Friday, July 27, 2007

INTEL guilty

EU found Intel guilty.

Some Intelers here say Intel is innocent until proven guilty. But it is the other way around in this case. EU has already done years of investigation and basically nailed Intel. It's not like Intel will get a jury. All Intel will get is a judge and summary judgment. They do that in U.S. too, no genuine issue of material fact and Intel is found guilty as a matter of law. Those EU charges are hard facts backed up by hard evidence-- and there is no jury, just a judge applying the rules and checking sentencing guidelines.

AMD's K10 running at 3GHZ.

Thursday, July 19, 2007

AMD grows units by 38% in 2Q07

AMD result is out. AMDers should cheer about the massive growth. It seems ATI is dragging a bit, but it's a temporary pain AMD must endure.

AMD's market share was 18.7% in 1Q07, 38% sequential unit growth should put AMD at about 26% . In other words, Intel market share is about 70%.

All AMD needs to do is keep pumping out those $70 dual core CPUs.

One interesting thing is AMD gained in server ASP. This is because Intel is now purely a low end 2P server chip vendor. AMD must be selling a lot of 8-way Opterons. With quad-core Opterons soon available, Intel's server performance will be pitiful. AMD will have 8P quad-core servers, total 32 K10 cores. Intel, on the other hand, is limited to 2P Clovertown, total 8 cores.

Intel's server performance will be less than 20% of AMD's.

Tuesday, July 17, 2007

Intel Core 2 servers are literally hot

Even Anand had to admit Intel servers are burning lots of watts. A 2.3GHZ Woodcrest server consumes 260 watts at idle. A 2.6GHZ Opteron server only consumes 160 watts. Intel dual core servers consume 60% more power than AMD. In benchmarks, we see the same thing, a 2.3GHZ Woodcrest generates more heat than a 2.6GHZ Opteron made on 90nm.

I told you long time ago. Core 2 architecture is a scam. You can see from the benchmarks, Intel has no advantage in performance, it only generates more heat.

Sunday, July 15, 2007

Core 2 Duo thermal throttles

Remember the old days when Intel rated Netburst CPU at 3.73GHZ but in reality it often runs at half the speed due the thermal throttling? Same is seemingly happening with Core 2 Duo within normal operating parameters. Basically, you get a Core 2 Duo rated at 2.93GHZ, when it is not doing anything stressful and the room temperature is cold, it runs merry at that speed, most of the circuits are turned off anyway. However, once you add some load, the Intel chip slows down to save itself from meltdown.

BTW, EU may fine Intel $3.5 billion for breaking the law.

The only thing that is keeping Intel alive

Intel is still alive for one single reason: AMD's complex 65nm SOI process is not producing the clockspeeds as expected.

As we have seen, although its 65nm SOI is low power, AMD is unable to push clockspeed higher. The X2 6000+ (3GHZ) is manufactured on 90nm process. The highest performing 65nm AMD processor is clocked at 2.6GHZ.

Once AMD optimize its 65nm process, Intel will be dead in less than three months.

PS: One reader made some reasonably articulated comments on AMD's strategic moves such as K10 design and ATI acquisition -- a rare thing from an Inteler. However, as an Inteler, he could only look forward seven inches. The x86 market is completely different today with AMD being the industry standard setter in server, desktop and mobile. Whatever Intel does, it is just copying AMD's ideas. The only thing that is holding AMD back is its apparent difficulty with 65nm SOI process. In terms of architecture, AMD is years ahead of Intel. Although Intel has copied AMD64 instruction set, NX bit and emulated multi-core, Intel's architecture is stuck with the primitive FSB design. Once AMD rolls out Direct Architecture 2.0 on fast clocks, Intel will look retarded. There are rumors that Intel will copy more from AMD, such as distributed memory architecture, direct connect and embedded memory controllers. But those are still on the paper.

AMD got everything in place to finish Intel. The only thing it needs to do now is to get 65nm SOI to work faster.

Sunday, July 08, 2007

Intel out of top 5

According to the June 2007 top 500 supercomputer list, Intel is nowhere to be found in the top 5. The fastest Intel powered super computer is ranked #8 with Clovertown. AMD, on the other hand, took #2 and #3 spots with Opteron.

With the death of Itanium and bankruptcy of SGI, Intel simply does not have a viable solution for high performance computing. The 1970s FSB architecture Intel still uses is simply too primitive -- the FSB simply chokes under load.

Once K10 is out, Intel will be totally driven out of the high performance market.

Wednesday, July 04, 2007

Core 2 Quad price starts the free fall

I previously projected that once K10 is out, all Intel Core 2 Quad, Core 2 and AMD K8 chips will be sold in a narrow price range of $60 to $120. Now, it's about two months before the arrival of K10, and you can get a Core 2 Quad 6600 + MB at Frys for a total of $300.

Cheap? Remember, an Intel quad core is not true quad core: you get 2.5-2.8x the performance in multi-tasking, instead of 4x.

Tuesday, July 03, 2007

Patty, Patty, where are you?

The K10 is here to frag Clovertown. The story was a 2.3GHZ K10 is 50% faster than a 2.6GHZ Clovertown. In other words, a 4GHZ Clovertown is needed to compete against a 2.3GHZ K10, and a 3.5GHZ clovertown is needed to compete against a 2GHZ K10.

You will see Paul and Graig unloading their options soon.