Intel’s CPUs use way more power than AMD’s latest chips.
Intel (INTC 3.12%) stumbled through manufacturing issues in the past, and only now is it starting to right the ship. The company’s entire desktop CPU lineup is still made on its Intel 7 manufacturing process, which is based on the same 10nm node that was originally slated for high-volume production in 2016. Meanwhile, rival AMD (AMD 4.52%) has taken advantage of TSMC‘s manufacturing progress with each new release.
A massive efficiency edge for AMD
Despite Intel’s current manufacturing disadvantage, the company has managed to stay competitive with AMD in terms of performance. Intel’s Raptor Lake Refresh desktop CPUs, launched in late 2023, continue to hold up well even as AMD launches its Ryzen 9000 series. Intel’s high-end 14900K, for example, beats AMD’s brand-new and more expensive Ryzen 9 9950X in single-threaded and gaming workloads.
Intel has had to sacrifice power efficiency to keep pace, and the gap between Intel and AMD is now a mile wide. In Tom’s Hardware’s review of AMD’s Ryzen 9950X, Intel’s 14900K used nearly 30% more power in some cases. The difference is even larger in the mid-range portion of the market, where Intel’s 14600K sometimes used twice the power of AMD’s Ryzen 9600X.
Power consumption and efficiency aren’t quite as important in desktops as they are in laptops, where battery life is a key selling point. However, those building gaming PCs or PCs for heavy-duty productivity tasks should have a preference for more efficient chips. A desktop PC with a power-hungry CPU paired with a graphics card can easily become a heater if it doesn’t have adequate cooling.
Arrow Lake could be the solution
Intel plans to skip over multiple generations of manufacturing technology for its follow-up to Raptor Lake Refresh. The company’s upcoming desktop CPU family is code-named Arrow Lake, and it’s set to launch before the end of the year. Arrow Lake is expected to be manufactured on the Intel 20A process, which should be a massive step forward from Intel 7. Intel 20A is a “2nm-class” process node, while Intel 7 is a “7nm-class” process node.
Intel hasn’t yet revealed much about Arrow Lake, but rumors and leaks point to a dramatic improvement in power efficiency. One report out of China states that Arrow Lake chips will consume 100W less power than Raptor Lake in some cases. For reference, Intel’s 14900K can use around 250W under heavy loads. If these reports are accurate, Intel could completely close the efficiency gap with AMD.
Questions remain
While Arrow Lake has the potential to vastly improve Intel’s competitiveness in the desktop CPU market, some questions remain. The company has struggled to churn out enough of its Meteor Lake laptop chips, which are built on its Intel 4 process, and it’s unclear how much volume the company will be able to muster for the Arrow Lake launch.
For Lunar Lake, Intel’s next-gen laptop CPU, the company opted to largely outsource manufacturing to TSMC, potentially because it wouldn’t have enough Intel 20A capacity for both product families. Time will tell whether Intel can produce enough chips for its purported efficiency gains to translate into market share gains this year.
AMD has a clear edge over Intel in the PC CPU market when it comes to power efficiency. With Arrow Lake, Intel has a chance to erase that edge as it plots its comeback.
Timothy Green has positions in Intel. The Motley Fool has positions in and recommends Advanced Micro Devices and Taiwan Semiconductor Manufacturing. The Motley Fool recommends Intel and recommends the following options: long January 2025 $45 calls on Intel and short August 2024 $35 calls on Intel. The Motley Fool has a disclosure policy.
