time is TSMC's 7nm which translates to more performance per watt so they can cram more cores into a chip at the same tdp level.
I think Intel will learn a lot from the challenges of both 14nm and 10nm and could tap into lessons learned from these processes going forward. On the architectural front, I see the beginning of the cache wars cropping up on us. Intel has no reason not to make the L4 cache a mainstay on their desktop chips, especially with the introduction of progressively beefier igps.
You do realize that they're pretty much in the same place IPC wise, just having taken different paths to get there?Intel's problem is not architectural design. It's funny that you think 10% (is that an average figure?) upon Skylake architecture is "paltry" when AMD is now getting to Skylake ipc level after how many tries? AMD's ONLY advantage at this time is TSMC's 7nm which translates to more performance per watt so they can cram more cores into a chip at the same tdp level.
I think Intel will learn a lot from the challenges of both 14nm and 10nm and could tap into lessons learned from these processes going forward. On the architectural front, I see the beginning of the cache wars cropping up on us. Intel has no reason not to make the L4 cache a mainstay on their desktop chips, especially with the introduction of progressively beefier igps.
AMD is now getting to Skylake ipc level after how many tries? .