Hacker News new | ask | show | jobs
by jzb 1147 days ago
“The “i” branding is extremely powerful with the average consumer and removing it is bound to cause confusion amongst consumers who aren’t technically-inclined.”

Who’s the “average” consumer here? I’m skeptical that the majority of PC buyers have internalized the “i” branding. Gamers, sure. But average consumers? I’ve talked a number of less computer savvy friends and family through computer purchases and there was awareness of Intel, but beyond that not so much.

I don’t think a rebrand is going to change much for Intel either way. Average consumers are either going to buy by price point and familiar brand (Lenovo, Dell, etc.) or phone a friend to get advice. Or use whatever they get handed by their employer.

4 comments

While somewhat ridiculous, I kind of feel like they might’ve been better off if they were currently selling the Pentium 17.

Anyone can look at 17 and compare it to the number on their older computer (say Pentium 14) and know it’s better.

Apple is doing this right so far with the M chips, but they’re about to be on #3. We’ll see how long they stick with it. They have so far in phones.

Apple is doing this right so far with the M chips

These?

    M1
    M1 Pro
    M1 Max
    M1 Ultra
    M2 Pro
    M2 Max
Besides the non-numerical suffixes being harder to sort, it still doesn't make it easy to compare performance between them at a glance. I will give them credit for not adding unnecessary "padding zeros", however.
Apple is not doing it right.

You have the "MacBook Pro with M1 Pro" which makes my eyebrow twitch.

You also have the "M1 Max" which is lower grade than the "M1 Ultra". Despite it you know, being the MAX.

I will say that this branding while uncharacteristically sloppy for Apple, is better than bullshit like the "Intel Platinum i6-43065GF Extreme edition" which is simply incomprehensible to the untrained eye while the trained eyes roll in the back of their skulls.

The thing is that these companies have so many SKUs because they're trying to address so many market segments and so newer is not always automatically better. Then technology (and competition) advances at an uneven pace so you end up with some generations (2 -> 3, 6 -> 7, 10 -> 11) where the generational difference is "who cares", and some generations (9 -> 10, 11 -> 12) where basically the entire old generation is obsolete.

Apple is tackling an easier problem since they're focusing on subdivisions of the high margin premium consumer segment. But there's a world of difference between the €1400 base model m1 macbook air and the €550 windows machines the standard consumer here buys. A lot of Apple's target market looks at the €1400 MBA doing quite well against comparably priced windows machines (well as long as you aren't gaming, running legacy business applications, or actually in need of more than 8gb of RAM), and sees the €950 price difference to what people actually buy as irrelevant, but for e.g. my parents that pays for their next laptop in a decade too.

Once you're into that price sensitive market, you end up needing from a business imperative that smattering of SKUs, since the person selling a €600 laptop wants something they can point to as better than the €550 laptop and at that market, people are not buying €50 for lighter or more aesthetically appealing laptops. Then a sort of mirror happens at the high end since it's users buying CPUs directly and therefore more informed so it's not as easy to just make some tiers and extract maximum consumer surplus for the higher tiers.

With the i branding, a desktop vendor could sell an “i9” using a less performant, old i9 chip that could perform much worse than a new i5. Conversely some i3 chips are rather good but at a glance a consumer will think they are too low-end. This is bad for Intel, a consumer may buy an i9 machine and think “wow, this is terrible for a highend chip, Intel really sucks” because it was a bad i9 or prefer a worse-suited i5 over an i3 due to assuming 5>3.

I’d prefer a more literate convention I described in other comments, but the new convention may at least prevent reductively bucketing skus. The old buckets were just descriptive enough to suggest a quality and performance hierarchy, but dangerously so, because even a 3 could be better than a 9 in some cases.

> “The “i” branding is extremely powerful with the average consumer and removing it is bound to cause confusion amongst consumers who aren’t technically-inclined.”

I feel like you're either in one of two groups:

1. a PC "nerd" (kind of deragotory) who is obsessed with specs all the way down to things like L2 cache size, etc.

2. somebody who has 0 clue what any computer/laptop specification means whatsoever

> somebody who has 0 clue what any computer/laptop specification means whatsoever

Life was so easy in the MHz-wars days.

The bigger number was basically always better, and you could easily tell by how much.

Nope, the IPC was hugely different in the MHZ wars days.

Pentium III had 20% higher IPC than Pentium 4 Northwood; AMD K8 (Athlon 64) had 35% higher IPC than Pentium 4 Prescott.

Same in the RISC vs CISC days: Pentium had 33% higher IPC than the original PowerPC 601. Late generation RISCs like Alpha were a completely different beast again, with the DEC Alpha 21264 (EV6) having 25% higher IPC than Pentium III.

It was never obvious back then, and it was also never obvious since then.

IPC keeps rising among different architectures. A couple years ago Apple overtook Intel, and ARM did as well.

Yeah it never worked across architectures, but up until the P4 came out it was quite reliable within a family.

The P4 is what killed the MHz Wars in my memory. It was clear it couldn’t keep clicking up and it’s poor performance relative to existing chips at similar clock speeds (as you noted) meant the game was over.

There’s also the “average PC gamer” that has some clue about computer parts, but wouldn’t be able to tell you what something like L2 cache is. I’d say that’s quite a large part of the consumer base as well.
I'm inclined to agree. Intel's iX branding sort of make sense at one point, i3 meant entry level/budget, i5 was mainstream, i7 was enthusiast.

Except actually Celeron was entry level, and i3 was entrish-level. Also i7 wasn't enthusiast, no you needed an i9 processor to be a REAL enthusiast. Also now mainstream tasks can run just fine on the i3's if not the Celeron chips.

I long thought their branding was over-convoluted MBA mess which was overdue for some simplification.