I hope not. There is no need for a core war right now. We need better IPC/clocks on 8 cores. Anyone needing 16+ can use HEDT. Amdahl's Law is a thing, and we are seeing diminishing returns in most software already.
Rumour: Ryzen 3000/x570 to have 16c/32t
He talked about AMD not wanting to lose, for instance, CS:GO users. It's not core count they're concerned about, iit's clock speeds.I hope not. There is no need for a core war right now. We need better IPC/clocks on 8 cores. Anyone needing 16+ can use HEDT. Amdahl's Law is a thing, and we are seeing diminishing returns in most software already.
These would be nice to see. I would also love some actual clarification on the 210w PL2. Because the Intel basically states in its sheet that it's the default target. But if Intel "suggests" or requires that the default be 210w it isn't in violation of their datasheet. What is the PL2 set to when some plugs in a 8700k? If it's ~113PL2, wouldn't this imply that Intel set the default for the 9900K and not every single mobo manufacturer ignoring the data sheet to illegally overclock your CPU for you?
https://videocardz.com/newz/amd-x570-chipset-with-pcie-4-0-could-launch-at-computex
Appears that Comet Lake won't be getting an updated chipset (although that doesn't necessarily mean it won't be on a different socket) and that Intel is going to release IGP-less models of the 9th series soon.
I hope not. There is no need for a core war right now. We need better IPC/clocks on 8 cores. Anyone needing 16+ can use HEDT. Amdahl's Law is a thing, and we are seeing diminishing returns in most software already.
I don't see them competing with their HEDT chips. They've also promised support for AM4 til 2020.
I hope not. There is no need for a core war right now. We need better IPC/clocks on 8 cores. Anyone needing 16+ can use HEDT. Amdahl's Law is a thing, and we are seeing diminishing returns in most software already.
He talked about AMD not wanting to lose, for instance, CS:GO users. It's not core count they're concerned about, iit's clock speeds.
Nah, more Cores, more better. AMD needs to continue pushing Intel hard, both for their own benefit and the Consumers benefit.
If there's going to be PCIe4 support, I guess that many buyers won't be overly bothered about backwards compatibility with previous boards; they'll be wanting PCIe4 boards.
Of course, many does not equal all, but AMD may consider that they'll bring over many new to AMD customers to their platform. Admittedly, some of the value goes out of the platform if a new board is required, especially for those that bought into AM4 on the back of the compatibility guarantee that AMD outlined.
We all might be quite dumbfounded later on for what AMD has done this time.
He talked about AMD not wanting to lose, for instance, CS:GO users. It's not core count they're concerned about, iit's clock speeds.
If there's going to be PCIe4 support, I guess that many buyers won't be overly bothered about backwards compatibility with previous boards; they'll be wanting PCIe4 boards.
Of course, many does not equal all, but AMD may consider that they'll bring over many new to AMD customers to their platform. Admittedly, some of the value goes out of the platform if a new board is required, especially for those that bought into AM4 on the back of the compatibility guarantee that AMD outlined.
It probably means Comet Lake-S won't come out in 2019.
The Comet Lake-S will use a new LGA 1200 socket. I'll be surprised if Intel doesn't rebrand the chipset to 4xx, even though it is the same KBL/CNL chipset.
The original leak said the 10C Comet Lake-S is planned to be released in Q1 2021. But since the information was shared at an Intel IoTG partner event, the plan for consumer desktop segment may be somewhat different, especially the K processors, which are likely to be available several months earlier.
I'd have to agree. While we were certainly short on cores pre-Ryzen, stuck at 4/8 for consumers it's time to let software catch up a bit. For a basic mom box, a 4/4 2200G is plenty. For a new gamer build I'd set the baseline at 6/12. Move up to 8/8 or 8/16 for a 'premium' gaming rig.I hope not. There is no need for a core war right now. We need better IPC/clocks on 8 cores. Anyone needing 16+ can use HEDT. Amdahl's Law is a thing, and we are seeing diminishing returns in most software already.
There was nothing in the original leak about a 2021 date or any other release date indication, it means your entire posting seems fake.
exactly...That's the downside to guaranteeing support for 4 years, new standards come out and you end up with situations like this.
I'm also sceptical about how budget B350 and B450 mobos with lower end VRMs will support a hypothetical 16C Zen 3000 chip. Many of those boards actually already have suboptimal VRMs for 8C Zen+ like the 2700X, so I understand the mobo makers concerns about doubling the core count, even on a more efficient 7nm process.
We don't know if this is true or not but the video suggested that AMD and (some) MB vendors have different opinions whatever current AM4 MBs can handle 16 cores (possibly overclocked) or not.
I'm also sceptical about how budget B350 and B450 mobos with lower end VRMs will support a hypothetical 16C Zen 3000 chip. Many of those boards actually already have suboptimal VRMs for 8C Zen+ like the 2700X, so I understand the mobo makers concerns about doubling the core count, even on a more efficient 7nm process.
Intel doesn't disclose transisitor counts. My guess is that it would be at least 140 mm2 if it was on 14 nm, and probably a bit more.
I'm also sceptical about how budget B350 and B450 mobos with lower end VRMs will support a hypothetical 16C Zen 3000 chip. Many of those boards actually already have suboptimal VRMs for 8C Zen+ like the 2700X, so I understand the mobo makers concerns about doubling the core count, even on a more efficient 7nm process.
How is choice ever a bad thing? You actually prefer being forced to buy a new board with each gen?exactly...
I am waiting for the ryzen 3000 series review with b350 middle budget board (not total low end)
My prediction is it won't work to the max CPU capabilities (XFR..) and if it does the memory for r3000 will have a major impact on performance and it won't work at high speeds
And to be rude, I am expecting the review sites won't do any reviews with that b350 middle level boards, but with the top end x570 boards with the reason to get max performance but also not to show the performance deficit with the budget b350 board...
I am really waiting to see if AMD and the MB makers deliver the real AM4 compatibility or it is just the same story as before- to get the max of your new CPU buy a new board, except the case where you have top level board for 200+ EUR +, where the MB makers are risking nerd rage
So to see if the result is the same as Intel's (where you need to change boards because of chipsets and ...and with the nerd rage again Intel su...s again new board) or ..
The original leak is earlier than Chrisdar's Nov. 8 post, you can check the source link I PMed you.
So let's assume 150mm2 to be on the safe side, for a 14nm i3-8121 (as an added bonus the iGPU might work!). That's a 2.14x increase in density, which is not 2.7x as advertised by Intel.
It probably means Comet Lake-S won't come out in 2019.
The Comet Lake-S will use a new LGA 1200 socket. I'll be surprised if Intel doesn't rebrand the chipset to 4xx, even though it is the same KBL/CNL chipset.
AMD has implied that AMD can double performance for nearly free. I expect something like a qualifcation list for the the highest X chips if they go higher than lets say 100w rated TDP. But really I don't know why we thing AMD couldn't offer more cores at competitive clocks in the same power range. Core clock wise we don't have a good idea how offloading the IO is going to affect the ability to clock the chiplets either.So let's assume 150mm2 to be on the safe side, for a 14nm i3-8121 (as an added bonus the iGPU might work!). That's a 2.14x increase in density, which is not 2.7x as advertised by Intel.
Those low-end VRMs won't work. 7nm will yield greater efficiency, but higher core counts = more of the power consumption biasing towards current rather than voltage. That hits minimalist VRM configurations where it hurts. You want a lot of phases when dealing with low voltage/high current (look at the 9900k on Z370 vs Z390).
The 2.7x is technically not a lie. But its only true for small cores like Atoms. Their Core was always 2x. Higher performing transistors have to be larger, higher performing circuits may be more complex(thus take more space) and higher thermal density may require spreading them out more.
But here we start to see the idiocy of their decisions. They risked killing the future of the company to enter new markets. 4 year delay on 10nm for what? To make chips that lost 4 billion/year for Tablets and sell 5 million Smartphone chips to Asus.
AMD has implied that AMD can double performance for nearly free. I expect something like a qualifcation list for the the highest X chips if they go higher than lets say 100w rated TDP. But really I don't know why we thing AMD couldn't offer more cores at competitive clocks in the same power range. Core clock wise we don't have a good idea how offloading the IO is going to affect the ability to clock the chiplets either.
The choice is ofc good thing. THe bad is the presentation. All around the web tells us AM4 B350 is ok.How is choice ever a bad thing? You actually prefer being forced to buy a new board with each gen?
Want the best & latest tech? Buy a new motherboard with that new CPU
Want a cheap increase? Buy the new CPU and stick it in your present board.
Nobody is forcing anyone to keep their motherboard, and these new tech features like PCIe4, XFR 3, better memory support, keeps the motherboard manufacturers happy as it generates new sales.
Best of both worlds.