Thread replies: 9
Thread images: 2
Thread images: 2
Maitland
Making a guide on Intel and AMD processor equivalents 2017-09-14 08:34:33 Post No.
[Report] Image search: [Google]
Making a guide on Intel and AMD processor equivalents 2017-09-14 08:34:33 Post No.
[Report] Image search: [Google]
File: samsung-32nm-wafer.jpg (300KB, 1000x734px) Image search:
[Google]
300KB, 1000x734px
Making a guide on Intel and AMD processor equivalents
Maitland
2017-09-14 08:34:33
Post No.
[Report]
I’ve been looking to update this crash-course guide to CPU’s and Motherboard types.
Intel’s line-up is quite confusing and AMD’s website reads more like a target pamphlet compared to the database that is Nvidea’s.
Here’s what I got so far for Intel:
Haswell-E (The E stands for Enthusiast)
Haswell-EP (Server)
Haswell-DT (Desktop)
The X-series are a collection of i5, i7, and the more recent i9 (Where’s the i8?) processors.
The numbers 5820 and up can all support the DDR4 Ram, but numbers 4960 and below are either the Sandy Bridge or Ivy Bridge, which can only use the DDR3 Ram.
All X-series numbers 5820K to 5930K and 5960X, use the Haswell x86 architecture on a 22nm process. The 5820K is part of the X-series, even though there is a K on the end of it.
These Haswell-E processors have 1.4 Billion transistors on a 13x14mm die.
The maximum amount of RAM supported is 64 GB in 4 slots (Quad-channel) and has a Bandwidth of 68 GB.
Are these processors considered to be the 4th generation of the 22nm process? The 4th tock?
Numbers starting at 6800K to 6950X and between, use the Broadwell x86 architecture on 14nm.
These are all marketed as i7, support a max RAM size of 128 GB in 4 slots, and have 40 PCI lanes, except for the 6800K which only has 28. Each Graphics card requires 16 PCI lanes, while an SSD needs 4 or 8. So a CPU can with 40 PCI lanes can support 2 Graphic Cards and 1 very fast SSD (or 2 slower SSD’s).
Haswell and Broadwell architectures use the LGA 2011-v3 socket, which require the X99 chipset.
Why does the 6850K cost about $200 more than the 6800K when there’s only a 0.20GHz difference?
>>
Continued:
https://trello.com/c/cXDDxrAk/2-ddr4-computer-ram
The numbers 7640X to 7740X are using the Kabylake architecture on the 14nm process.
These processors (including those of Skylake) use LGA 2066 pin socket, which requires the X299 chipset (The successor to the X99 and its LGA 2011-v3 socket).
Finally, all numbers beginning with 7800X to 7920X have the Skylake architecture (Also on 14nm).
When Broadwell is said to be the 5th generation, is this referring to the 5th gen 14 nm process?
While Kaby Lake is the 3rd gen of the 14 nm process which uses the same Broadwell architecture?
Are these generations (‘Tocks’?) the optimisation of the same ‘tick’ (Micro-architecture)?
What important information needs to be included or corrected?
Now, since I know almost nothing about AMD and Radeon, could you provide an equivalent explanation on what these two products have to offer? How does their line-up work? With Intel, the chips made for Enthusiast-grade consumers seem to be grouped together under the X-series and have separate price-ranges listed as i5, i7, and i9.
Memory Speeds (Fastest to slowest): Registers, Cache, Main Memory (RAM), Hard-Drive (HDD).
As the memory size decreases, its speed increases.
>>
Continued:
The numbers 7640X to 7740X are using the Kabylake architecture on the 14nm process.
These processors (including those of Skylake) use LGA 2066 pin socket, which requires the X299 chipset (The successor to the X99 and its LGA 2011-v3 socket).
Finally, all numbers beginning with 7800X to 7920X have the Skylake architecture (Also on 14nm).
When Broadwell is said to be the 5th generation, is this referring to the 5th gen 14 nm process?
While Kaby Lake is the 3rd gen of the 14 nm process which uses the same Broadwell architecture?
Are these generations (‘Tocks’?) the optimisation of the same ‘tick’ (Micro-architecture)?
What important information needs to be included or corrected?
Now, since I know almost nothing about AMD and Radeon, could you provide an equivalent explanation on what these two products have to offer? How does their line-up work? With Intel, the chips made for Enthusiast-grade consumers seem to be grouped together under the X-series and have separate price-ranges listed as i5, i7, and i9.
>>
>why does the 6850K cost $200 more for a 200MHz bump
((()))
No but seriously, ((()))
AMD's consumer stack is easy.
R7 = 8C/16T, various clockspeeds
R5 = 6C/12T or 4C/8T
R3 = 4C/4T
For all of them, X adds an extra couple hundred MHz of automatic overclock and better binning. All processors are unlocked, B350 and X370 boards support overclocks.
>>
>>62432455
Tick tock broke down after Broadwell. Skylake was the last tick. Kaby and Cannon Lake are neither new architecture nor new process nodes, they're "refinements".
Radeon is currently:
RX(XY0), where X is generation and Y is SKU level. RX570 is the "-70" level SKU, of gen 5.
Vega is a fucking meme, much like the Fury X was. It doesn't fit into the normal convention because it's basically a marketing stunt. Expect RX680s next year using improved Vega architecture.
>>
>guide
no one has an attention span long enough for reading make an infograph instead
>>
>>62432413
>while an SSD needs 4 or 8
you're retarded anon. no one here is going to have any modern SSD which requires 8 lanes. M2 uses 2 or 4
>>
File: 6173330+_8a79d7cee099af053da6ca522b141bc9.png (65KB, 326x242px)
65KB, 326x242px
>>62432538
Do I sound like an American to you?
>>
>using NSA hardware
EXCUSE ME?
Thread posts: 9
Thread images: 2
Thread images: 2