Author: batteryforpc

We’re just weeks away from the launch of AMD’s RDNA 2 GPUs, but a firmware listing may have prematurely outed the specs for the the company’s highly-anticipated Big Navi cards.

An eagle-eyed Redditor found a listing in the new ROCm (Radeon Open Compute) firmware that reveals some of the specs for the so-called Sienna Cichlid GPU – otherwise known Navi 21 or Big Navi – which will arrive as the flagship of the RX 6000 series

The listing hints that the GPU will feature 80 compute units (CUs) and a 256-bit memory bus. If each Compute Unit (CU) in AMD’s RDNA 2 architecture still equates to 64 Stream Processors (SPs), that means Big Navi will 5,120 SPs. 

If this is true, and if the GPU runs on TSMC’s latest 7nm process, the flagship RDNA 2 processor could deliver Nvidia RTX 3080 levels of performance. After all, AMD has already promised that RDNA 2 will deliver 50% more performance per watt than its first generation RDNA architecture. 

What’s more, a recent rumor suggests the GPU will feature 16GB VRAM – more than the 10GB Nvidia has loaded onto its Ampere GPU. 

The firmware update also references a graphics card codenamed Navy Flounder, believed to be Navi 22 or Navi 23. This mid-range GPU features 40 CUs and a 192-bit memory bus, according to the listing, which amounts to 2,560 SP.

This card will likely arrive as a replacement for the AMD RX 5700 XT, but with its newer RDNA 2 architecture, it will support real time ray tracing and be capable of gaming in 4K. We don’t have any details about performance just yet, but it could give Nvidia’s RTX 2070 a run for its money. 

It won’t be long until we find out for sure, as AMD will announce these new GPUs under the RX 6000 series branding on October 28. 

A new series of Intel’s Next Unit of Computing (NUC) has been detailed. 

The next generation Tiger Lake NUC has been detailed

Intel is adding a new series of NUCs called 11 PRO (Tiger Canyon). The PRO series is meant for business and commercial use. These are not the same series as Panther Canyon (NUC 11 Performance), also featuring Tiger Lake, as those are meant for consumers.

Intel Tiger Canyon aka NUC 11 PRO will feature up to 11th Gen Core i7-1185G7 processor based on Tiger Lake-U architecture. This is a quad-core CPU with eight threads offering up to 4.8 GHz turbo clock at 28W. The NUC 11 PRO would also be offered with lower-tier Core processors going down to Core i3-1115G4, which is a dual-core processor.

The NUC 11 PRO would offer dual HDMI 2.1 display connectors (supported by integrated Xe-LP Graphics) and dual Thunderbolt 4 connectivity. Since these devices are often used for home theatre media boxes, such options will certainly not go unnoticed. In businesses, this might mean connecting multiple high-resolution screens for presentations.

Intel’s new MiniPC supports two M.2 slots for (key M for PCIe NVME SSD and key B for other peripherals). It supports WiFi 6 and memory up to 64GB DDR4-3200. The device does not offer discrete graphics options, but one would guess that these might come next year when Intel’s DG1 graphics become available.

The device will be offered in two sizes: ‘tall’ H chassis (117x112x54mm) and ‘slim’ K chassis (117x112x37mm), both were pictured on a leaked slide.

The Samsung Galaxy S20 Fan Edition is a more affordable version of the flagship phone ‘for the fans,’ which means it prioritizes cameras and display quality at a more favorable price than the rest of the Galaxy S20 line, starting at $699 / £699 / AU$1,149 for the sub-6 5G model.

The phone goes as low as £599 / AU$999 for the 4G-only version, which isn’t coming to the US, though a different mmWave 5G $749 version is coming to the US through Verizon. All S20 Fan Edition models can be pre-ordered starting September 23, while the S20 Fan Edition release date is set for October 2, except in Australia where pre-orders start September 25 and the release date is October 9.

For those prices, the S20 Fan Edition packs most of the best perks of the flagship S20 line: a Snapdragon 865 chipset (or Exynos 990 for the 4G version), 6GB of RAM, and 128GB of storage (expandable to 1TB via microSD). The phone has a 4,500mAh battery, maxing out at 25W wired fast charging and 15W wireless charging. 

A few compromises for lower cost

The S20 Fan Edition makes some compromises compared to its pricier siblings, though it has some advantages, too. Its 6.5-inch OLED display only has Full HD Plus resolution (2400 x 1080) instead of the WQHD Plus of the main S20 line, but the S20 Fan Edition retains a 120Hz refresh rate. The phone also has a polycarbonate rather than glass back, but it comes in six colors (navy, red, lavender, mint, white, orange).

Despite releasing in October 2020, the phone comes with Android 10; an expected upgrade to the recently-released Android 11 will be one of a guaranteed three major operating system updates coming in the next few years.

Ultimately, Samsung believes the S20 Fan Edition makes the right cuts and comes in at the right price to appeal to budget-conscious shoppers who still want good specs and features. It’s this phone generation’s version of the Samsung S10e or S10 Lite to tide buyers over until the 2021 slate of Samsung phones.

We were fairly impressed with Adata’s first foray into gaming PCs, which began a few months ago with the XPG Xenia 15, a genuinely impressive laptop with better-than-expected battery life. For its followup act, Adata is trying its hand at a mini PC with its new XPG Gaia.

The Gaia is based on Intel’s NUC 9 Extreme Kit, or otherwise known as Ghost Canyon. That is not necessarily a bad thing, just perhaps a bit uninspired for Adata’s introductory desktop.

Built for “gamers, esports pros, and tech enthusiasts,” Adata appears to have simply taken the Ghost Canyon NUC and stuffed its own brand memory and storage inside, then slapped an XPG logo on the front. The result is a compact 5-liter PC that should be relatively fast out of the box, though gamers, esports pros, and tech enthusiasts will undoubtedly want to add a discrete GPU.

The system is built around an Core i9 9980HK processor, a relatively meaty 8-core/16-thread chip with a 2.4GHz base clock, 5GHz boost clock, and 16MB of L3 cache. This also provides the graphics, by way of Intel’s UHD 630 GPU (24 execution units, 350MHz to 1,25GHz clockspeed). That’s fine for playing less demanding games like League of Legends, but is not going to cut it for more graphically intensive titles (Notebookcheck has a nice rundown of benchmarks).

One of the neat things about the Ghost Canyon NUC is that it can fit a discrete graphics card up to 202mm (7.95 inches) in length. There are even graphics cards designed specifically with the Ghost Canyon NUC in mind, like the Asus Dual GeForce RTX 2070 Mini OC Edition.

As far as what comes with the Gaia, in addition to the 9980HK CPU, Adata ships it with 32GB (2x16GB) of DDR4-3200 “sorted” RAM (undoubtedly from its own stock) and an XPG SX8200 Pro SSD, a fast storage drive rated to deliver read and write speeds of up to 3,500MB/s and 3,000MB/s, respectively. Adata does not mention the capacity, so it’s likely there will be multiple SKUs with different storage allotments.

The system also boasts Wi-Fi 6 (802.11ax) and Thunderbolt 3 connectivity, and is powered by a 500W PSU that is 80 Plus Platinum certified.

A sharp-eyed redditor found a listing in the new ROCm (Radeon open compute) firmware update that reveals some of the specs for the highly-anticipated RDNA2 Sienna Cichlid (Big Navi) and Navy Flounder (Navi 22 or 23) graphics cards. Things could change over the next month as AMD fine-tunes its gear, so take these specs with a grain of salt – even though the numbers come from an official firmware update, these specs aren’t confirmed.

The firmware update indicates Sienna Cichlid (Navi 21, i.e., Big Navi) will feature 80 CUs and a 256-bit memory bus. The CU count for this GPU is interesting, indicating all that we know on Big Navi could be legit, with most signs pointing to 80 CUs (or around 80 CUs) to be the spec for Big Navi. If this is true, not to mention if the GPU runs on TSMC’s latest 7nm process, we could see RTX 3080-levels of performance, but with much better efficiency.

The firmware specifications also list AMD’s Navy Flounder with 40 CUs and a 192-bit memory bus. This GPU appears to a direct replacement to the RX 5700 XT and/or RX 5700 on the new RDNA2 architecture. If each CU has the same 64 stream processors as RDNA1, then Navy Flounder will have identical core counts to the 5700 XT.

Strangely, the memory bus is narrower than Navi 10 chips (RX 5700/XT) at 192-bits, AMD could be pulling an Ampere tactic by using higher-frequency GDDR6X memory to compensate for the lower bus width. Unfortunately, there is not enough information to indicate just where these GPUs will be placed in AMD’s lineup – they could compete with the rumored RTX 3060 Ti, possibly the RTX 3070, or even other SKUs in the future.

Hopefully, this time around, AMD can make a bigger dent in Nvidia’s market share over the upcoming months and years. AMD will announce these new GPUS under the RX 6000 series branding on October 28th.