Intel Coffee Lake Hits SiSoft Sandra Benchmark Database With 6 Cores And 12 Threads, 4.2GHz Turbo Boost

We’re just getting our heads wrapped about Kaby Lake-X and Skylake-X (aka, the Core X-Series), and here we are staring at Coffee Lake. Coffee Lake represents the fourth and latest major iteration of Intel’s 14nm process, which started with Broadwell in 2014. Coffee Lake processors are set to debut in 2018, and will offer up to six cores for mainstream PCs.

Today, we’re witnessing a supposed Coffee Lake-S chip in action via a SiSoft Sandra benchmark. The processor is identified as a Genuine Intel CPU 0000, which denotes an engineering sample. Other specifics regarding the processor identify it as a Kaby Lake-S part (which is expected since it is not fully supported in software yet). What’s interesting, however, is that rather than having four cores like most Kaby Lake and Kaby Lake-S processors, this processor has six cores and 12 threads, which confirms what we’ve previously heard about Coffee Lake processors for the mainstream segment.

Intel manufacturing D1X 1

This particular engineering sample features a base clock of 3.1GHz and a turbo clock of 4.2GHz. 256KB of L2 cache is provided per core and there’s 9MB of L3 cache shared between the cores.

Coffee Lake SANDRA Listing

Earlier reports suggest that there will be quad- and hexa-core versions of Coffee Lake. The quad-core parts will reportedly have a die size of 126 mm2, while the hexa-core variants will some in a bit larger at 149 mm2. LGA 1151 has served us well with both Skylake and Kaby Lake processors, but it’s likely that Intel will move to yet another socket/platform with Coffee Lake.

But what about Cannon Lake? We know that it can get a bit confusing with all of these “Lakes” being splashed around, but Cannon Lake is still on track to ship by year’s end. Cannon Lake will be Intel’s first 10nm processor family, and will initially be available for low-power applications. So, you can expect the first chips to be in the U-Series (15W TDP) and Y-Series (5W TDP) families.

Intel recently bragged that not only is Cannon Lake on track, but that its second generation 10nm architecture, Ice Lake, has been taped out:

Intel’s Cannon Lake processors will reportedly deliver a 25 percent performance boost compared to Kaby Lake, while offering a 45 percent reduction in power consumption. Intel’s 10++ process, which will be deployed with Ice Lake, will offer another 15 percent performance boost over Cannon Lake while adding in a further 30 percent reduction in power consumption.

Intel 10nm Die Area Scaling

AMD is currently strutting its stuff with the 14nm Zen architecture, which is already making waves the Ryzen family of processors. The company is also looking to shake up the HEDT market with Ryzen Threadripper and the server market with EPYC. Intel is making it clear, however, that it will use its lead in process technology to fend off all challengers going forward. Earlier this year at Intel’s Technology and Manufacturing Day, the company proclaimed that its first generation 10nm process technology is already a full generation ahead of the “other guys”.

Intel EVP Stacy Smith added, “Today we have line of sight to 7 and 5 nm. We may not know exactly which approaches will prove best for 5 nm yet, but our culture thrives on those challenges. It has for generations.”