As we're approaching the rumored Intel Comet Lake-S desktop launch, we're slowly starting to spot more performance figures that fall in a much more believable range. The latest test submission comes from Tum_Apisak, who has kindly trawled through the Geekbench results databases and found some figures for the alleged i9-10900K.
The Intel Core i9-10900K is expected to be the successor to the Core i9-9900K, packing an additional two cores and featuring slightly uplifted frequencies. Whereas the Core i9-9900K has a base frequency of 3.6 GHz and a maximum boost clock of 5.0 GHz, the i9-10900K is expected to bump those figures to 3.7 GHz and a maximum clock of 5.1 GHz.
Purported Benchmark Results
The results found by Tum_Apisak place the i9-10900K at a single-core Geekbench score of 1437 points, and a Multi-Core score of 11390 points. For comparison, the Intel Core i9-9900K scores 1340 points in single-core and 8787 points in multi-core tests.
|Cores/Threads||8 / 16||10 / 20|
|Base Frequency||3.6 GHz||3.7 GHz|
|Max Boost Frequency||5.0 GHz||5.1 GHz|
|Geekbench 5 Single-Core||1340 pts||1437 pts|
|Row 4 - Cell 0||Improvement||+ 7%|
|Geekbench 5 Multi-Core||8787 pts||11390 pts|
|Row 6 - Cell 0||Improvement||+ 30%|
*Core i9-10900K specs in table not confirmed
Looking purely at core count, you would expect the Core i9-10900K to notch a 25% increase in multi-threaded performance, but the 30% figure jumps over that by a significant margin. The slightly elevated clock speeds would account for that somewhat, but not by the entire 5%, so we reckon there is something more going on.
It's been known for a while that Intel has been having problems with the power consumption of its new Comet Lake Desktop chips, as these are still to be on the 14nm fabrication process, and packing this many cores into a CPU at that node size makes for some very power-hungry silicon. Nevertheless, the spotted results suggest Intel has been able to tame the power consumption somewhat, thus enabling higher sustained boost frequencies and the higher performance that comes with that.
Also note that it's highly unlikely that the two chips were tested on the same motherboard with identical cooling, which can also account for differences. The i9-10900K was benchmarked in the ASRock Z490M Pro4 together with 32GB of DDR4 memory.
Of course, we won't know for sure until the chips come out and we get them in our labs for review. The biggest bit of information that we're still missing is what they will cost -- though given that AMD's Ryzen 3000 CPUs are brutally competitive and Intel does need to catch up, we somehow doubt the price hikes will be all too steep. Even so, Intel says it might take until late 2021 for Intel to catch up with AMD.
What did I miss?
Barring some trickery, many might be pretty underwhelmed with a 3-5 fps increase in gaming framerates at 1080P in CS:GO! :)
Upping the frequency by just 2% shouldn't do that!
One would think with all the money they've made sitting on their laurels for so many years and not actually making much in the way of advances in CPU technology, that they could easily afford providing exceptional coolers for their excessively power-hungry CPUs.
Since when has 5% become a significant margin?
Is this something that is new in our now post Corona world?