AMD's 6th Generation Carrizo APUs Officially Launched and Detailed - Upto 15% IPC, 3rd Gen GCN Cores, DirectX 12 and HSA 1.0 Support

AMD has officially launched their latest Carrizo APUs at Computex 2022 and detailed additional $.25 and pieces nigh their next generation mobility processors. The AMD Carrizo mobility processors come up in two packages, the power optimized 15W variants and the more than performance oriented 35W variants which volition be shipping in several notebooks later this month.

AMD's sixth Generation Carrizo APUs Officially Launched and Detailed

The processors were detailed in a printing conference by Tinhte.vn and while the NDA lifts on 2nd June, we tin can now tell you lot what can y'all expect in terms of performance from the latest sixth generation processors. When talking about basic features, first of all, we should know that Carrizo is based on the 28nm process node and comes in the FP4 package. The Carrizo fries feature 4 x86 Excavator cores with ii MB L2 cache, 3rd generation GCN GPU (integrated) that pack 8 graphics compute units or 512 stream processors and 2 RBs. The chips support DDR3 dual channel memory with speeds of up to 2133 MHz and are designed to feature total back up for HSA 1.0 spec. The chips also integrate the southbridge on die and have several I/O technologies along with new software tier back up that we will detail in but a bit. The Carrizo APUs will be branded equally the 6th generation AMD FX, AMD A10 and AMD A8 series fries and we already know some product names such as:

AMD FX-8800P
AMD PRO FX-8800B
AMD A10-8700P
AMD A8-8600P
AMD A6-8500P
AMD PRO A10-8700B
AMD PRO A8-8600B
AMD PRO A6-8500B
AMD RX-418GD
AMD RX-216GD

Then at that place's the round-up of what to wait from Carrizo APUs merely let's get more than technical. From the ISSCC 2022 presentation, we know that Carrizo features a nominal 5-fifteen% IPC gains from the new Excavator cores which shows AMD is following Intel footsteps in this field with the blue team as well offering a similar IPC improvement on their latest 14nm Broadwell Uarch. The dice is still based on a 28nm node even so AMD has managed to optimize the overall chip design past adding 29% more transistors than Kaveri thanks to the high-density design library. This results in a 3.1 Billion transistor die that delivers forty% lesser power consumption and 23% lesser dice area than its predecessor. The H.265 encode support allows 3.v times transcode performance of Kaveri while the compute architecture enables the eight GCN compute units (512 stream processors) a reduction of 20% in power consumption.

When specifically talking about Excavator cores, we get improved and larger enshroud sizes that permit prefetch improvements and lower latency. Amend branch prediction leads to l% increase in co-operative target buffer size (512 to 768 Entry)) and accelerated flush in the FPU. New instruction support include AVX2, MOVBE, SMEP and BMI1/two along with more power gating options to cut down ability when the bit remains dormant or doesn't gets utilized to full extent. The nigh significant gains in frequency come to 15W models while the 35W models really able to push IPC with and 0-five% clock speed bumps. The 15W variants get a 25-45% frequency push and increase in IPC by x%.

In terms of size, the Carrizo die measures at 244.62mm² on the 28nm node while Kaveri measures at 245mm² on the same process. The divergence between both chips is that Carrizo ups the transistor count to 3.1 billion from Kaveri's 2.41 billion count. The sudden reduction in the size of the die even when adding more better x86 operation was due to the fact that Excavator cores are smaller than Steamroller cores, measuring at but fourteen.48mm2 with a cadre transistor count of 102 meg transistors. The L1 cache has besides doubled on Carrizo to 32 KB per core from 16 KB. The overall core structure has 690 million transistors crammed in i partition while the rest of the transistors are dedicated to GCN cores that utilize HSA and compute engine advantage in general purpose computing environments.

AMD is also giving a boost update to the GCN architecture with their 3rd generation GCN cores integrated inside Carrizo. These are the aforementioned architectural enhancements as featured on Tonga and the presently to be released Fiji graphics card. The iGPU has 512 KB L2 cache, 819 GFlops of compute performance and HSA acceleration via ATC. Some features such equally DirectX 12 (Level 12), improved tessellation functioning, loss less delta colour compression, updated ISA instruction set, loftier quality scaler unit, cache coherent fabric interface are available on the new GCN unit of measurement.

The nigh interesting thing most Carrizo, aside from its technical specifications is also the design of the chip itself. AMD for the get-go fourth dimension is aiming for a true SOC design eliminating the need of a separate FCH every bit was the case with Kaveri mobile which requires Bolton FCH for additional connectivity options. The FCH will exist integrated on the dice itself which will deliver Security, Brandish, Audio, PCI-e, SATA, SD, USB, Multimedia, UART/12C. CLCKGen and Misc I/O connectivity. AMD is aiming for UVD6, VCE3 and a audio co-processors with H.264 encode while feature a display control engine "DCE11″. With HDMI two.0 that provides upward to 3 display interfaces and PCI-e Gen three.0 x8 for discrete GPU expansion and PCI-e 3.0 x4 for GPP, the APU begins to look like a decent improvement over Kaveri from a design perspective. The FCH tin evangelize 4 USB 3.0 / ii.0 ports, 4 USB 2.0 ports and 2 SATA 3 ports while the memory controller permit for Dual Aqueduct DDR3 memory rated at 2133 MHz in SoDIMM grade gene (One per channel).

AMD Carrizo APU Comparison Chart:

	AMD Trinity APU	AMD Richland APU	AMD Kaveri APU	AMD Carrizo-Fifty	AMD Carrizo APU
Core	x86 Piledriver	x86 Piledriver	x86 Steamroller	x86 PUMA+	x86 Excavator
Cores	2-4	2-iv	2-4	2-iv	ii-4
GPU	HD 7000 VLIW4	HD 8000 VLIW4	second GCN Sea Islands	2nd GCN Sea Islands	3rd GCN Volcanic Islands
GCN Cores	384 SPs	384 SPs	512 SPs	128 SPs?	512 SPs
Chipset	A85X/A75/FCH	A88X/A78/FCH	A88X/A78/Bolton	SOC	SOC
Socket	FM2/FM2+/BGA	FM2/FM2+/BGA	FM2+/BGA	BGA	BGA
Memory	DDR3	DDR3	DDR3	DDR3	DDR3
TDP	17/25/35W	17/25/35W	17/19/35W	10/25W	fifteen-35W
HSA Support	No	No	Aye	No	Full HSA i.0

AMD Carrizo APU Die:

AMD FX-8800P Benchmarks and Confirmation:

One of the slides in the presentation has confirmed the FX-8800P as the flagship Carrizo APU of the lineup which can be seen in the gallery mail below along with the previously leaked benchmark: