If it’s not obvious enough (somehow), we’re in the midst of an incredible time for processor tech. Remember when the first 1GHz chips hit the market? That was exciting. It was also exciting when we saw the world’s first dual-core chips. Fast-forward to right now, and it’s exciting because after a decade of iterative releases from Intel, and a general lack of competition from AMD, we’re now seeing a real fight waged. And, by many accounts, the underdog is winning.
When Intel unveiled its Core X-series of processors at Computex last week, I didn’t put that much thought into it because of everything else that was going on. Since being back at home, though, I’ve found myself pondering about this lineup, ultimately curious about some of Intel’s decisions. Looking around the web, I can tell that I haven’t been alone.
It seems certain at this point that Intel didn’t quite anticipate the punch AMD was going to deliver with Ryzen, and based on that series’ horsepower, Threadripper on the horizon, and Zen soon to hit the enterprise (as EPYC), Intel is finding itself in a bit of a rough place. Before Computex, I didn’t really expect Intel to unveil a lineup like X-series, but the landscape right now is not normal.
In some ways, I feel like Intel is being a little too hasty with its actions, and the Core X-series naming is a big point that ties into that. When looking at its Ark product database the other day, I noticed that Intel had renamed its “High End Desktop” category to “Core X-series”. In doing so, Intel retroactively applied the Core X-series branding to all previous high-end desktop SKUs. This means that Core X-series isn’t just a name to represent a one-off launch: it’s the new name for Intel’s enthusiast-grade chips. 8th gen (and future) Core should sport the same name for high-end SKUs.
There’s nothing wrong with that move, per se, but I feel like Intel is bastardizing what its “high-end” chips are. That’s not just odd, it’s going to lead to confusion. Heck, it’s confusing me.
Take a look at Intel’s Core X-series lineup, as it stands today:
|
Clock |
Turbo |
Cores |
PCIe |
Memory |
Power |
Price |
i9-7980XE |
– |
– |
18 (36T) |
– |
– |
– |
$1,999 |
i9-7960X |
– |
– |
16 (32T) |
– |
– |
– |
$1,699 |
i9-7940X |
– |
– |
14 (28T) |
– |
– |
– |
$1,399 |
i9-7920X |
– |
– |
12 (24T) |
– |
– |
– |
$1,199 |
i9-7900X |
3.3 GHz |
4.3 GHz |
10 (20T) |
44 |
Quad |
140W |
$999 |
i7-7820X |
3.6 GHz |
4.3 GHz |
8 (16T) |
28 |
Quad |
140W |
$599 |
i7-7800X |
3.5 GHz |
4.0 GHz |
6 (12T) |
28 |
Quad |
140W |
$389 |
i7-7740X |
4.3 GHz |
4.5 GHz |
4 (8T) |
28 |
Dual |
112W |
$339 |
i5-7640X |
4.0 GHz |
4.2 GHz |
4 (4T) |
28 |
Dual |
112W |
$242 |
Etc |
Before this lineup was revealed, no one in their right mind would have called a Core i5 chip without HyperThreading that only supports dual-channel memory a “High End” chip – yet, here it is in this lineup along with the actual high-end chips. While it’s nice to have an overarching brand, like X-series, the inclusion of many chips in the above lineup make a complete mockery of the X moniker. “X” in Intel’s lineups has always referred to “Extreme Edition”, and today, the bottom rung of the above lineup does not deserve that. “Extreme” and “dual-channel memory controller” don’t go together.
After handing out Xs like candy, Intel must have realized that it’s a bit odd to put a $1,999 chip in the same class as a $242 one, so “XE” was introduced. I have a beef with this one, too, because “XE” on only the top-end chip implies that the $1,199, $1,399, and $1,699 chips are somehow undeserving, even though they should absolutely be XE, as well. I’d even argue that the $999 i9-7900X deserves to be included.
Is this entire lineup the result of Intel simply being too hasty in response to AMD? It’s hard to say, and only those working inside Intel know the answer for sure.
While this post was created more to talk about the odd branding than anything else, that branding is really the least of Intel’s worries right now. With Threadripper, AMD is planning to deliver an almost “too good to be true” solution, one that is rumored to give us 16 cores and 32 threads for under $1,000, and 64 PCIe lanes, across the entire range of TR chips. Intel’s $1,000 solution will give us 44 PCIe lanes and 12 fewer threads.
One thing that works to Intel’s favor is that its IPC (instructions per clock) performance is better than AMD’s, which means single-threaded performance will generally be better on its chips. In many scenarios (like general computing), that’s important. In situations where more cores benefits a task, processors (like Threadripper) that can offer more cores for less money can come well ahead.
Whether you’re an AMD or Intel fan, hopefully you’re enjoying the show. There is so much good stuff going on right now, it feels like we’re trying to make up for an entire decade all at once.