This site may earn affiliate commissions from the links on this page. Terms of use.

Ever since AMD announced Carrizo, nosotros've been cautiously optimistic about the chip's performance and improvements over Kaveri. No one expected miracles from AMD's sixth-generation APU, but the new chip included additional CPU improvements, significantly better video playback, and information technology was AMD's offset fully-integrated SoC. There was a chance that Carrizo could win AMD some respect in the low-power 15W markets where the fleck was meant to play.

An absolutely enormous review from Anandtech dashes those hopes. First off, when I say "enormous," I mean it — the five-style comparison betwixt four Carrizo systems (built by HP, Toshiba, and Lenovo) and one Kaveri organization totals over 23,000 words. Nosotros'll hit some of the high points here, but if you're curious to come across the subtleties, nosotros highly recommend reading the original piece.

Carrizo vs. Kaveri

Anandtech first compares Carrizo against Kaveri before taking it out against Core Thou. The proficient news in this match-upwardly is that Carrizo often delivers the improved functioning and reduced power consumption AMD promised. In a scattering of cases, the gains are considerable — bank check Cinebench 15's unmarried-and-multi-threaded performance graphs below for an case of how much Carrizo tin ameliorate over Kaveri in a all-time-case scenario.

CB15MT

Carrizo vs. Kaveri in Cinebench. This is a best-example scenario for the new silicon. Images by Anandtech.

CB15ST

Carrizo vs. Kaveri in Cinebench. This is a best-case scenario for the new silicon. Images by Anandtech.

Fifty-fifty the slowest Carrizo APU is eleven% faster than Kaveri in unmarried-threaded style and fifteen% faster in multi-threaded mode.

Unfortunately, all of the Carrizo systems tested share a common handicap — they all transport in unmarried-channel retention configurations. The HP Carrizo systems can all exist upgraded to dual aqueduct functioning, at to the lowest degree, only the Toshiba and the supposedly high-end Lenovo rigs are both stuck in single-channel manner.

The reason for this is price. When AMD launched Carrizo and Carrizo-L (refreshed Beema), information technology talked virtually how using a unified platform for both fries would allow OEMs to salve money on blueprint costs. Unfortunately, Carrizo-50 is a single-channel design. Instead of building motherboards that could handle two channels and only disabling one when using Carrizo-Fifty, several manufacturers opted to limit Carrizo to a single aqueduct in the first place.

That single-channel isn't fifty-fifty loaded with high-speed RAM. The Carrizo systems AT evaluated all used DDR3L-1600. That'south a non-trivial additional operation hitting in GPU workloads; AMD'south APUs typically encounter near-linear performance gains from increasing the RAM clock. That'due south particularly true in a single-channel system, where bandwidth is even more limited.

Compared with Kaveri, Carrizo improves in multiple areas, including a huge reduction in video playback power consumption and official support for smooth 4K video. Unfortunately, that's not the whole story.

Carrizo vs. Core M / Broadwell

The Core M / Broadwell comparisons aren't and so bully for AMD. Carrizo sometimes wins against Core K (albeit with a much higher TDP limit) and in ane test (POV-RAY) fifty-fifty manages to compete with Broadwell. By and large, however, it's stuck betwixt the two form factors.

Carrizo still competes better against Cadre Thousand / Broadwell than Kaveri did, fifty-fifty handicapped by single-channel memory, but here's where the screws really bite. In theory, Carrizo should be showing up in cheaper SKUs. AMD charges $150 for the Carrizo processors AT compares, versus Intel's $300 for the Cadre M / Broadwell flavors.

In practise, OEMs are building single-channel Carrizo systems with $150 APUs and pricing them every bit against dual-channel Intel systems with $300 SoCs. This doubly penalizes Carrizo. Not only is it going upwardly against Intel processors with far more memory bandwidth, it's not being allowed to compete on price.

OEM sabotage

HP's $one,049 Elitebook 840 G3 has an Intel Core i5-6200U and 8GB of DDR4-2133 in a dual-aqueduct configuration. The Dell XPS 13 uses the same CPU, but offers 8GB of LPDDR3-1866.

HP's AMD-powered Elitebook 745 G3, in dissimilarity, has just 4GB of DDR3L-1600 and the same price tag as the $1,049 Intel organisation. Given that we know AMD is selling the SoC for $150 less than the official list cost on the Intel counterpart, that should heighten more than a few eyebrows.

"Sabotage" is a strong word to throw around, simply in my opinion these system prices (and the by and large depression quality of the Carrizo systems themselves) warrant it. Even in a best-example scenario, Carrizo can't exist said to lucifer Core Grand or Broadwell — but the functioning gap between AMD and Intel should exist kickoff by a lower arrangement price. The supposed toll savings AMD offered OEMs past designing Carrizo and Carrizo-L to a common platform aren't being passed on to the consumer, and neither are the price savings from using an AMD APU.

Anandtech dives into how positioning Carrizo / Carrizo-L as unified low-price selections may have encouraged OEMs to handicap the final products out of the gate. Equally things stand today, without significant discounting, information technology's hard to recommend Carrizo hardware — not because AMD didn't improve over Kaveri, but because OEM decisions and pricing take crippled the platform.

When AMD briefed us on Carrizo nearly a year ago, it made information technology clear that the chip was targeting the $400 to $700 laptop market, with Carrizo-L shipping into the $250 to $400 range. Betwixt $400 and $500, AMD would exist competing confronting the lower-terminate Cadre i5-5200U from Intel, also as various Haswell-era Celerons and lower-end Intel parts. Instead, OEMs are edifice low-quality hardware at premium prices, tossing Carrizo against Intel fries it wasn't meant to compete confronting. This feeds a vicious cycle — consumers who buy AMD and are dissatisfied with the result are less inclined to consider AMD in the future — for problems that ultimately, are beyond Chimpzilla's control.