Monday, May 29, 2017

Opinion: Xiaomi’s Love for the SD625 (and Other Sidegrades) Impacts the Value of its Affordable Smartphones

When I had the opportunity to review the Xiaomi Redmi Note 3, I went in with low expectations. After starting my Android journey with a low-end device and struggling to get a better smartphone for a few years, I certainly did not look forward to downgrading from my OnePlus One (considered close to a flagship then) to a device competing in the budget segment.

But my preconceived notions as an amateur reviewer at that time couldn't have been more mistaken.

The Redmi Note 3 was a fantastic device which punched way above its selling price. The overall experience it offered did not resemble an 'entry level' device at all. Certain aspects of the device even came close to the flagship experience, and some like its battery life even went beyond that. As an occasional gamer, the Redmi Note 3 was a joy to use too — no signs of thermal throttling, even when subjected to intense gaming scenarios for long, sustained sessions, and a battery that let you stretch those gaming sessions even longer than usual. It ultimately redefined my understanding of low-end smartphones today, their potential, and how some companies can do it right.

A good chunk of credit for the Redmi Note 3's remarkable performance and experience went to the SoC inside of it. The Redmi Note 3 that was sold in India came with the Qualcomm Snapdragon 650, a mid-range hexa-core processor with 4x Cortex-A53 cores for daily tasks and power efficiency, and 2x Cortex-A72 cores for when you need that extra performance. It was built on a crusty 28nm process, but that didn't stop it – at the time – from punching above its segment. This particular combination resulted in smooth and efficient performance for daily tasks which was sustained in heavier use cases, too.

So when the Redmi Note 4 was announced, I was disappointed with the choice of SoC that Xiaomi opted for. The Snapdragon 660 was not official back then, so Xiaomi went in with the Snapdragon 625 – an octa core SoC with a dual cluster setup of Cortex-A53.

Again, I came out surprised with the resulting real-world performance of the device, mostly because I expected the lack of a heavy cluster to deeply impact the intensive usage of the device. As noted in my review, the Redmi Note 4 is still a theoretical downgrade when you consider where the Redmi Note 3 was sitting, because otherwise the Note 4 can hold its own in real world performance. But still, it was a marked, literal downgrade year-to-year and from one revision to the next, with slight advantages in power efficiency (but battery life was exceptional on its predecessor, anyway).

Even on the GPU end, the Snapdragon 625 with its Adreno 506 GPU performed worse when compared to the Adreno 510 on the Snapdragon 650. The Adreno 506 has a higher clock speed (650 MHz vs 600 MHz) and is built on the 14nm fabrication process, but it has lesser number of ALU's (96 vs 128) and manages to score lesser GFLOPS (130 vs 180). Benchmarking scores place the Adreno 506 well below the Adreno 510 as it managed to score about ⅔ of the framerates on the same benchmarks, pointing towards a marked downgrade in graphics performance.

A similar situation or "downgrade" crops up with the Xiaomi Mi Max and Xiaomi Mi Max 2. The original Mi Max came with a beefy Qualcomm Snapdragon 652 on the higher end variants — an SoC which was a slight step up from the 650 by adding an additional 2x Cortex-A72 to the performance cluster.

But with the Xiaomi Mi Max 2, we see Xiaomi switching lanes as the Qualcomm Snapdragon 625 makes an appearance again in this device. We repeat, the Snapdragon 625 is not a bad SoC by itself — particularly if paired with decent software, but the performance difference between the Mi Max and Mi Max 2 would be larger than between the Redmi Note 3 and Redmi Note 4. This immediately makes the Mi Max 2 a definite downgrade in terms of peak performance. Consumers looking to purchase the Mi Max 2 would have to rely on other changes to the device, like the larger battery and Quick Charge 3.0 capabilities through the USB Type-C port, in order to justify the purchase over the Mi Max. Xiaomi, like all other OEMs, also tends to cease production and sales of older devices when newer versions are released, so that its current products do not face competition from the value propositions of its past products.

The choice of Snapdragon 625 on the Mi Max 2 is a bigger deal than on the Redmi Note 4 because of the existence of (or close proximity to) alternatives. The Redmi Note 4 was announced when the SoC choices available to Xiaomi would be to move along the 65x lineup and choose the Snapdragon 652 or newer 653 with their caveats being that they were built on the 28nm fabrication process, now old and rusty. Or, Xiaomi could choose the SoCs built on the new 14nm fabrication process and opt for the Qualcomm Snapdragon 625 or newer 626. Choosing the 625 back then to focus on battery life was more digestible, but again, these phones really weren't lacking in terms of battery life anyway.

But with the Mi Max 2, most of the rumors and leaks surrounding the device indicated a higher-specced device coming in with the Qualcomm Snapdragon 660 SoC — a much more exciting SoC with its octa-core Kryo setup on a 14nm fabrication process. Coupled with the larger battery and Xiaomi's affordable pricing strategy, the Mi Max 2 would have been the beastly phablet to watch out for, a flagship-on-a-budget device for Xiaomi that would be one of the first devices to sport the 660, and show us what it can be capable of. That alone, I think, would have driven a lot of interest its way.

The Snapdragon 660 was announced in early May 2017 and the Mi Max 2 came out just a few days ago. The time gap between the two indicates that the Mi Max 2 may have already been past the drawing board and into production when Qualcomm presented its SoC upgrade to the OEM [Note that public releases obviously do not coincide with partner unveilings]. Thus for Xiaomi to upgrade to the 660 from the 625 would have meant starting back again from scratch on various aspects of product development.

It is also probable that Xiaomi knew of the existence of the Snapdragon 660 and went ahead with the Mi Max 2 and Snapdragon 625 anyways. This may have been because of the trickle down effects of production issues that have been plaguing the Snapdragon 835. With Qualcomm pushing forth heavily and focusing resources on the Snapdragon 835, there might not have been enough built up stock of the Snapdragon 660 for Xiaomi to go ahead with full scale production without delaying their product release. It might have been circumstantial, it might have been calculated, but either way it resulted in one of those rare instances in mobile tech where you can point at a clear step backwards (or backward-sideways) in a new device's specification.

Xiaomi's decision to go with the Snapdragon 625 on the Mi Max 2 may be wholly unrelated to Qualcomm as well. Xiaomi is known for half yearly 'product refreshes', which usually include the addition of the alphabets C, S or I to the name of the last released generation. So while the Mi Max 2 comes with the Snapdragon 625 SoC, Xiaomi could be looking at a half yearly upgrade with the Mi Max 2C/I/S (or a different alphabet for that matter) with the Snapdragon 660 SoC. The half yearly refresh could even feature Xiaomi's own in-house SoC, but we wouldn't count on that.

Going for the Snapdragon 625 may also be a conscious decision from the get go. The Snapdragon 625 is a good SoC for mainstream consumers and brings with it a balanced compromise of performance and battery life. The 14nm fabrication process allows for improved efficiency in a manner that the mainstream consumer could feel and appreciate, while taking away from the peak performance that mainstream consumers would seldom reach. The average Joe does not know, or care, about the differences between the 625 and the 650, so a swap out would be a calculated maneuver. But still, it's a rare occurrence to see a company willingly go for the "lower numbers" in such important lines, particularly some that are praised for their value or bang-per-buck.

Choosing the 625 would also help keep costs of the product down, letting Xiaomi pass on benefits to the consumer through lower selling prices or other additions that increase its product value. Part of this was reflected in the Xiaomi Redmi Note 4, which began sales at the same price as its predecessor while still offering more base storage.

At the end of the day, the Xiaomi Mi Max 2 with the Snapdragon 625 is what we have for now. The Max 2 and the Redmi Note 4 for that matter, did not generate as much excitement in me as their predecessors did. All of Xiaomi's smartphone products have great price-to-value ratio, but the Redmi Note 3 and the Mi Max just set the bar too high for Xiaomi to beat their own previous efforts. It is partly disappointing to see successors to some of Xiaomi's best smartphones come out with objectively-worse chipsets and not be true and complete year-on-year upgrades. Becoming complacent in this highly-competitive smartphone might just present a rare opportunity for another OEM to take advantage of these strategies.

With their future releases that already sport high capacity batteries, we hope Xiaomi goes back to more performance-focused SoCs and make use of 2017's exceptional-looking range of processors.

What are your thoughts on Xiaomi's recent adoption of the Snapdragon 625? What should Xiaomi do for future releases? Let us know in the comments below!



from xda-developers http://ift.tt/2qAfXhv
via IFTTT

No comments:

Post a Comment