Saturday, January 7th 2023
AMD Ryzen 7040 Series "Phoenix Point" Mobile Processor I/O Detailed: Lacks PCIe Gen 5
The online datasheets of some of the first AMD Ryzen 7040 series "Phoenix Point" mobile processors went live, detailing the processor's I/O feature-set. We learn that AMD has decided to give PCI-Express Gen 5 a skip with this silicon, at least in its mobile avatar. The Ryzen 7040 SoC puts out a total of 20 PCI-Express Gen 4 lanes, all of which are "usable" (i.e. don't count 4 lanes toward chipset-bus). This would mean that the silicon has a full PCI-Express 4.0 x16 interface for discrete graphics, and a PCI-Express 4.0 x4 link for a CPU-attached M.2 NVMe slot; unlike the "Raphael" desktop MCM and the "Dragon Range" mobile MCM, whose client I/O dies put out a total of 28 Gen 5 lanes (24 usable, with x16 PEG + two x4 toward CPU-attached M.2 slots).
Another interesting aspect about "Phoenix Point" is its memory controllers. The SoC features a dual-channel (four sub-channel) DDR5 memory interface, besides support for LPDDR5 and LPDDR5x. DDR5-5600 and LPDDR5-7600 are the native speeds supported. What's really interesting is the maximum amount of memory supported, which stands at 256 GB—double that of "Raphael" and "Dragon Range," which top out at 128 GB. This bodes well for the eventual Socket AM5 APUs AMD will design based on the "Phoenix Point" silicon. Older Ryzen 5000G "Cezanne" desktop APUs are known for superior memory overclocking capabilities to 5000X "Vermeer," with the monolithic nature of the silicon favoring latencies. Something similar could be expected from "Phoenix Point."The iGPU of the Ryzen 7040 series in its top avatar will have the branding "Radeon 780M," an upgrade from the "Radeon 680M" of the top iGPU option available with the "Rembrandt" silicon and its RDNA2-based iGPU. The new 780M is based on the latest RDNA3 graphics architecture, and packs 12 compute units (768 stream processors), with the same dual-instruction issue rate capabilities as the desktop Radeon RX 7900 series GPUs; and matrix-math accelerators (these are besides the dedicated XDNA AI accelerator present on the "Phoenix Point" silicon). The iGPU has engine clocks as high as 2.90 GHz.
The iGPU of "Phoenix Point" is confirmed to feature AMD's latest Radiance Display Engine, with support for DisplayPort 2.1 UHBR10 and HDMI 2.1, with native support for 8K 60 Hz displays with a single cable. It also features the latest VCN media engine, with hardware-accelerated AV1 encoding up to 4K @ 240 Hz 10 bpc, and 4320p @ 175 Hz 8 bpc H.265; and hardware-accelerated decoding of nearly all standard resolutions/bit-depth/framerates of MPEG2, VC1, VP9, H.264, H.265, and AV1.Built on the 4 nm EUV foundry node at TSMC, the "Phoenix Point" monolithic silicon has a die-area of 178 mm², and a transistor-count of 25 billion. Besides the iGPU, it features a single 8-core "Zen 4" CCX. Each of the 8 CPU cores has 1 MB of dedicated L2 cache, and share 32 MB of L3 cache.Many Thanks to TumbleGeorge for the tip!
Source:
AMD
Another interesting aspect about "Phoenix Point" is its memory controllers. The SoC features a dual-channel (four sub-channel) DDR5 memory interface, besides support for LPDDR5 and LPDDR5x. DDR5-5600 and LPDDR5-7600 are the native speeds supported. What's really interesting is the maximum amount of memory supported, which stands at 256 GB—double that of "Raphael" and "Dragon Range," which top out at 128 GB. This bodes well for the eventual Socket AM5 APUs AMD will design based on the "Phoenix Point" silicon. Older Ryzen 5000G "Cezanne" desktop APUs are known for superior memory overclocking capabilities to 5000X "Vermeer," with the monolithic nature of the silicon favoring latencies. Something similar could be expected from "Phoenix Point."The iGPU of the Ryzen 7040 series in its top avatar will have the branding "Radeon 780M," an upgrade from the "Radeon 680M" of the top iGPU option available with the "Rembrandt" silicon and its RDNA2-based iGPU. The new 780M is based on the latest RDNA3 graphics architecture, and packs 12 compute units (768 stream processors), with the same dual-instruction issue rate capabilities as the desktop Radeon RX 7900 series GPUs; and matrix-math accelerators (these are besides the dedicated XDNA AI accelerator present on the "Phoenix Point" silicon). The iGPU has engine clocks as high as 2.90 GHz.
The iGPU of "Phoenix Point" is confirmed to feature AMD's latest Radiance Display Engine, with support for DisplayPort 2.1 UHBR10 and HDMI 2.1, with native support for 8K 60 Hz displays with a single cable. It also features the latest VCN media engine, with hardware-accelerated AV1 encoding up to 4K @ 240 Hz 10 bpc, and 4320p @ 175 Hz 8 bpc H.265; and hardware-accelerated decoding of nearly all standard resolutions/bit-depth/framerates of MPEG2, VC1, VP9, H.264, H.265, and AV1.Built on the 4 nm EUV foundry node at TSMC, the "Phoenix Point" monolithic silicon has a die-area of 178 mm², and a transistor-count of 25 billion. Besides the iGPU, it features a single 8-core "Zen 4" CCX. Each of the 8 CPU cores has 1 MB of dedicated L2 cache, and share 32 MB of L3 cache.Many Thanks to TumbleGeorge for the tip!
83 Comments on AMD Ryzen 7040 Series "Phoenix Point" Mobile Processor I/O Detailed: Lacks PCIe Gen 5
Hope I can get something decent for like $599. Doubt I get that lucky, but hey I can dream.
Here's to hoping this is the year stronger just CPU options become popular again as unlikely as it is.
This APU is actually exactly what you are looking for if I am reading you correctly.
I personally am going to keep my eye out for it, hopefully it comes out in next couple months. Hopefully in a Dell laptop, cause my company gives me discount through Dell.
There is one model at Bestbuy. The $1000 Asus VivoBook S 14X OLED.
The 6800U was a great performer but the typical 25W TDP throttled away a lot of its potential. The same silicon given a 45W TDP as the 6800H was never spotted outside of chungus gaming laptops with stupid RGBLED and childish robot slashes/angles/vajazzle adorning their cheap plastic shells.
The "default" (non U-series) Intel consumer laptop without dGPU has been 35-47W going back at least as far as Skylake. Across that many years, manufacturers and model variants, there must easily be 10,000 different models sold with 35W+ TDPs and no dGPU. Why is it so hard for AMD to do that? The whole point of an APU is that they have the best IGP, so that you can get away without needing a hungry dGPU. They've utterly failed to capitalise on that USP. No, these APUs, 'codename Phoenix' are going to be the U-series, 7800U etc. They will be cut-down silicon with a maximum of 8 cores and low cTDPs of likely 15~28W again.
Both @trsttte and I are talking about the H-series, 'codename Dragon Range' which will be 35W+ 16-core APUs that will likely come with the default, non-gaming RDNA2 IGP that Zen4 desktop CPUs have. They're definitely not 3D powerhouses, but that basic 2CU RDNA2 IGP is plenty for most non-gaming applications and even works for most CAD software. As a mobile editing rig or workstation, 16 cores in a thin-and-light is amazing, and AMD simply haven't had any design wins for a performance-focused thin-and-light to date.
geizhals.eu/?cat=nb&xf=14265_H~14265_HS~14265_HX~18366_16+-+Rembrandt~19538_16+-+Zen+3++(ab+2022)&asuch=&bpmin=&bpmax=&v=e&hloc=at&hloc=de&hloc=pl&hloc=uk&hloc=eu&plz=&dist=&mail=&sort=p&bl1_id=30
I saw at least one 6600HS model at Lenovo's US site.
It is however impossible to find a Rembrandt laptop under
€900€899. Given that some models still are listed as "coming soon", things must be going really slow.I have no idea what TSMC's 6 nm chips are used for, except for one Radeon model. Maybe Raphael IO chips?
That's why you can find Ryzen 6000 models listed as "coming soon" right now.
Chip shortage affected everyone, not just TSMC.
I wonder what went wrong with 6000?
Hope AMD pushes harder in laptops now that the discrete GPU market is lost for them.
I wonder what the performance & price delta will be between them.
So, I hope this graph is an indication of new AMD laptops in the works from OEMs and not just a glitch, which is normal at the first weeks of every new quarter in PassMark's market share graphs