Gulftown Review (Intel's upcoming 32nm 6-core processor)

[VirtualRain]

http://pclab.pl/art39718.html

I wouldn't expect a price cut on the next rev of the Mac Pro... although you may end up getting 6 cores for the price of today's 4 cores providing a nice compromise for those torn between today's octo or quad...

Intel has not yet shown a silicon wafer with the Gulftown cores or the processor itself with the core uncovered. According to unofficial sources, its size should be approximately 21.1 × 11.4 mm. The core surface area is probably around 240 mm2, i.e. about 10% less than the Bloomfield core (263 mm 2). This means that if the yield on Intel's 32-nanometer process technology is as good as that on 45-nanometer technology, the manufacturing cost of a single Gulftown is less than that of a Bloomfield. But let us not count on Intel's six core price tag to be low. Firstly, Intel's throughput in the 45-nanometer process is much higher – only one plant (D1D in Oregon) is ready to produce 32 nm circuits. Another three are being converted. Meanwhile, 45-nm systems are produced in three facilities. In addition, the Gulftown prices will be tailored to their target consumers – prepare to spend about 1,000-1,200 dollars for the desktop version.

Clock-for-clock single threaded performance remains nearly identical to Bloomfield...

They had this to say about the multi-threaded performance (graph below)

In single-threaded mode, all Nehalem-based processors obtain similar results. In multithreaded mode, the 32-nanometer six-core has about 33% more performance than the eight-thread Core i7. Why not 50%? The Cinema 4D engine takes a long time to complete a thread and start a new one, which means that for part of the time, not all 12 threads are being used. As the number of threads grew, scaling would keep deteriorating.

Early BIOS issues affecting memory performance bunged the Photoshop test...

The only real good news is the lower operating temps and power consumption which may indicate the possibility of higher turbo boost clocks.

 

[smacman]

The only real good news is the lower operating temps and power consumption which may indicate the possibility of higher turbo boost clocks.

So not much good news for us then!?

 

[Pressure]

So not much good news for us then!?

If you render or encode video, this is great.

 

[gugucom]

This test is a joke. The CPU wasn't properly enabled by the firmware and as a result the memory controller was run at a fraction of it's capacity. Practically every high performance benchmark is affected by this. Even power and temperature will be different in the final version compared to this "test".

The main result from this review are: Improved encryption instructions and faster memory will be introduced. All the important messages have come from other sources.

For the Xeons Apple could rise memory speed by 40% from 1066 MHz to 1600 MHz. Combined with 50% more core capacity we can simply expect 50% higher benchmarks on all multi threaded apps and similar or lower energy consumption. Mono threaded apps will also have the potential to be slightly faster in comparison to the 2009 MP because the frequency range of 5600 is supposed to go higher than 5500.

High end CPU prices by Intel are going to be the same which means 1600$ for the top end. Low end CPU prices are supposed to be higher than the 3500/5500 range. If Apple follows that pricing the low end hexads will go up in price against the 2009 quads. The twelve core machines could keep the price level of the octads. This assumes that Apple will simply replace 3500/5500 with 3600/5600 which I think they will.

 

[gotzero]

This seems like great news to me, and the previous poster is correct that it looks like the tests were not "fair". Bring on the dirty dozens!

The 27" iMac is enough machine now for those whose primary concern is price, so I would rather have them focus on performance.

 

[VirtualRain]

I wouldn't say the test is a joke... but like most early CPU engineering sample reviews, results are (of course) not fully representative of the final product, however it's interesting to see early benchmarks to get a feel for what workloads might benefit the most from these new CPU's so buyers on the fence about buying a 2009 Mac Pro vs. waiting for the 2010 Gulftown version can make a somewhat informed decision.

BTW, here's another early review...

http://www.pcgameshardware.com/aid,...nds-on-test-of-the-six-core-Gulftown/Reviews/

Note that again, the 6 cores are not scaling as you might expect in Cinebench, which may be Cinebench related or issues with the early bios/memory/silicon.

Other noteworthy news here is that under load, the Gulftown with it's two extra cores and 2MB of L3, consumes less power than an equivalent clocked Bloomfield.

EDIT: The other possibly good news is that with engineering samples already making their rounds, it's very likely we could see production silicon in systems as early as a few months from now.