The ACTUAL Difference Between Intel and AMD

The ACTUAL Difference Between Intel and AMD

As much as AMD and Intel fanboys like to scream at each other online, the average user would be hard-pressed to notice the difference between these companies' CPUs in everyday use – unless you're like us and benchmark things religiously. But on the inside, modern chips from Team Red and Team Blue are built quite differently. So let's take a look at how, starting with AMD.

AMD Ryzen Architecture AMD Ryzen CPUs are made up of at least one Core Complex or CCX. Despite the name, it's not a complex thing to understand. A CCX is just a group of cores along with some cache memory. Newer Zen 3 chips have CCXs with 8 cores, while Zen 2 CPUs have 4 cores per CCX. Each CCX lives on a chiplet called a CCD.

The setup on Zen 3 is straightforward: one CCX per CCD, with up to 2 CCDs per processor for a total of 16 cores. On Zen 2, each CCD can instead hold two of these smaller CCXs, resulting in the same 16-core maximum.

Infinity Fabric What links all these core complexes together? AMD's Infinity Fabric, a high-speed interconnect that can run between CCXs or between CCDs. Although it works well with Ryzen beating out competing Intel chips in benchmarks, it’s still not as fast as having the cores more directly connected because there’s higher latency with Infinity Fabric, and the connection is serial rather than parallel. However, there are real benefits to this approach.

More smaller chiplets mean big cost savings due to better yield. If one chiplet is bad, you can discard just that one without wasting as much silicon from throwing away a whole CPU. This CCX plus Infinity Fabric approach also scales quite well so you can just add more chiplets if you need more cores instead of spending more time and money designing a whole new chip.

Intel’s Approach Intel, in contrast to AMD, still uses monolithic chip designs where cores are right next to each other, at least for the time being. This approach allows the cores to talk with each other more directly but it has the opposite problems as AMD’s chiplet design – lower yield, less scalability, and higher cost.

But Intel isn't going to be left behind as the industry moves toward chiplets. They call them “tiles” instead because they just had to be different. In 2023, we’re expecting a lineup of mainstream processors called Meteor Lake, based on tiles connected by something called EMIBs. Although it has a much less cool name than Infinity Fabric, we've already seen EMIBs in the Kaby Lake G processors and Intel’s Ponte Vecchio supercomputer chips.

EMIB Technology According to Intel, EMIBs (Embedded Multi-die Interconnect Bridge) offer a more elegant solution to AMD’s Infinity Fabric. Instead of having a large slab of silicon that the chiplets sit on, an EMIB is a smaller piece of silicon that connects tiles on either side. This approach reduces silicon usage, keeping costs lower, and allows easier scalability by using multiple EMIBs to build outward.

Unlike Infinity Fabric, EMIBs move data in parallel, giving them higher bandwidth while also offering lower latency and less power loss. However, there are probably limits on how many tiles can be connected before hitting a thermal ceiling.

Foveros Technology Intel also has another method to connect chiplets called Foveros, which involves stacking chips on top of each other. But this hasn't seen much adoption in the consumer space yet, as Lakefield, the specific lineup that featured Foveros chips, was discontinued due to poor demand.

Future Perspective Even though both Intel and AMD are trying to break up the classic monolithic chip design, it won't go away completely. These fancy methods of linking chiplets together cost money, so simpler monolithic designs for systems that don’t need tons of computing power will be with us for quite some time.

So if you’re in the market for something cheaper but really want a chiplet, there’s always those broken Pringles at the bottom of the can.

Keywords

• AMD
• Intel
• Ryzen
• CCX
• CCD
• Infinity Fabric
• Monolithic Architecture
• EMIB
• Foveros
• Meteor Lake
• Zen 2
• Zen 3
• Chiplet Design
• Scalability
• Silicon Yield

FAQ

Q: What is a Core Complex (CCX) in AMD processors? A: A Core Complex, or CCX, is a group of cores along with some cache memory. In newer Zen 3 chips, each CCX contains 8 cores, while Zen 2 CPUs have 4 cores per CCX.

Q: How does AMD connect its core complexes? A: AMD uses a high-speed interconnect called Infinity Fabric to connect core complexes (CCXs) and chiplets (CCDs).

Q: What is Infinity Fabric? A: Infinity Fabric is AMD's interconnect technology that runs between CCXs or CCDs. It provides high-speed data transfer, although it has higher latency and a serial connection compared to more directly connected cores.

Q: How does Intel's monolithic chip design differ from AMD's chiplet design? A: Intel's monolithic design places all cores next to each other on a single piece of silicon, whereas AMD uses separate smaller chiplets connected via Infinity Fabric. Monolithic designs offer lower latency but come with higher costs and lower scalability.

Q: What are EMIBs used by Intel? A: EMIBs (Embedded Multi-die Interconnect Bridge) are small pieces of silicon that connect tiles on either side, facilitating parallel data transfer with higher bandwidth and lower latency compared to serial connections.

Q: What is Foveros technology? A: Foveros is Intel's technology for stacking chips on top of each other. Though it has potential advantages, it has not seen widespread adoption in the consumer market yet.