Intel: A Core New Low February 14, 2006 Jakub Wojnarowicz

AMD vs Intel

Page:: ( 1 / 4 )

I am not someone who is an AMD fan in particular, nor do I have anything against the company. I like competition. I think AMD’s surge forward is an excellent event for the computer industry and it provides pressure on its competitors (namely Intel) to provide a better product at a lower price. While prices had been falling during Intel’s monopoly, AMD’s rise exerted extra pressure on the market and has led to the excellent conditions we have now. With that out of the way…

There’s a tendency in every industry, especially by the youth market, to rip on the “big guy”. In computers, the targets have been SUN, Microsoft, Novell, IBM, Electronic Arts, and so on. People like cheering for the underdog, whether out of a desire to be different or from a genuine sense of fostering competition, that depends on the individual. The point is that big companies are rarely popular.


Intel had managed, for most of its existence, to avoid this stigma. It was a popular company well into the 2000s, and certainly there was no common dislike of the firm until the late 1990s. The reasons for Intel’s continued popularity are many. For starters, unlike many big companies, Intel was rarely content to rest on its laurels. It relentlessly pushed technology and processor speeds, and the market was always able to gobble those extra CPU cycles up. The 8088 gave way to the 8086, then to the variants of the 286 which were quickly replaced by the 32-bit 386 (available in both DX and SX models in various speeds, with a 32-bit and 16-bit bus respectively), which was replaced equally quickly by the 486. Finally, came the Pentium and its successors.

A top-of-the-line 386 PC would cost in the neighborhood of $6000-12000 when the 386 was just launched. When the Pentium II was launched, it was one of the pricier chips Intel had made in some time for the consumer segment, yet prices had fallen to under $4000 for a top end rig by then. When faced with competition from AMD’s K6, Intel launched the Celeron and quickly improved it to make it competitive even though that undercut the market for the Pentium II. Such actions left little doubt that Intel was an aggressive company determined to hold onto market share, to innovate even if it cost money in the short run. That Intel, the old Intel, led the way.

How times have changed.

The problem is that Intel isn’t responding to the AMD challenge. Or rather, the problem is that Intel is merely responding. There was a time when Intel dictated the features processors would have. Intel wasn’t the company responding to competitors, rather it was they who were playing catch-up.

How low can Intel go?

Page:: ( 2 / 4 )

While it is difficult to tell when Intel lost the initiative internally, it is not difficult to tell where it began losing it relative to the competition. That occurred in 1998, with AMD’s launch of the K6-2. Until the K6-2, AMD was at best able to offer some extra clock speed over Intel’s chips (remember the 486DX-40?) The K6-2 introduced “3DNow!”, a much-hyped technology which promised to address the K6’s poor floating point performance. With the K6, AMD had simply been caught flat-footed by the gaming market and the move to 3D. The original K6 design was actually superior to the Pentium in integer performance, clock-for-clock, and it was scaling better to boot.

Then Intel launched the Pentium II, a processor based off the Pentium Pro core that featured two major improvements – one was that it was focused from the start on 32-bit applications rather than mixed 16-bit and 32-bit. The second improvement was a massive gain in floating point capabilities. Whereas the K6 was capable of roughly 70-80% of the Pentium’s floating point calculations per clock, the Pentium II stomped it silly, delivering double what the K6 could do.


Designing a new processor is not easy. It is expensive, time-consuming and a generational investment. The 386, 486, Pentium, and Pentium Pro/II/III line all saw about a four- to five-fold gain in clock speed over their existence. 386s started at 8MHz on the low end and reached 40MHz. The 486 started as slow as 25MHz and reached 100MHz (120MHz in AMD’s variations). The Pentium Pro/II/III line, technically still alive today in the Pentium M, started at 200 and went up to over 1GHz. Those three processor cores were very closely related, so distinctions between them are too thin to worry about.

As you can see then, AMD faced the terrible problem of being significantly behind in a key area of its core with no easy way to update it. Their solution was 3DNow!, a SIMD set of operations for floating point calculations. In retrospect, it’s a fairly ugly solution and a hack. Even at the time, most knowledgeable computer users weren’t keen on it because it required – at the least – a recompilation of existing code by a game developer to take advantage of it. A better option was to write a game with 3DNow! SIMD instructions in mind from the get-go. And really, who would do this for a company as small and unknown as AMD?

Except, luckily for AMD, the internet was gaining steam at the same time and slowly but surely all the myths built around non-Intel processors were being wiped away. Just like 10 years prior there was a lot of concern about PCs being “100% IBM PC compatible” – meaning not junk like Tandy computers – in the 1990s there was worry that processors weren’t compatible with each other. This was partly FUD, partly caution with such pricey investments, and definitely not helped by the issues surrounding Cyrix processors. With the early internet there was also a lot of openness and interaction by hot, young development studios like id Software, Epic, 3DRealms and others. Gamers were keen on getting more bang for their buck so they harassed developers into including at least nominal 3DNow! support. It still wasn’t as good as a Pentium II, and Intel quickly retaliated with SSE in the Pentium III, but that’s the point – Intel had to respond. It was not the first on the market. (MMX provided SIMD for integer performance a few years prior, but no one cared about that even when integer performance was key.)

Intel responds, again

Page:: ( 3 / 4 )

From that point on, we see Intel increasingly in the position of responding to AMD. Intel was no longer setting the tempo all of the time, and as the years flew by, it would increasingly find itself reacting rather than being pro-active.

AMD was first to 1GHz with the Athlon. AMD bet on 64-bit x86 while Intel was pushing Itanium and the EPIC architecture as the next generation. The whole time, Intel was denying it would make a 64-bit Pentium 4 while secretly working on one. Heck, the entire Pentium 4 and NetBurst architecture didn’t live up to expectations either.


Take a look back at Brandon’s Pentium 4 Performance Preview. You can almost smell the expectations Intel had of creaming AMD. Not only was Intel expecting to outperform the upstart Athlon, but they knew they had the marketing MHz advantage. The P4 traded blows with the Athlon, but dominated key benchmarks like 3DMark and Quake III Arena. Moreover, NetBurst promised almost limitless MHz, it had pipelines way longer than anything else out there. Sure, its per-MHz performance was disappointing, but it was going to have 50% or more cycles per second than the best comparable Athlon, right?

Of course, the failure of the Pentium 4 isn’t quite Intel’s fault. What the company encountered, the new roadblock to faster clockspeeds, was foreseen - just not expected as quickly. The problem is of heat density – the amount of watts put out by a specific area. With a smaller area relative to the number of transistors and the clock speed, you end up with a more focused point of heat. The limited area means the contact patch of a heatsink is small, while the heat is greater than previous processors. Intel bet that it wouldn’t hit the wall until sometime around 2010. For whatever reason, they were wrong and got stopped dead in their tracks in 2005.

The heat issues were unexpected, but our disappointment lies more with Intel’s inability to lead the way. They focused too much on what they have to lose rather on what they have to gain, they put the customer in second place to their own goals. Intel’s plan was to live out the Pentium 4 in 32-bit, slowly transition away from x86 towards EPIC and the consumer successors to Itanium. EPIC would be a proprietary architecture, with no nasty competitors like AMD or Cyrix around to spoil the party. It wouldn’t have any of the legacy problems of x86 (which are far too many to list), but consumers would also lose out on choice. If Intel had pioneered 64-bit x86, that would have undercut their plans for EPIC. They thought about what they had to lose, so they didn’t do it. If Intel had thought about the consumer, they could have had 64-bit processors before AMD.

Intel could also have had dual-core before AMD, and possibly a superior implementation rather than the inferior design out now. Yet again, the company got held up by not wanting to undercut its glorious Itanium dreams for the future, rather than trying to lead the market. And yet again, they got burned for it by AMD.

The new low

Page:: ( 4 / 4 )

Intel’s solution to its current problems is to look to the past, to the abandoned Pentium Pro family – now represented by the revamped Pentium M. This isn’t a bad thing. Due to the unexpected heat issues at high clock speeds, a more clock-efficient processor design is a good idea.

AMD might have stumbled on the best compromise between clock speed, per-clock performance, and heat management with the Athlon. Processors have several limitations. Engineers can’t just tack on however many transistors they want – this will lead to heat issues. Also, as the Pentium 4 has shown, engineers can’t just design a chip with a huge pipeline for maximum clock speed – not only will this reduce per-clock performance, but the processor will hit the MHz wall the P4 did. Finally, processors need to scale well. The AMD K5 was an excellent performer per clock that soundly beat a same-clock Pentium in integer performance, though it was lacking in floating point capability. The problem was, it couldn’t scale. The fastest K5 released reached a mere 116MHz for a “PR” (Pentium Rating) of 166, compared to the 233MHz the Pentium hit.

The Pentium III was generally as fast as the Athlon, clock for clock. It didn’t scale as well, and was thought to reach its peak at about 1.13GHz, but the Intel team in Israel kept the design alive and cleaned it up to make the Pentium M that has thus far reached about 2.26GHz. This compares well with the 2.4GHz of the Athlon 64 4000+. Intel has also included dual core on the upcoming desktop variant, and plenty of rumor supports the idea that it does indeed have 64-bit support.


On a technical basis, then, I have nothing to gripe about with the new Intel chip. Yet I am still unhappy with the company for two reasons:

One: Intel is still responding to AMD. This is not the dynamic company that launched new designs, features and ideas year after year a mere decade ago. Microsoft constantly re-invents and challenges itself to respond to all sorts of threats, it has gone on the offensive against Sony, IBM, and Google – three of the most dynamic major companies around and each competing in a different field. Intel is apparently content to let AMD come up with an idea, and then implement their slightly different (64-bit), maybe better (SSE), maybe worse (dual-core) version. Does that seem like a winning strategy to you, always coming from behind to try and tie, always playing defense?

Two: What’s the brand name of the new Intel chip, since Pentium is being dropped? “Core”. That’s right, Intel has made “Core” into a trademarked brand. One of the most common words used to describe a processor, and Intel hijacks it in a lame attempt at marketing exposure. That’s right, every time you ask for a “dual core” Athlon, Intel hopes that the resident idiot across the counter at the computer store will assume you mean “Intel Core Duo”. Perhaps they’re hoping to confuse search engines, or establish grounds for a lawsuit to keep AMD from advertising the Athlon X2 as “dual core”. I don’t know – quite honestly, the idiocy baffles me.

If the naming scheme is not the single most pathetic attempt at gaining an advantage over a competitor, I don’t know what is. Someone, anyone, please tell me what happened to the Intel of the mid-1990s. Back then, I was buying up every PC-related magazine at a prodigious rate for news and benchmarks of the Pentium and Pentium Pro. Not only had marketing convinced me they were going to be the Next Big Thing, but the reality was that they were the Next Big Thing. That Intel, the real Intel didn’t resort to desperate gimmicks like hijacking a common words for cheap attention – instead it pummeled its competition with better processors, newer features, and quality. Now? What’s Intel’s next great plan, to take Intel CEO Paul Otellini into a beauty salon, give him a makeover, dye his hair blond, dress him up in a revealing miniskirt and teach him how to bat his eyelashes suggestively?

Where’s the self-respect, Intel? Are you that desperate that you’re going to steal the word “core” from our lexicon? That’s pathetic, that’s something a company from the third world trying to compete here might do; that’s like naming your store “Wall-Mart” in the hopes that someone gets confused and walks in. Stop the gimmicks and come up with a real product with real innovation. Maybe “Core” is the Next Big Thing, and maybe not, but either way it deserves its own name.

Put some clothes on, Paul.