player-graphics-card:-lenovo-confirms-geforce-rtx-3060-with-12-gb-of-memory

Player graphics card: Lenovo confirms GeForce RTX 3060 with 12 GB of memory

Lenovo has prematurely released the data sheet for the upcoming Legion R5 complete system 28 IMB 05 put online. The PC manufacturer confirms the existence of other mid-range graphics cards from Nvidia’s current Ampere family: GeForce RTX 3050, GeForce RTX 3050 Ti and GeForce RTX 3060.

The latter model is said to be in response to AMD’s expected Radeon series RX 6700 with 12 GByte graphics memory also with 12 GByte GDDR6-RAM on 192 Data lines appear – Lenovo also specifies the amount in the data sheet. The GeForce RTX 3060 would therefore have more graphics memory than its big sister RTX 3060 Ti, RTX 3070 and RTX 3080.

This is unusual, but it would not be the first case of this kind: The GTX 780 and GTX 770 were available as new editions with 6 and 4 GB of RAM, the GTX 780 Ti, however, only with 3 GByte. The rumor mill spoke for a long time about new versions of the RTX 3080 and RTX 3070 with 20 or 16 GB of memory, which Nvidia is said to have deleted. A GeForce RTX 3080 Ti with 20 GByte should appear at the beginning 2021.

GeForce RTX 3050 as ray tracing entry According to the often well-informed Twitter user “kopite7kimi” should the GeForce RTX 3060 come with 3840 shader cores – 21 Percent less than the GeForce RTX 3060 Ti. Below the GeForce RTX 3060 Lenovo also names the two models GeForce RTX 3050 Ti and GeForce RTX 3050. The Ti version should be hardly slower with 3584 shader cores than the GeForce RTX 3060, but according to Lenovo only uses 6 GB of GDDR6-RAM.

The GeForce RTX 3050 is According to “kopite7kimi” with 2304 shader processing cores and 4 GByte GDDR6 memory clearly lagged behind. For comparison: The cheapest RTX graphics card from the 1920 generation, the GeForce RTX 2060, used 1920 shader, but with higher computing power per core.

(mma)

rocket-lake-engineering-samples-benchmarked-against-zen-3

Rocket Lake Engineering Samples Benchmarked Against Zen 3

(Image credit: Chiphell.com)

As outlined in a recent post on Chip Hell, one of its users reportedly grabbed an early B560 motherboard and engineering samples of three of Intel’s new Rocket Lake CPUs, including the Core i7-11700, Core i9-11900, and Core i9-11900K. 

The tested pitched each processor against AMD’s Zen 3-powered Ryzen 7 5800X to see just how they compare to AMD’s best eight-core chip. Since these are engineering samples, the Intel chips’ clocks speeds are significantly lower than we would likely see with retail models. The poster also threw in Intel’s previous-gen Core i9-9900K and Core i7-10700K as well to compare gen-on-gen performance gains.

The testbed used the same B560 board discussed above, a B550 Taichi Razer Edition for the AMD tests, ASRock Radeon RX 6800 Taichi, 2x8GB kit of ZADAK Spark DDR4-3600 RAM, 1000W Antec HCG-X1000 power supply, and a 360mm AIO liquid cooler.

Here are the Rocket Lake engineering samples tested:  

  • QV1J, Core i7-11700 ES — 1.8GHz base frequency, 4.4GHz boost frequency.
  • QVTE, Core i9-11900 ES — 1.8GHz base frequency, 4.5GHz boost frequency.
  • QV1K, Core i9-11900K ES — 3.4GHz base frequency, 4.8GHz boost frequency.

Even though there is a wide selection of benchmarks posted at Chip Hell, we’re only covering the locked 4GHz benchmark results. We chose to focus on this test because the engineering samples for Rocket Lake are clocked so low that any performance benchmarks from these samples are specific to these samples alone, and will not represent actual Rocket Lake performance when the retail SKUs hit shelves this year.

The “Locked at 4GHz” Benchmark

4GHz Locked R15 Benchmark (Image credit: Chiphell.com)

Chip Hell ran the Core i9-9900K, Core i7-10700K, Core i9-11900K ES, and Ryzen 7 5800X in Cinebench, but with all the chips locked at 4GHz, allowing us to see how much of an IPC gain Rocket Lake-S has purely from an architectural standpoint, as clock speed is no longer the deciding factor to performance.

Processor Cinebench R15 Single-Threaded Cinebench R15 Multi-Threaded
Ryzen 7 5800X 221 1,121
Core i9-11900K ES 200 1,029
Core i7-10700K 176 888
Core i9-9900K 168 852

We can see the gains from Comet Lake to Rocket Lake are quite decent; Rocket Lake commands a 13% lead over its predecessor. Compare this to the generational leap from Coffee Lake to Comet Lake at just 4%.

However, despite the architectural gains, it’s not enough for Intel to beat AMD’s Ryzen 7 5800X, which wins by 8%.

It appears that Intel will use clock speed rather heavily to try to gain an advantage over AMD’s Zen 3 architecture, as AMD still appears to win on clock-for-clock performance. But at least the architectural changes were significant enough to give Rocket Lake a decent IPC increase over Comet Lake. The overall performance gap should also widen rather significantly if clock speeds are higher on shipping Rocket Lake models.

However, Intel will continue to struggle against AMD’s Zen 3 platform, whether or not Intel manages to beat AMD in the single-threaded battle, as Rocket Lake will be severely behind in core count and that won’t change until Intel’s 12th-Gen Alder Lake architecture arrives.